I

I

DOT I FAA/PM-86/7 ,I

Program Engineering and Maintenance Service Washington, D.C. 20591

11'£tBN l C:>' i: ::.J."'rn U MAll .(..;,:t.hN.fl' r ··;"1 1'<U :\lU~

. Investigations for Improving Operational Reliability and Maintainability of Instrument Landing System Components

Volume I, Text and Appendices

Richard H. McFarland Walter D. Phipps Larrv D. Bradv

Jeff Dennis Joe D. Longworth

Avionics Engineering Center Department of Electrical and Computer Engineering

Ohio University Athens, Ohio 45701

February 1986

This Document is available to the public through the National Technical Information Service, Springfield, Virginia 22161.

US. Department of Transportation

Federal Aviation Administration

i

l· l \ .

T •o:hniccd R<tport Oocumentation Page

1. Report No.

DOT/FAA/PM-86/7, I

4. Tirlo an<i s.,brirle

,3. R•cil'i-r' s Corolo9 No.

I ' S. Rlf!lcwt Oar•

Fe~ruary 1986

I INVESTIGATIONS FOR IMPROVING OPERATIONAL RELIABILITY AND MAINTAINABILITY OF ILS COMPONENTS VOLUME I, TEXT AND APPENDICES·

~·~----~------~----------4 ~---~----------------------------+ S. p,.,fo,..ing O•gani&atle" Re,.... No. 7. Aurho,l a)

R.H. McFarland, W.D. Phipps, L.D. Brady, OU/AEC/EER 79-1 T<>ff T'l<>nni co · <~nr! .T T'l T.nnc,....rn,.rh

Avionics Engi~eering Center Department of Electrical and Computer Engineering 11. Conr.acr or c;,_, No.

Ohio University nT1?A01 -A1-C-200:' 5 Athens Ohio 45701 1J. Tytte of~ ... .,. -• Pllrioci Cowoweci ~~~~~~~~~~ .. --~------------------------------~ 12. s .. - ... ;,.9 .a.9 ... cy N-· -• 4clcire••

I l

U.S. Department of Transportation Federal Aviation Administration Program Engineering and Maintenance Service Washington. D.C. 20591 IS. Suppl_r,.,., Hates

! i 6. Altsrracr

Final Report Feb. 1983 - Feb. 1986

i Evaluations and performance analysis are presented for Dallas-Fort Marth, TX.; I wheeling, WV.; Bristol, TN.; Parkersburg, WV.; Pontiac, MI.; and DeKalb-Peachtree, I GA. An investigation of anomalous performance of the glide slope at Lambert-I St. Louis is reported. An analysis of Instrument Landing System (ILS) Reference 1 Datum Heights (RDH) is reported, as well as the results of an investigation of

f

glide slope critical areas. A study into the feasibility of developing ground­.based checking techniques for validating the Sideband Reference (SBR) glide.slope

I is presented~ as well as an assessment of the current standards and tolerances used to qualify glide slope structures. In addition, a proposed methodology for

, mathematical mndel validation is discussed .

. 17. ICey Words

ILS, Critical Areas, Glide Slope, Reference Datum Height, Anomalies, Mathematical Modeling, Ground Monitor ing

This document is available to the U.S. public through the National Technical Information Service, Springfield, Virginia 22161

19. Security Claui I. (of this raportl 20. s .. cvrity Clnasil. (of this p•11•l 21• No. of Pages 22. Price

t1nclassified Unclassified 262

Form DOT F 1700.7 ca-72) RopraductiiHI of completed pnge outhorized

· English/Metric Con-version Factors

Length

~ 0 Cm m Km in ft s mi nmi

Cm 1 0.0·1 1x1o-s 0.3937 0.0328 6.21x1o·6 5.39x1o·6

m 100 1 0.001 39.37 3.281 0.0006 0.0005 Km 100,000 1000 1 39370 3281 0.6214 0.5395 in 2.540 0.0254 2.54x1o-5 1 0.0833 1.58x1o·S 1 .37x1o·S

ft 30.48 0.3048 3.05x10"4 12 1 1.89x10"4 1.64x1o·4

Smi 160,900 1609 1.609 63360 5280 1 0.8688 nmi 185,200 1852 1.852 72930 6076 1.151 1

Area

~ 0 cm2 m2 Km2 in2 ft2 S mi2 nmi2

cm2 1 0.0001 1x1o·10 0.1550 0.0011 3.86x10"11 5.1 1x1o·11

m2 10,000 1 1x1o-s 1550 10.76 3.86x1o·7 5.1 1 x1o·7

Km2 1x1010 1x106 1 1.55x109 1 .08x107 0.3861 0.2914 in2. 6.452 0.0006 6.45x1o·10 1 0.0069 2.49x1o·10 1.88x1o·10 ·ft2 929.0 0.0929 9.29x1o-s 144 1 3.59x1o-s 2.71x10"8

S mi2 2.59x1o10 2.59x106 2.590 4.01x109 2.79x107 1 0.7548 nmi2 3.43x1010 3.43x106 3.432 5.31x109 3.70x107 1.325 1

Volume

I~ 0 cm3

Cm3 1 liter 1000 m2 1x106 in3 16.39 ft3 28.300 yd3 765,000 fl oz 29.57 fl pt 473.2 fl qt 946.3 gal 3785

Liter . m3 in3 ft3 yd3

0.001 1 x10·6 0.0610 3.53x1o·S 1.31 X 1 o·S 1 0.001 61.02 0.0353 0.0013 1000 1 61,000 35.31 1.308 0.0163 1.64x1o·S 1 0.0006 2.14x1o·S 28.32 0.0283 1728 1 0.0370 764.5 0.7646 46700 27 1 0.2957 2.96x1o·S 1.805 0.0010 3.87x1o·S 0.4732 0.0005 28.88 o .. 0167 0.0006 0.9463 0.0009 57.75 0.0334 0.0012 3.785 0.0038 231.0 0.1337 0.0050

Mass

~ 0 g Kg oz lb

g 1 0.001 0.0353 0.0022 Kg· 1000 1 35.27 2.205 oz 28.35 0.0283 1 0.0625 lb 453.6 0.4536 16 1 ton 907,000 907.2 32,000 2000

Temperature

°C a 5/9 (Of - 32)

Of • 9/5 (OC) + 32

ii

fl oz fl pt fl Qt

0.0338 0.0021 0.0010 33.81 2.1 13 1.057 33.800 2113 1057 0.5541 0.0346 2113 957.5 59.84 0.0173 25900 1616 807.9 1 0.0625 0.0312 16 1 0.5000 32 2 1 128 8 4

ton

1.10x10·6 0.0011 3.12x1o·5 0.0005 1

gal

0.0002 0.2642 264.2 0.0043 7.481 202.0 0.0078 0.1250 0.2500 1

I

II

TABLE OF CONTENTS

Page No.

List of Tables••••••••••••••••••••••••••••••••••••••••••••••• viii

INTRODUCTION • •••••••••••••••••••••••••••••••••••••••••••••••• 1

Purpose.................................................... 1 Background. . . . • • . . • . • • • • • • • • . • • • . • • • • • • • . . • • • • • . • • . • . • . • • . • 1 Pertinent Sections......................................... 1

ILS ANOMALY INVESTIGATIONS••••••••••••••••••••••••••••••••••• 3

SELECTED SITE ANOMALIES•••••••••••••••••••••••••••••••••••• 3

INVESTIGATION OF ANOMALOUS PERFORMANCE OF THE LAMBERT-ST. LOUIS INTERNATIONAL AIRPORT RUNWAY 30R ILS CATEGORY II GLIDE SLOPE•••••••••••••••••••••••••••••• 3

Summary and Conclusions................................ 3 Background. • • . • • • • • • • • • • • • • • • • • • • • • • • • • . • • • • • • • • • • • . • . . 3 Discussion............................................. 3

Initial Measurements................................. 3 Calculations......................................... 4 Final Measurements................................... 4 Recommendations...................................... 4

EVALUATIONS AND PERFORMANCE ANALYSIS FOR SELECTED ILS SITES 5

EVALUATION OF GLIDE SLOPE PERFORMANCE OF ILS SERVING RUNWAY 18R AT DALLAS-FORT WORTH REGIONAL AIRPORT......... 5

Summary and Conclusions................................ 5 Introduction and Background............................ 6 Data Collection........................................ 7 Data Analysis and Discussion........................... 9 Comments on Order 8240.47.............................. 10 Glide Slope Siting Considerations...................... 13 Recommendations........................................ 14

ANALYSIS OF STRUCTURE ROUGHNESS OF WHEELING, WEST VIRGINIA GLIDE SLOPE. • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 15 .. ELECTRICAL MODIFICATION TO CAPTURE EFFECT GLIDE SLOPE FACILITY TO ELIMINATE REQUIREMENT FOR SITE RELOCATION.... 17

RESULTS FROM EXPERIMENTAL APPROACH INVESTIGATING IMPROVEMENT OF TRI-CITY GLIDE SLOPE PERFORMANCE TO CATEGORY II STANDARDS•••••••••••••••••••••••••••••••••••• 19

iii

TABLE OF CONTENTS (continued)

Page No.

RESULTS OF FLIGHT MEASUREMENTS PERFORMED ON CATEGORY II GLIDE SLOPE SERVING RUNWAY 22 AT TRI-CITY AIRPORT, BRISTOL, TENNESSEE. • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 20

RECOMMENDATIONS FOR RESTORING THE PARKERSBURG GLIDE SLOPE TO SERVICE..................................... . . . . . . . . . . 21

RESULTS OF OPTIMIZATION WORK ON DEKALB-PEACHTREE WAVEGUIDE GLIDE SLOPE•••••••••••••••••••••••••••••••••••••••••••••• 23

INVESTIGATION OF PONTIAC GLIDE SLOPE PERFORMANCE......... 25

Summary and Conclusions................................ 25 Introduction and Background............................ 25 Initial Phase of Investigation......................... 26 Pilot Interviews....................................... 27 Pilot Reports of Interest.............................. 30 Ground-Based Data Collection Setup..................... 30 Flight Data............................................ 31 Tracking of Aircraft................................... 31 Snow Effects........................................... 31 Analysis of Ground Data................................ 32 Recommendations........................................ 32

ILS CRITICAL AREAS••••••••••••••••••••••••••••••••••••••••• 33

THEORETICAL INVESTIGATION OF NULL REFERENCE, SIDEBAND REFERENCE, AND CAPTURE EFFECT GLIDE SLOPE SIGNAL SCATTERING FOR CRITICAL AREA DETERMINATION.......................... 33

Conclusions............................................ 33 Introduction and Background............................ 33 Objective of Work...................................... 34 Approach to Solution••••••••••••••••••••••••••••••••••• 35 Mathematical Model Description......................... 37 Calculation Work....................................... 37 Recommendations........................................ 56

ILS REFERENCE DATUM HEIGHT ANALYSIS........................ 58

DETERMINATION OF AIRCRAFT DELIVERY HEIGHTS AT THRESHOLD FOR COUPLED ILS APPROACHES TO RUNWAY 18R AT DALLAS-FORT WORTH REGIONAL AIRPORT ••••••••••••••••••••••••••••• ·•••••• • • • • • • 58

Summary and Conclusions................................ 58· Introduction and Background............................ 59 Data Collection........................................ 59

iv

TABLE OF CONTENTS (continued)

Page No.

Equipment............................................ 59 Procedure••••••••••••••••••••••••••••••••••••••••;... 59

Data Presentation and Analysis......................... 60 Specific Findings...................................... 61 Recommendations........................................ 61

SITING EVALUATION WITH RESPECT TO REFERENCE DATUM HEIGHTS FOR THE GLIDE SLOPE SERVING RUNWAY 18R AT THE DALLAS-FORT WORTH REGIONAL AIRPORT•••••••••••••••••••••••••••••• 63

Summary and Conclusions................................ 63 Introduction and Background............................ 64 Method of Analysis..................................... 65 Specific Results and Findings.......................... 66 Conclusions............................................ 69 Recommendations......................................... 70

ANALYSIS OF EFFECTS ON REFERENCE DATUM HEIGHTS OF A PROPOSED TAXIWAY ADDITION FOR RUNWAY 9R AT THE CHICAGO O'HARE INTERNATIONAL AIRPORT•••••••••••••••••••••••••••••••••••• 71

Summary and Conclusions................................ 71 Introduction........................................... 71 Analysis and Results................................... 71 Recommendations........................................ 72

EFFECTS OF IRREGULAR PATH-FORMING TERRAIN ON GLIDE SLOPE REFERENCE DATUM HEIGHTS•••••••••••••••••••••••••••••••••• 73

Summary and Conclusions•••••••••••••••••••••••••••••••• 73 Introduction and Background............................ 74 Analytical Method•••••••••••••••••••••••••••••••••••••• 75

Irregular (rough) Terrain............................ 76 Uniform Slopes....................................... 80

Lateral Slopes..................................... 80 Longitudinal Slopes................................ 80

Specific Findings•••••••••••••••••••••••••••••••••••••• 81

Irregular (rough) Terrain............................ 81 Uniform Terrain Slopes............................... 82

Lateral Slopes..................................... 82 Longitudinal Slopes•••••••••••••••••••••••••••••••• 82

v

III

IV

v

TABLE OF CONTENTS (continued)

Page No.

Discussion........................................... 82 Wheeling, WV (HLG)................................. 82 Greenville, SC (GMU)............................... 83

Recommendations...................................... 83

ILS MATHEMATICAL MODELING VALIDATION....................... 85

PROPOSED METHODOLOGY FOR LANDING SYSTEMS MATHEMATICAL MODEL VALIDATION............................................... 85

Executive Summary•••••••••••••••••••••••••••••••••••••• 85 Introduction........................................... 85 Methodology............................................ 85 Implementation......................................... 90 Existing Data Base..................................... 92 Summary and Conclusions................................ 94

USERS' MANUAL FOR LANDING SYSTEMS MATHEMATICAL MODEL VALIDATION PROCEDURE••••••••••••••••••••••••••••••••••••• 98

Introduction........................................... 98 Discussion............................................. 98

EDUCATION SUPPORT FUNCTION••••••••••••••••••••••••••••••••••• 108

SEMINAR ON ENDFIRE GLIDE SLOPE••••••••••••••••••••••••••• 108

LECTURE ON RELEVANT ISSUES CONCERNING CONSTRUCTION PRACTICES AS THEY MAY AFFECT PKOPAGATION OF ELECTROMAGNETIC SIGNALS USED FOR AIR NAVIGATION •••••••••• 108

VERIFICATION OF STANDARDS AND TOLERANCES ••••••••••••••••••••• 109

INITIAL ASSESSMENT OF APPROPRIATENESS OF QUANTITATIVE VALUES USED TO QUALIFY GLIDE SLOPE STRUCTURES •••••••••••• 109

Summary and Conclusions •••••••••••••••••••••••••••••••• 109 Introduction and Background •••••••••••••••••••••••••••• 112 Approach. • • • • • • • • . • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 113 Data Base •••••••••••••••••••••••••••••••••••••••••••••• 113 Data Collection •••••••••••••••••••••••••••••••••••••••• 114 Data Analysis and Discussion ••••••••••••••••••••• _...... 116 Recommendations •••••••••••••••••••••••••••••••••••••••• 119

DESIGN AND DEVELOPMENT OF GROUND-BASED TECHNIQUES FOR VALIDATION OF GROUND FACILITY PERFORMANCE ••••••••••••••••••••

vi

120

TABLE OF CONTENTS (continued)

GROUND-BASED TECHNIQUES FOR VALIDATION OF SIDEBAND REFERENCE GLIDE SLOPE PERFORMANCE ••••••••••••••••••••••••

Summary and Conclusions •••••••••••••••••••••••••••••••• Introduction and Statement of the Problem •••••••••••••• Approach • •••••••••••••••••••••••••••.•••••.•••••••••.•• Data Collection and Discussion ••••••••••••••••••••••••• Recommendations ••••••••••••••••••••••••••••••••••••••••

VI INVESTIGATORS AND ACKNOWLEDGEMENTS•••••••••••••••••••••••••••

VII BIBLIOGRAPHY ••••••••••••••••••••••••••••••••••••••••••••••.••

VI I I APPEND ICES • ••••••••••••••••••••••••••••••••••••••••••••••••••

A. Data from Pontiac Near-Field Monitor ••••••••••••••••••••• B. Data from Pontiac Far-Field Monitor Lower Antenna •••••••• c. Data from Pontiac Far-Field Monitor Upper Antenna •••••••• p. Data from Pontiac Near-Field Monitor at Time of

Page

120

120 121 122 123 128

131

133

139

140 145 150

Crash on December 13, 1984 ••••••••••••••••••••••••••••••• 155 E. Analyzed Data•••••••••••••••••••••••••••••••••••••••••••• 157 F. Landings Tracked at DFW •••••••••••••••••••••••••••••• , ••• 162 G. Measurement Sensitivity Analysis ••••••••••••••••••••••••• 180 H. Effects of Antenna Location on Reference Datum Heights

for an Ideal Glide Slope ••••••••••••••••••••••••••••••••• 182 I. Determination of the Exact Equation for the Glidepath

for Uniform, Lateral Terrain Slopes and Effects on the Reference Datum Heights•••••••••••••••••••••••••••••••••• 203

J. Bias Elimination ••••••••••••••••••••••••••••••••••••••••• 208 K. Terrain Plate Selecting Process •••••••••••••••••••••••••• 211 1. Bias Elimination ••••••••••••••••••••••••••••.•••••••••.•• 218 M. Outline for Endfire Glide Slope Seminar •••••••••••••••••• 221 N. Strip Chart Data ••••••••••••••••••••••••••••••••••••••••• 223 0. Sample of an Analog Strip Chart Recording •••••••••••••••• 237 P. Legend Page for Providing Identification and Quantitative

Assessment of the Analog Strip Charts.................... 239 Q. FORTRAN Program Listings ~or the Computation of Glide

Slope Reference Datum Heights •••• : ••••••••••••••••••••••• 241 R. Maintenance and Refurbishment of Tamiami Test Facility... 258 S. Initial Assessment of Localizer Standards and Tolerances. 261

vii

No.

Table

2-1

2-2

2-3

2-4

2-5

2-6

2-7

2-8

2-9

2-10

2-11

2-12

2-13 -

2-14

2-15

2-16

LIST OF TABLES

Antenna position variables used as input data for calculations.

Basic assumptions and variables used as input data for calculations.

Summary of tolerances applied to all calculated CDI values.

ILS ca~egories for which no critical area maps are presented. No out-of-tolerance locations identified for X ) 50 feet and Y ) 0 feet.

Regions for which no contour maps are presented. Maximum contours are less than 10 ~A.

Critical area vs. aircraft orientation for null reference glide slope, dipole antennas.

Critical area vs. aircraft orientation for sideband reference glide slope, dipole antennas.

Critical area vs. aircraft orientation for capture effect glide slope, dipole antennas.

Critical area vs. aircraft orientation for null reference glide slope, directional antennas (FA-8976).

Critical area vs. aircraft orientation for sideband reference glide slope, directional antennas (FA-8976).

Critical area vs. aircraft orientation for capture effect glide slope, directional antennas (FA-8976).

Critical area vs. aircraft orientation for null reference glide slope, 2-lamdba antennas (GRN-27).

Critical area vs. aircraft orientation for sideband reference glide slope, 2-lamdba antennas (GRN-27).

Critical area vs. aircraft orientation for capture effect glide slope, 2-lamdba antennas (GRN-27).

Results of sensitivity analysis for three terrain depressions.

Results of sensitivity analysis for modification to terrain between antennas and taxiway W-19.

viii

Page No.

39

40

42

44

45

47

48

49

50

51

52

53

54

55

67

68

LIST OF TABLES (continued)

Table Page No.

2-17 t values for calculating confidence intervals. 89

2-18 Rectangular coordinates converted from RTT CD! data. 91

2-19 Type-1 site data base with. capture effect facility. 93

2-20 Type-3 site data base with capture effect facility. 95

2-21 Type-S site data base with capture effect facility 96

2-22 Type-4 site data base with capture effect facility. 97

2-23 Permittivity and conductivity of selected ground 101 conditions.

2-24 Rectangular coordinates converted from RTT CDI data. 105

2-25 Type-1 site data base with capture effect facility. 107

5-1 Data summary for SBR ground monitoring. 129

ix

CHAPTER I

INTRODUCTION

PURPOSE.

This report documents the results of engineering and technical services provided to the Airways Facilities Division of the Federal Aviation Administration during the period of February 25, 1983 to February 25, 1986. These services consisted of assigned tasks in 4 areas:

1. ILS Anomaly Investigations: Studies of ILS facilities reported to be performing in an unusual or unexplainable manner.

2. Education Support Function: Workshops and semin&rs in ILS related subjects.

3. Verification of Standards and Tolerances: A validation of the requirements for ground and airborne operating parameters.

4. Design and Development of Ground-Based Techniques for Validation of Ground Facility Performance: Studies of ILS parameters nor­mally examined by flight checks that might instead be validated by ground measurements.

In addition to these tasks, the ILS test facility at Tamiami Airport (TMB) was repaired and refurbished.

BACKGROUND.

Much of the information contained herein has been previously published in Ohio University Technical Memoranda G-1 through G-10, as well as assorted Technical Precis. These individual memoranda were intended to provide a more timely reporting basis on the individual efforts. This document com­piles the previously published work, and completes the documentation of data which may have been in a preliminary form until this date.

This report is organized into discrete sections presenting information and pertinent results aligned with the contractual tasks as listed above. As such, the conclusions and other data are contained with the applicable tasks and are not repeated in a separate section. In addition, there is some repetition in the introductory material.

PERTINENT SECTIONS.

Chapter II of the report is divided into several sections.

The first section details the results of an investigation of anomalous per­formance of the Category II glide slope serving runway 30R at St. Louis International Airport. The next section documents the results of eva­luations and performance analysis at several selected ILS sites. Then, the results of an investigation to determine critical area requirements for the

-1-

null reference, sideband reference, and capture effect glide slopes are presented. The fourth section presents an analysis of Reference Datum Height. This area contains the results of 4 separate studies. The first documents aircraft delivery heights at threshold for runway 18R at Dallas-Fort Worth while the second presents a siting evaluation of runway 18R with respect to Reference Datum Heights (RDH). This work is followed by a generalized study of the effects of irregular path-forming terrain on glide slope reference datum heights. An analysis of the effects on reference datum heights of a proposed taxiway extension for runway 9R at Chicago O'Hare International Airport is also documented. The final section of Chapter II documents efforts towards ILS mathematical model validation.

Chapter III describes a seminar and a lecture presented under the education support function. Chapter IV presents an assessment of the standards and tolerances used to qualify glide slope performance. Finally, Chapter V documents the feasibility of evaluating the sideband reference glide slope system through a method of ground checks.

-2-

CHAPTER II

ILS ANOMALY INVESTIGATIONS

INVESTIGATION OF ANOMALOUS PERFORMANCE OF THE LAMBERT-ST. LOUIS INTERNATIONAL AIRPORT RUNWAY 30R ILS CATEGORY II GLIDE SLOPE.

SUMMARY AND CONCLUSIONS.

The investigation into the structure roughness problem on the 30R GRN-27 capture effect glide slope facility at the St. Louis Airport has resulted in the conclusion that the glide slope monitor tower for the Mark 1E system located 25 feet from the GRN-27 is the reflection source. The facility was measured with the monitor tower removed. This resulted in a virtual elimi­nation of all roughness on the glide slope structure.

BACKGROUND.

A Wilcox Mark 1E capture-effect glide slope system currently provides ILS Category I performance for runway 30R. This facility is located 375 feet from runway centerline and 1200 feet from runway threshold. This location would not meet the required obstruction clearance criterion for a CAT II facility and thus a location offset by 400 feet was chosen for the CAT II install-ation.

Some structure roughness does exist on the CAT I facility, but the facil­ity remains well within Category I tolerances [1]. After installation of the GRN-27 facility at the 400 feet offset location, precommissioning flight measurements indicated that the structure for this facility would not meet Category II tolerances. In fact, the facility would not meet Category I tolerances. With this information FAA engineers proceeded to investigate the potential reflection sources which might cause the dif­ficulty noted.

In April, 1983, Ohio University was tasked with investigating the 30R glide slope structure roughness problem. FAA flight recordings were reviewed and on April 23, 1983, a series of measurements were conducted on the facility.

DISCUSSION.

Initial Measurements. On March 23, 1983, an Ohio University team traveled to St. Louis to perform measurements on the· 30R CAT II glide slope. The results of these measurements indicated that indeed a severe roughness problem did exist. A digitized plot of the measured system as found is contained in figure 2-1. In addition to the basic measurement, recordings were made on the localizer course edges, above the glide angle, below the glide angle, and on the Wilcox CAT I system located 25 feet closer to the runway centerline.

Each of the measurements resulted in a substantial change in the periods of oscillation. As an example, figures 2-2 and 2-3 show the measured traces for the left of course measurement and the CAT I measurement respectively.

-3-

At the site, no terrain variations or obstacles are apparent which might cause the severe roughness of the measurements. In addition, the periods of oscillation of the measured traces indicate that the potential reflec­tion source is located near the array (less than 1000 feet away). With this in mind it was decided that a series of calculations should be per­formed to determine the source.

Calculations. The first calculation performed was of the CAT II system performance with a capture effect monitor tower located 317 feet in front of the GRN-27 system. Figure 2-4 compares the calculated trace with that measured for the GRN-27. Note that the period of oscillation for the calculation is essentially the same as the measurement. With this infor­mation, the CAT I system was modeled (25 feet nearer to runway, same backset). Figure 2-5 shows the comparison of the calculated and measured traces. Again, the periods of oscillation are very close.

The comparisons made in figures 2-4 and 2-5 indicate that a reflecting structure located near the monitor tower position could cause the path roughness noted. Based upon these comparisons it was recommended that measurements be performed on the GRN-27 with the Mark 1E monitor tower removed.

Final Measurements. On May 31, 1983, Ohio University engineers again visited the St. Louis site to perform measurements on the 30R GRN-27 glide slope. FAA personnel had previously removed the Mark 1E monitor tower, and flight measurements of this configuration were performed.

Five recordings of the facility were conducted, three structure measure­ments and two angle and width checks. A structure measurement is shown in figure 2-6. It is noted that the oscillations of the previous measurements (fi~ure 2-1) are virtually eliminated. The resultant structure reaches less than SO percent of Category II tolerances with no reversals in slope. An angle change of approximately 0.09° was also noted. This is also attri­buted to the removal of the monitor tower and may be corrected through antenna height adjustment.

RECOMMENDATIONS.

From the investigation it is noted that the removal of the Mark IE monitor tower should result in acceptable Category II performance of the facility. At the St. Louis facility the tower will be removed upon the commissioning of the CAT II facility; however, it is recommended that if it is required that this type of structure be utilized as a monitor antenna platform at other facilities an investigation should be made into the glide slope structure roughness.

-4-

EVALUATION OF GLIDE SLOPE PERFORMANCE OF ILS SERVING RUNWAY 18R AT DALLAS-FORT WORTH REGIONAL AIRPORT.

SUMMARY AND CONCLUSIONS.

An investigation has been completed concerning the applicability of FAA Order 8240.47 generally and then specifically with respect to the glide slope serving Runway 18R at the Dallas-Fort Worth Regional Airport, Texas. Additionally, the performance of the 18R glide slope was evaluated prin-cipally by means of flight measurements and data analysis. The conclusions are as follows:

1. The glide slope serving Runway 18R meets the requirements of the U.S. Flight Inspection Manual, FAA Handbook OA P 8200.1 217.5b and should be commissioned.

2. Order 8240.47, while providing a significant benefit in allowing decoupling of the aiming point from the glide slope transmitting antenna mast, has certain aspects which produce unfortunate circumstances, viz, prohibiting the commission of an otherwise acceptable facility. The cause of the unacceptable reference datum height identified by the order is a decrease in path angle of .011 degree over a 3 1/2 mile long path segment.

3. It is not practical to develop a universal order which will pro­vide adequately for sets of airborne equipment that are not designed to a specific standard and which as a result possess considerable variability.

4. There are serious questions as to whether it is proper to extra­polate beyond the data field which has been used in the linear regression analysis. Further there is no known. flight director computer that makes use of such a computation to provide guidance.

5. There are also serious questions as to whether complete desen­sitization of the coupler to glide slope guidance at 100 feet is the safest procedure. It may be efficacious but can also be undesirable, particularly when certified glide slope guidance information for CAT II operations is presently available~

6. Modern technology provides techniques that make desensitizing of airborne equipment not the way to go in producing acceptable flight trajectories; instead, quality, sophisticated filtering of less than that the experienced ideal signals could be expected to do the job of delivering ·the aircraft to the proper threshold location with the proper attitude.

7. Glide slope structures currently allowed by OA P 8200.1 do indeed provide a basic reference for aircraft delivery to the runway, given that some smoothing is provided.

8. A more extensive review of data relevant to CAT I and II opera­tions performed with and without couplers needs to be made to

~s-

place the FAA in a strong position with respect to a new order. Consideration should be given to generating more data using flight simulators.

9. The practice of having some facilities approved under one order and having some approved under another with some not meeting both (in other words, selective application of orders) should be aban­doned because the rationale is difficult to defend.

10. The statistical approach used to arrive at an RDH should be carried to completeness. The RDH is in itself a statistical quan­tity, i.e., it has a distribution and a confidence interval. There are statistical considerations of Type One Errors where one establishes an hypothesis and ·then rejects it because of data. This is then the circumstance when one determines a site to be inadequate when it in fact is. A confidence interval can be cal­culated; this is a-consideration in this work and it should be done. The totality of the statistical domain should be applied.

11. It is desirable to have a prescription for preparing a site with­out resort to an iterative process. Practicality and economics dictate that such a plan exist. Because of the procedures spelled out in Order 8240.47 it is not easy to determine a priori what

~-the optimum site location is for the glide slope. Mathematical modeling of the site with detailed knowledge of the terrain is the only conceivable practical approach to avoid iteration.

12. The glide slope serving 18R at DFW is not optimum because of antenna heights and ground plane preparation. The ground plane is not graded well.

13. The absence of a facility has historically been shown to be a more dangerous condition than a glide slope with a structure or alignment problem.

INTRODUCTION AND BACKGROUND.

This task effort is a response to a condition that exists with the glide slope serving Runway 18R at the Dallas-Fort Worth Regional Airport (DFW), Texas. This runway is new and the glide slope is intended to be ready for Category II commissioning ?sing GRN-27 equipment transferred from another location.

FAA personnel have completed the usual activities preparatory to the planned commissioning of mid-December, 1983. To the concern of many, the facility did not pass the commissioning flight check principally because it did not meet the requirements of FAA Order 8240.47 [2] or the companion order, 8260.34 [3]. These orders, in effect, place new requirements on the performance of the glide slope. The orders specify that the requirements be applied to facilities that are being commissioned and that existing facilities be examined referencing the u.s. Flight Inspection Handbook OA P 8200.1.

-6-

Unfortunately, the path characteristics produced by the 18R glide slope do not meet these specifications. As a result, the much needed new runway is not currently served by a glide slope and certain questions have been posed, the most fundamental of which concerns what to do to obtain a com­missioned facility. Questions concerning whether the terrain at the site, and/or the equipment, is generating problems or whether the problem is related to some inappropriate aspects of Order 8240.47 are addressed in this study. A corollary to the siting issue is that of providing guidance for correcting any siting defects to eliminate the unacceptable conditions as identified by Order 8240.47.

Some additional constraints have been present, particularly those pre­venting consideration of Category I options. Charting of facilities for the operational use and the assignment of common frequencies for the 18L and 18R ILS transmitters essentially preclude the implementation of a Category I operation on 18R without losing the capability of Category II on 18L.

An Ohio University team has pursued a solution to this problem of glide slope acceptability at DFW. The initial approach has been through discussions with FAA personnel in Washington, Fort Ft. Worth, and Oklahoma City and by examination of the site at DFW. Further, and quite impor­tantly, flight measurements were made using a radio telemetering theodolite as a reference to allow in part for a historical perspective and a com­parison of Category II path characteristics found from previous experience.

While DFW Runway 18R is an early example of difficulty with glide slope acceptability under Order 8240.47, it is not the first. In the fall of 1983, Ohio University experienced difficulty with glide slopes at Gillette, Wyoming, and Norfolk, Nebraska, which were not immediately being considered as acceptable ~ith application of Order 8240.~7. This is in spite of the fact that these glide slope structures comfortably met the requirements of OA P 8200.1 for Category I operation [4].

DATA COLLECTION.

Even though the FAA had collected data using an inertial reference, the decision was made to augment this by obtaining data with an optical reference, viz, the traditional RTT (radio telemetering theodolite). Accordingly the Ohio University Mini-Lab Mark Ilia was flown in Beechcraft Model A36, N25688, to collect data relevant to the vertical path structure and the position and character of the glide slope on-course in space. The RTT reference position was surveyed in accordance with OA P 8200.1 217.25. For background it is well to note that this position was developed to pro­vide a hyperbolic-shaped reference path whose asymptote intersects the run­way abeam the antenna mast [5, 6].

In addition the prescribed position provides for minimum error in locating the on-course by making the sensitivities of the electronic and reference paths practically identical by making the axes of the glide slope and reference cones colinear. Admittedly in some cases this is compromised slightly to allow the human tracker to stand at a normal height (62

-7-

inches). The compromise takes place as the theodolite is slid along the ideal reference cone until 62-inch eye-piece height is reached. The direc-tion is along the 3-degree cone in a direction towards the runway threshold to provide for minimum positioning error above the threshold.

A major objective in using the theodolite is to track the receiving antenna on the aircraft and locate, in terms of angular measure, the position of the glide slope on-course. Telemetry by use of the Reaction Instruments UHF transmitter allows onboard real time subtraction of the reference signal from the received glide slope indication to give the difference that is the path position independent of the specific flight track.

By accurate tracking and careful positioning of the flight check aircraft near or on the on-course, very good quantitative information is obtained as to the location of the on-course signal in space. Repeatability is the principal means of ascertaining if problems existed with tracking or flight along the on-course. For this reason multiple runs were made and the ones that repeated were selected for analysis.

Analog strip chart recordings of the differential amplifier (analog of the true path position), the CDI (course deviation indicator) indication, and the theodolite look angle (angular position of the aircraft) were made onboard the aircraft.

Approach speeds used for the data collection were approximately 145 knots. This represents the approach speed of many jet aircraft. The flight con­ditions for the data collections were good. Only light turbulence was encountered with surface winds being westerly 10 to 12 knots. Sky con­ditions were clear with unrestricted visibilities. The tracker, pilot, and electronics panel operator were well experienced in this rather routine operation. Excellent cooperation was obtained from Air Traffic Control, thus minimizing the time airborne and allowing completion of all flight data collection between 1100 and 1230 CST. Two level runs to check the vertical structure and seven low approaches were made.

Data collection was concerned principally with the flight data. With the RTT as reference, both level runs at 1000 feet AGL and low approaches from the outer marker to the threshold were accomplished. In addition, in cooperation with FAA personnel terrain profiles from the site to the threshold and from the site approximately 1400 feet out on a line parallel to the runway were measured.

Data obtained from the FAA-furnished topo charts showed that the differ­ences in elevation were:

Threshold - Runway abeam mast 1.97 ft.

Runway abeam mast - base of mast 3.04 ft.

This latter was confirmed when the theodolite position was surveyed in using ·oA P 8200.1 217.25.

-8-

Also reported were:

Distance of mast back from threshold = 1090 ft. Distance of mast from runway centerline = 400 ft. Distance from threshold to ILS Pt. C = 856 ft. Distance from threshold to middle marker= 2764 ft Distance from threshold to outer marker= 30,896ft Glide slope frequency: 331.1 MHz Localizer frequency: 111.9 MHz.

Figures 2-7, 2-8 and 2-9 illustrate the 18R glide slope site, the tracking reference equipment, and the airborne data collection· equipment.

DATA ANALYSIS AND DISCUSSION.

Two level runs were completed which gave

RUN ANGLE WIDTH 190 \lA SYMMETRY

4-4B 3.04° 0.76° 2.03° 42:58

4-5B 3.03° 0.74° 1.99° --46; __ 54 --- -----

These were acceptable but it was noted that symmetry was very close to CAT II tolerance limits. Comments to the local FAA maintenance personnel con­cerning this prompted explanations that the lower antenna had been moved in an attempt to obtain improvement to meet the tolerances of Order 8240.47 and that this change had degraded symmetry. The indication is that reloca­tion of the lower antenna to a 1.0:2.0 ratio from the 1:2.11 now existing will improve the symmetry. Figure 2-10 is a copy of record 4-5B showing the smooth crossover obtained from measuring the vertical path structure.

Seven low approaches were made while recording outputs of the pilot's cross pointer (CDI), the theodolite indication (angular position of the aircraft) and the differential amplifier (difference of CDI and the reference theodolite). One was deficient because the track was on the nose of the aircraft rather than the antenna. The remaining six repeated well; however, it was possible to select three that were best and of these Run 4-12A is used as the example for this report. Run 4-10A was used for ma~y of the discussions held in Dallas.

Figure 2-11 shows the trace of the differential amplifier for Run 4-12A. Three characteristics are noteworthy. First, in ILS Zone 2 a gradual, almost continuous, decrease in path angle is evident from ILS Point A to Point B. There is a decrease from a high value of 0.07 degree at ILS Point A to a low value of 0.04 degree at ILS Point B. Second, .just before and including ILS Point B, the path has a depression with a maximum value of 0.08 degree. This consistently appears on the records and is a product of an imperfection in the ground plane as. evidenced from the terrain profile

-9-

observed at the Runway 18R glide slope site. Third and finally, the path has a slight flare downward-in the last 3000 feet before the threshold. This produces an on-course which is at 2.89 degrees when referencing the runway elevation abeam the mast and a height above the threshold of 56.6 feet. The actual path angle defined by Zone 2 data is 3.04 degrees.

Calculations using the measured terrain slope and reported heights of the antennas indicate that the path should be 58.8 feet above the runway threshold. The reader should not be confused with these path height calcu­lations. They are not intended to represent TCH (threshold crossing height). Rather they are intended to allow a comparison between measured and predicted values for the on-course position. TCH for this facility is given by

TCH = (1090 tan 3) - 2 = 55.1 feet.

If one disregards the decrease in path angle and the depression evident in Zone 2, the noise in the path structure is found to be approximately 5 microamperes. An· AFIS ILS 3 run would likely show 15 microamperes, however. Irregularities of the path in Zone 3 amount to 10 microamperes from the graphical average path. The worst case in Zones 2 and 3 show that 76% of the tolerance values are consumed.

An examination of the CDI indication at the t~reshold for the three best runs reveals an average displacement of the aircraft high on the path by 15 microamperes as it was hand flown. The data provided by FAA indicated that their flight check aircraft crossed the threshold an average of 51 feet above the threshold as measured by the radar altimeter. To make the com­parison, the 6-foot differential between wheel height of the Saberliner referenced by the radar altimeter and the receiving antenna should be added to the 51-foot reading. Thus 57 feet should be compared to the 59-foot value obtained by Ohio University.

Path values and characteristics obtained and observed from the flight measurements performed indicated that the tolerances of OA P 8200.1 217.5b were met. The only marginal condition appeared in the symmetry.

COMMENTS ON ORDER 8240.47.

The purpose of Order 8240.47 dated May 10, 1983, is to prescribe a method by which the actual flight inspection glidepath angle, a newly defined ILS reference datum height (RDH) and achieved RDH (ARDH), and a ground point o~ intercept can be determined. Historically there has not been a measured value used in qualifying a glide slope facility with respect to the expected height at which the user aircraft will be delivered, other than the Zone 3 structure tolerance. TCH (threshold crossing height) published for the aviator is a calculated value and is not related to final perfor­mance of the facility.

The rationale for the difference between the TCH and the calculated RDH based on flight data is that the TCH is strictly a calculated quantity based on surveyed terrain heights of the base of the antenna mast, the run-

-10-

way elevation abeam the mast, and the height of the runway threshold. The RDH is derived from flight data of the glide slope recording. This recording is of a path which is dependent on detailed terrain features in the reflecting ground plane. The distinction is essentially that of gross versus detailed terrain information.

Obviously, the height at which the aircraft is to be delivered over the threshold is of paramount importance to the safety of the flight operation. ICAO has recognized this in their publication of Annex 10 [7]. FAA has recognized this need for a quantitative determination of a parameter asso­ciated with crossing height and an expected representative descent angle by issuing Order 8240.47.

At this point in time there has been no extensive experience with applica­tion of this ·order. Consistent with history, initial application of this order, as with previous orders of this kind, produces difficulties; certain facilities, while appearing acceptable according to earlier standards, do not meet the new requirements. The question logically to be asked, then, is whether or not the new standard is appropriate and acceptable. Experience has to be the guide in helping to obtain an answer.

There is no dispute regarding the past operational safety of the contem­porary ILS. There has not been a known accident due to an improperly sited or maintained facility in the 40-year history of ILS.

An important additional note is that absence of facilities are considered by some (including this author) to have been heavy contributors to several fatal aircraft accidents.

Application of Order 8240.47 implies knowledge of several factors. Since the order prescribes use of a best fit straight line and this is assumed to be a good representation of the ultimate flight track of the aircraft, good knowledge of the characteristics of the airborne equipment is essential •. Knowledge of this is particularly important in terms of what area should be used in establishing the line of best fit. In other words, should the line of best fit be determined for Zone 2, the first 6000 feet or some other region of the path? The answer is very difficult to give, if for no other reason than that the performance of the airborne systems is different from aircraft to aircraft and from manufacturer to manufacturer. Research to date has not revealed an existing standard.

Further, an important consideration is that the delivery of aircraft using presently commissioned Category II glide slopes is apparently safe and acceptable. The interest should be (and probably is) to eliminate any peculiar, potentially unsafe facilities and to allow others which have heretofore not been commissionable to be made available to users. Again, this availability is very important because, based on experience, .the absence of a facility provides the most adverse impact on safety.

Information obtained during the research for this report reveals that the order has been tested on several facilities. The experience appears to have been different in different cases. In some cases it has taken facili-

-11-

ties that were unacceptable and made them acceptable. In other cases it has confirmed that a facility qualified with OA P 8200.1 is acceptable. In at least two recent cases, one of them being Dallas-Fort Worth Regional Airport Runway 18R, application of Order 8240.47 prohibits commissioning the glide slope thus depriving the user of a glide slope.

The observed variability of response to application of the order is a con­cern. Safety of operation has historically been consistent. Therefore appropriate consideration must be given to a reference for minimizing variability while maintaining a perfect safety record.

The application of this order has the potential of degrading keeping certain facilities from being available to the user. of this discussion is to offer some comments that will allow perspective as to applicability for glide slopes.

safety by The purpose

a better

A few years ago the u.s. Air Force published a report which suggested a linear regression analysis for improving the positioning of the RTT [8]. In 1980, Ohio University published report FAA-R-6750.3, AAF 420 which discussed at length some of the problems associated with theodolite place­ment [9]. One of the problems addressed, that of close-in path shape, would be alleviated or eliminated if the reference theodolite placement would be decoupled from the glide slope antenna mast. Order 8240.47 accomplishes this goal. This provides a flexibility not previously available. It allows for shaping the path essentially by allowing flexibi­lity in the point from which it is viewed. The specification, of course, must be that the intended point of landing be safely on the runway.

The concept of applying a linear regression analysis to landing system guidance has some attractive as well as some unattractive features. For example this type of analysis permits the aiming point for the aircraft to be at any location with respect to the transmitting antennas. On the other hand, it involves definition of a point outside the data field (extrapolation), whereas some authorities recommend that use of the regression analysis be confined to interpolative use only.

The concept of developing a s.traight line analysis appears to have virtue since an aircraft on an approach desirably flies a straight line until flare for landing. The earlier concept of using a hyperbolic line was simply one of convenience, since the hyperbola was inherent in generating the glide path under ideal conditions. Reported as a motive for developing a straight line based on early flight track history is the fact that many approach couplers desensitize to glide path information as a function of radar altitude information. Research into this scheme reveals that there is reasonable probability that it can lead to disaster.

First, there is no known consistency among the flight director or coupler designs and the programs for desensitization. Some are based on radar altitude, some on time, and some do not desensitize at all.

Further there is no standard or specification for the design, and accep­tance is based on satisfactory flight performance during a number of

-12-

approaches. There is no assurance that all possible terrain profiles under the approach paths ca~ be accommodated simply because they have not been tested. Design of a standard for a guidance signal with these variables imposed must be regarded as difficult at best.

It is paradoxical that with new designs making use of glide slope infor­mation which is replaced by radar altimeter guidance information that a standard has to be applied to early-path shape or orientation. There appears to be no flight director system that uses inertial references to permit calculation of a line which will provide guidance for the last segment of the descent. ·

There is a condition reported with couplers in which the system is desen­sitized to the glide slope almost totally at ·100 feet elevation, and no radar altitude is available to the coupler for continuation. The pilot, whether he knows it or not, is on his own and has only his visual cues to descend safely for landing.

GLIDE SLOPE SITING CONSIDERATIONS.

Over the years there has been continuous concern that installation person­nel be able to place a facility in a specific location on an airdrome with the ultimate result that a commissioned glide slope will become available. Siting manuals have been produced to serve as guides and in most cases these have been useful.

Most everyone is aware that the glide slope produced ·in space is very much dependent on the character of the reflecting ground. Unfortunately, unless one knows in considerable detail what the precise character is, he cannot accurately place the facility and produce the optimum path.

The advent of Order 8240.47 adds some considerations which make it all the more important to know the character of the path even before the foundation is placed for the antenna tower. This order prescribes new quantities of ARDH (achieved reference datum height) and RDH (reference datum height) which have tolerances and result from measured data. This measured data, of course, are dependent on existing terrain.

The only practical approach to producing a path which meets Orders 8240.47 and 8260.34 is to mathematically model the facility using detailed terrain information. The predictions of the model can be made to include ARDH and RDH. If the terrain is known well enough, then the values of RDH and ARDH can be determined. If they are found to be unacceptable the facility can be moved inexpensively in the modeling process.

Anyone having long time experience in siting glide slopes knows that it is at times very expensive with the present constraints. The additional requirements of the two orders cited makes the siting operation even more speculative if modeling is not used. Iterative processes involving flight operations will be regarded by many as uneconomical and impractical, based on available flight time.

-13-

The bottom line is that without detailed terrain information and extensive calculational efforts, it is not possible to prescribe exactly where a glide slope facility should be placed to meet the new orders.

RECOMMENDATIONS.

The following recommendations are based principally on the work performed during evaluation of the performance of the 18R glide slope at Dallas-Fort Worth Regional Airport and the experience of working with placement of reference systems beginning in 1969.

1. Place the implementation of Order 8240.47 in abeyance until cer­tain analyses and investigations are completed.

2. Commission the DFW 18R glide slope under Order OA P 8200.1 because experience shows that glide slopes with that structure are safe and that unavailability of glide slopes at times result in decreased levels of safety.

3. Develop and promulgate standards for airborne flight directors and couplers to obtain optimum performance and safety and to insure uniformity.

4. Modify the present Order to recognize the standardized designs that will be forthcoming due to the standards discussed in 3 above.

5. Continue the concept of decoupling the reference point from the glide slope mast because airborne flight check computational capa­bility is present to allow this.

6. Modify the antenna heights of the glide slope serving 18R to a 1.0:2.0 ratio which should produce improved symmetry.

7. Improve the grading at the 18R glide slope site to give uniform, minimum slopes.

8. Make maximum use of flight and computer simulation to examine the allowable limits and standards for ILS and MLS.

9. Provide a thorough foundation and insure that it is mathematically rigorous if a statistical approach is going to be used for an order.

-14-

ANALYSIS OF STRUCTURE ROUGHNESS OF WHEELING, WEST VIRGINIA GLIDE SLOPE.

During mid-year 1984, the glide slope transmitting facility at Wheeling, WV, was moved 200 feet further back from the threshold to produce a higher crossing height for the glide slope on-course serving Runway 3 at the Ohio County Airport, Wheeling, wv. The approach plate indicates a threshold crossing height of 32 feet.

FAA flight inspection found when attempting to recommission the system that the structure in particular was out of tolerance. In addition, application of FAA Order 8240.47 precluded commissioning because prescribed tolerances could not be met.

An Ohio University measurement team proceeded to Wheeling on August 21, 1984, and obtained the following data which are presented with data the FAA had obtained earlier.

Path Angle Path Width A~H RDH

Ohio 3.01 0.60 59' 76'

FAA 3.03 0.72 57' 70'

Additional information can be obtained from figure 2-12 which reveals that there are only minor non-linearities in the vertical structure of the ·path. Figure 2-13 shows the glide slope structure as measured with the new transmitting site 1250 feet back from the threshold. Clearly, there are major path excursions pressing or exceeding the Category I tolerances. The theodolite (RTT) reference was set as prescribed in the u.s. Flight Inspection Manual 8200.1 217.25. Historically, the structure was found to be within tolerance but not by a great margin.

The site environment is shown in figures 2-14 and 2-15. The long row of trees running parallel to the runway, 50 feet to the outside of the glide slope mast, is evident. This tree line is suspected to be the principal cause of the glide slope roughness because 1) as the site was moved back, a greater surface area of trees appeared in front of the transmitting anten­nas, 2) the trees being close to the transmitting site produce a low grazing angle which supports a reflecting capability, and 3) the flight recording of path structure recorded in line with the trees shows the roughness to be eliminated.

Figure 2-16 shows the flight recording made in line with the row of trees. Except for the excessive flare, which is due to the flight not tracking the reference cone established for approach to the runway, the structure is extraordinarily good.

The conclusion is that the Wheeling glide slope will be improved to be well within tolerances by accomplishing a modification to the tree line.

-15-

Complete removal of the trees might not be required. This would aid in minimizing impact to the natural environment.

-16-

ELECTRICAL MODIFICATION TO CAPTURE EFFECT GLIDE SLOPE FACILITY TO ELIMINATE REQUIREMENT FOR SITE RELOCATION.

The existence of a 5 to 20-foot high runway pedestal associated with a glide slope site will present apparent path structure problems. For example, an evaluation of performance of the Greenville, SC, Downtown Airport glide slope, serving runway 18 located on an 18-foot pedestal, revealed that an out-of-tolerance condition produced by the pedestal exists. Fortunately, application of current technology can make it accep­table. This is achieved most simply by advancing the phase of the current in the upper antenna by 20 degrees. An alternative is to move the facility further down the runway.

The Greenville image glide slope site appears to be a very poor one. The base of the antenna mast is 18 feet lower than the runway, which is on a pedestal that extends forward all the way to include the threshold. The reflecting plane for the path is not smooth and has a downward slope. Additionally, the runway slopes downward 8.7 feet from abeam the mast. All in all, one has to be an optimist to believe that any image facility at this site can meet flight inspection tolerances.

Practically, a pedestal can be accommodated by moving the facility further down the runway to allow threshold crossings at an adequate height [10]. Presently, with the new FAA Order 8240.47, the reference no longer need be associated with the antenna mast. The reference can remain fixed, while the path is adjusted to meet tolerances associated with the reference or the opposite, the reference can be moved leaving the facility fixed. The constraint is that the reference datum height (RDH) must be maintained within tolerances. With a pedestal, adjustment of the path is sometimes accomplished by relocating a facility, but in the past, moving the facility meant moving the reference and no net gain was generally achievable [11].

FAA flight checks showed the Greenville glide slope flared downward, out­of-tolerance at point B. This is expected because the old reference located abeam the mast established an ideal hyperbola which was approxima­tely 18 feet higher than the actual path. Be~ause of the expense of relo­cating the facility, consideration was given to moving the reference forward which is appropriate under Order 8240.47 provided the crossing height, viz, RDH is satisfactory.

An Ohio University team made measurements of the path on November 14 and 15, 1984. Results confirmed the flare downward due to the reference loca­tion abeam the mast; however, ground plane slopes also are producing a progressive decrease in path angle as the aircraft moves through ILS Zone 2. Figure 2-17 shows a trace of the differential amplifier output when measuring the path against a reference located 600 feet back from the threshold.

Relocating the reference, of course, will not correct for the Zone 2 flare­down because the path structure of interest is too far from the touchdown zone.

-17-

Along with the linear regression technique prescribed in Order 8240.47 is the requirement that the elevation of the regression line, viz, the RDH, is at least 30 feet high over the threshold. The minimum height of 30 feet is prescribed in FAA Order 8260.34 for runways serving general aviation, small commuter aircraft, and corporate turbojets.

The RDH for the path shown in figure 2-17 is 24.4 feet. This is produced even with the path angle increased to 3.15 degrees. The reader should note that the RDH is independent of the tracking reference location. The non­congruence of the geometries of the tracking system and the electronic glide slope, which in the past mandated the link between the reference and the transmitting antenna mast, has not been_ignored in the data collection. Special concern was given to flying the Greenville glide slope on-course closely.

Because of the persisting low RDH, electronic modification of the system was implemented as motivated by theoretical, calculational, and practical studies of the Kansas City Category II glide slope performed in 1979 [12]. Similar benefits were achieved at Shreveport, LA in 1981 [13].

Results of the modification can be seen in figure 2-18. The path most noticeably has lowered in angle in Zone .2, which is consistent with expec­tations. The path ang~e, which had earlier been raised 0.15 degree, pro­vided compensation. The resulting lower actual path angle was measured as 3.05 degrees. Most importantly, however, is the slight rotation of the path in Zone 2, with the far out portion lowering and the close-in portion rising. This makes for a more consistent angle throughout the zone, and an RDH of 33.1 feet. This is clearly within tolerance.

Conclusion. Electronic adjustment of a capture effect glide slope facility can produce changes in the path which allow tolerance to be met without physical relocation of the facility or terrain modifications. The FAA Order 8240.47 allowing a divorcing of the reference path from the antenna mast is a benefit in permitting facilities to meet tolerance.

-18-

RESULTS FROM EXPERIMENTAL APPROACH INVESTIGATING IMPROVEMENT OF TRI-CITY GLIDE SLOPE PERFORMANCE TO CATEGORY II STANDARDS.

An experimental investigation was performed on February 28 and March 1, 1985, for the purpose of investigating the achieved ILS reference datum heights and correcting the problems that were preventing the glide slope serving runway 22 at the Tri-City, Bristol, TN, Airport from meeting Category II standards. FAA flight inspection had rejected the facility based principally on structure roughness being outside limits for the 30-20 microampere, tapered tolerance section of ILS Zone 2.

Success was achieved when good quality Category II performance was obtained through the use of new phase settings and reduced sideband power. Structure values were reduced to 65% ·of allowable tolerances, with ARDH and RDH values of 54.6 and 50.2 feet respectively.

An Ohio University team travelled to the Tri-City Airport on February 28, 1985, and performed measurements with the Mini-Lab Mark Ilia flown in a Beechcraft Model 36. Ground reference was provided by means of a Warren Knight WK-83 radio telemetering theodolite and a Reaction Instruments tele­metry transmitter. The signal standard was an IFR 401-L with calibration traceable to NBS.

Figui"a .. 2-l9_ shows the initial as-found conditions. Figure 2-20 shows the final structure produced by changes in the phase and sideband power.

A check of M-array performance revealed it to be remarkably good with no false courses appearing below the 3.00 degree path angle and· good width and structure being available.

The site is not an extremely difficult capture effect site. However, because the site has a bare, good-reflecting upslope in front of the array, it is important the the capture effect system be properly phased to effect the needed signal cancellation. Several different phase settings were exa­mined by means of the_ flight measurements. Some of these settings were those on record from previous FAA data gathering sessions. The final set­tings were those derived from FAA records and which produced the very good structure shown in figure 2-20. Further examination of the performance with these settings indicated that the below-path signal cancellation was reasonable, the 15-degree middle antenna dephasing tests were good, and attention of the upper antenna by one dB produced no significant degrada­tion.

-19-

RESULTS OF FLIGHT MEASUREMENTS PERFORMED ON CATEGORY II GLIDE SLOPE SERVING RUNWAY 22 AT TRI-CITY AIRPORT, BRISTOL, TENNESSEE.

On October 18, 1985, an Ohio University flight crew traveled to Bristol, Tennessee, for the purpose of evaluating the performance of the glide slope serving Runway 22 at the Tri-City Airport. FAA flight inspection had downgraded the facility to Category I after measuring out-of-tolerance structure in ILS Zone 2.

Earlier in 1985, Ohio University performed work on February 28 and March 1, at Tri-City to restore Category II performance. At that time a maximum structure value of 65% of tolerance was obtained with reduced sideband power and new phase settings.

Work was performed on the station. since March, because it was believed that some of the system components were causing system parameters to drift in value. The changeover relays and phasors were replaced. Some line lengths were changed to ensure that the nominal phase settings occurred at the center of the phasing scales.

A level run was made which showed a path angle of 2.97 degrees and a path width of 0.68 degree. A low approach was made to measure the structure. This recording is reproduced in figure 2-21, showing that measured struc­t~re is only 50% of Category II tolerances.

TABLE OF VALUES ON GLIDE SLOPE SERVING RUNWAY 22

AT TRI-CITY AIRPORT, BRISTOL TENNESSEE

PARAMETER MEASURED VALUE CAT II TOLERANCE/LIMIT

Angle 2.97° 2.95 - 3.05°

Width 0.68° 0.65 - 0.75°

Symmetry 51% - 49% 67% - 33%

Structure -11 llA 30 + 20 J.IA

Conclusion. The glide slope performance at Tri-City Runway 22 is comfor­tably within Category II structure tolerances.

Recommendation. Restore the glide slope to Category II operation.

-20-

RECOMMENDATIONS FOR RESTORING THE PARKERSBURG GLIDE SLOPE TO SERVICE.

On June 4, 1985, an FAA flight inspection crew flew one of several attempts to restore the Parkersburg (PKB) null reference glide slope to service, following the station's modulator and antennas change-outs. Flight inspec­tion declined to restore the facility due to lack of repeatability of width valves taken from level-type, data-collection runs.

On June 14, 1985, Ohio University responded to an FAA request to collect data on the PKB glide slope problem. Measurements were made and the data analyzed. Recommendations derived from these data are:

1. The PKB facility is safe and should be restored to service by flying the glide slope consistent with FAA Flight Inspection Handbook 8200.1 217.3308. This permits the average value of width to be taken from low approaches, 75 microamperes above and below the path.

2. The facility should be converted to a capture effect type system because the ground plane is too short to serve well for a null reference type. The second system choice is a sideband reference type.

A variety of flight measurements was made using the Ohio University Mark Ilia Mini~Lab operating in a Beechcraft Model 36. References for the measurements were made through the use of a Warren Knight WK-83 Radio Telemetering Theodolite.

The modulation balance was checked and the phasing was adjusted. Sig­nificant data were collected using level runs at 1000 and 1500 feet AGL. These data clearly indicate that the vertical structure of the glide slope contains many non-linearities, making it difficult to obtain repeatable flight data on the vertical. path width. Variations in width from 0.64° to 0.73° were observed. Interestingly and in contrast, path structure data obtained on low approaches indicated values in ILS Zone 2 of only 55% of Category I tolerances.

The ground plane serving the PKB glide slope was created by providing a monstrous earth fill. The ground plane is reported to be 930 feet in extent in front of the glide slop~ mast. Experience has shown this to be unsatisfactory in producing quality signal in space from a null-reference system. Undoubtedly aggravating the problem of the short ground plane is a lip at the edge rising approximately one wavelength above the ground level. This likely serves as a knife edge diffractor.

Because the FAA in the past has experienced difficulty in obtaining repeatability in values from width measurements obtained from level passes, a special procedure was conceived and published in paragraph 217.3308 of the Flight Inspection Handbook. In essence, this describes an alternative, averaging process through the use of low approaches 75 microamps above and 75 microamps below the normal angle of the glide slope. The procedure is similar to that used for obtaining a single number value for the path

-21-

angle, the actual path angle taken as the arithmetic mean for path angle values obtained in ILS Zone 2. The arithmetic mean for the path, the mean for approximately .35° below the path, and for .35° above are determined and the lower subtracted from the upper in order to determine the course width. Symmetry is obtained by comparing the values referencing the on­course. While it is recognized that thi~ averaging method is more time­consuming and expensive, the criticalness to flight safety of the glide slope encourages the use of this procedure on certain occasions. To mini­mize long-term commitment to this more expensive data collection process, the capture effect system should ·be installed. This system is known to be capable of operating with a short ground plane [14, 15].

Path angle values obtained during the level runs at PKB were 3.04°, 3.05°, and 3.10°. A 3.05° value repeated for the actual path angle on two low approach runs. Width values of 0.64° and 0.73° were measured on 1000-foot AGL runs, and 0.63° was obtained from a 1500-foot AGL run. Corresponding symmetries were 56/44 and 38/62 which clearly indicate gross non-linearity by the inversions of the dominant section. Structure angles observed were 2.07° and 2.05°; therefore, there is no pro~lem with clearance. Note should be made that the PKB localizer is offset to the right of the runway centerline.

Conclusions. Non-linearity in the vertical path structures at PKB produced non-repeatability in path width and angle values taken on level runs; data indicate that path structure disturbances and non-linearities increase with elevation angle; the path structure near 3° is remarkably smooth, reaching only 55% of tolerance; all measurements indicate that the system is per­forming to give a safe path in space; and the present problem is created, in part, by flight data collection which uses level runs rather than an averaging process on data taken from a low approach.

' -22-

RESULTS OF OPTIMIZATION WORK ON DEKALB-PEACHTREE WAVEGUIDE GLIDE SLOPE.

The waveguide glide slope serving Runway 20L at the DeKalb-Peachtree Airport (PDK) historically has operated close to CAT I structure tolerance limits as prescribed by the u.s. Flight Inspection Manual. Efforts by the FAA and Ohio University in November, 1984, resulted in slight improvement; however, an apparent recent deterioration motivated further work. On October 7 and 8, 1985, an Ohio University team working with the FAA regional and sector personnel was able to reduce the path structure rough­ness from 97% to 83% of tolerance limits by increasing the coning from 8 to 10 dB, and moving the waveguide base forward 3 1/8 inches. This move com­pensated for the .15 degree lower path angle produced by the coning adjust­ment. Ohio University made the flight measurements with the Mini-Lab Mark I·Ila, flown in a Beechcraft Model 36. A radio telemetering theodolite was used as a reference.

When there are variations in the three-dimensional glide slope such as exist at PDK, good repeatabilities in path values can be difficult to obtain. Twelve ·low approaches and 6 level runs were made of the as-found conditions. These served as baseline references for determining whether improvements were to be obtained.

A site inspection was performed and it became apparent that a relatively smooth area of the earth's surface existed from 800 to 1200 feet in front of the antenna. This surface has the capability of reflecting lower side lobes of sideband-only radiation into the approach path. Ground measures were made of the incident radiation from the sideband system, and several electrical changes were made to minimize this radiation. The most success­ful was to retard the auxiliary antenna radiation by 15 degress with respect to the main antenna. This condition was flown and the amount of reduction of structure roughness was within the amount of variability in the measurement.

A second approach was investigated, that of increasing the coning from 8 to 10 dB. Improvement was noted, and the decision was made to change the back-tilt of the waveguide to restore a 3-degree path angle. Following this, three runs were made which produced structure values 87%, 83%, 83%, of tolerance.

' The principal aberration in the path consists of an oscillation which requires application of the reversal tolerance. Results of this indicate that this tolerance is not met. This will require a restriction for coupled approaches.

Symmetry was found to be 54/46 (90 Hz/150 Hz) and is well within tolerance. Figure 2-22 shows the differential amplifier trace representing the path before the 10 dB coning and antenna tilt change, and figure 2-23 shows the results. Width values were 0.71° and 0.67°, respectively.

The conclusion based on this work is that significant improvement can be obtained by iacreasing the coning and decreasing back-tilt of the antenna. One strong recommendation is to scarify the smooth earth from 500 feet to

-23-

1500 feet in front of the antenna. This will break up the reflected signal which contributes to path derogation.

There are no significant rapid path oscillations to suggest existence of multipath sources off-centerline-from the antenna.

-24-

INVESTIGATION OF PONTIAC GLIDE SLOPE PERFORMANCE.

SUMMARY AND CONCLUSIONS.

A variety of pilot reports and several aircraft accidents which occurred during approaches to the Oakland-Pontiac Airport over the past five years has motivated an investigation of the performance of the ILS glide slope serving Runway 9R. The appearance was that there was an intermittent malfunction in the glide slope signal which caused at the very least pilot comment, and at the worst, accidents. FAA flight checks and maintenance­could not substantiate any of these claims of malfunctions.

Early in 1984, Ohio University began a special investigation and evaluation of the facility which lasted nearly 15 months. The investigation included special flight measurements, review of logs and records, interviews with operators and members of air crews, implementation of a special receiving (monitoring) and recording facility, and analysis of over 25,000 hours of analog, strip chart, glide slope data. The conclusions reached are as follows:

1. There has been no aircraft accident due to a malfunctioning glide slope at Pontiac.

2. All technical data indicate that the Pontiac glide slope operates as a typical facility which serves the aviation user flawlessly.

3. There do not appear to be any problem areas with the Pontiac glide slope facility or its performance.

4. Pilot reports can, in certain instances, be totally incorrect with respect to glide slope performance. Two reasons for this are that subtle air currents may displace the aircraft from the position believed to be correct by the pilot, and airborne equipment may at times give erroneous indications.

5. While radio frequency interference conceivably could be a cause of problems for a user of the glide slope, there is no evidence that this does exist.

6. Anomalous glide slope operation, in one important instance involving an aircraft accident in which the pilot alleged glide slope anomaly, can be traced to a cockpit instrument malfunction.

7. The Mark 1B glide slope transmitting equipment produces a very stable signal in space.

INTRODUCTION AND BACKGROUND.

The Oakland-Pontiac Airport has 4 runways one of which, 9 Right, is served by an instrument landing system. The airport is principally a general aviation airport with at least one Part 135 operator, PDQ Executive Air Service, which has nightly runs carrying for the most part cancelled checks.

-25-

Over the past 10 years, there have been a number of accidents or incidents involving aircraft operating at or near the airport. At lease two resulted in claims for damages being filed against the government. Since these two, plus several others, involved aircraft making an approach to landing, con­cern developed over the operation of the ILS, particularly the glide slope.

Of special interest, was an accident involving a Beechcraft Model 18 which crashed into Lake Pontiac, .4 mile from the runway, during a final approach. Concern over a finding by the after-accident flight check that the broad alarm limit was 0.92 degree instead of 0.90 or.less, and their listing it as a serious discrepancy, could not have affected the flight of the accident aircraft. There are several reasons. First, the amount of excess beyond tolerance is within the measurement tolerance of the flight check aircraft, and is not significant quantitatively. Second, the flight check finding was that the path width was at 0.73 degree and was normal, and not near the tolerance limit.

Reports obtained by Air Traffic Control (ATC) on May 19, 1983, from two aircraft of "pitchdowns" while on the approach 1 1/2 miles out, and two more reports of anomalies on May 22, 1983, resulted in a systematic inquiry of pilot users by ATC on June 3, 1983, as to their impressions of the glide slope performance. Thirteen comments received from solitations on that IFR day, suggested a variety of conditions might exist ranging from upswings of the_ needle, to fluctuations, to downward deviations. There was no con­sistency to the comments with the result that no firm conclusions could be drawn concerning the performance of the glide slope. Because there was concern, ATC began issuing an announcement to all pilots who began an ILS approach when the ceiling and visibility were less than 1000 feet and 3 miles respectively. This announcement was, "Use caution for possible glide slope deviations."

The glide slope is a null reference type with Mark 1B electronics. The transmitting antennas are the APC FA 8976 type with 3 dipoles arranged in a colinear array and backed by a corner reflector. The ground plane is large and essentially flat. In the recent past, a taxiway was completed from some hangars northwest of the airport to feed Runway 9R, and this cut per­pendicularly in front of the glide slope, 500 feet distant. Except for the taxiway, the Pontiac glide slope, by most any standard, must be considered a typical null reference type facility.

INITIAL PHASE OF INVESTIGATION.

Initially the FAA's interest was to have an inspection of the facility by Ohio University, a review of flight check records, a check of any pilot contacts that could be made, and a review of maintenance records with an interrogation of maintenance personnel. All of this was to be done with the purpose of uncovering any evidence that could be related to possible production of intermittent fly-down commands in the approach region at or near 3.0 degrees elevation.

This process was begun by Ohio University making visits to the Pontiac site, flying the glide slope and interviewing personnel at the PDQ

-26-

Executive Air Service, the FAA ATC facility, and the FAA glide slope main­tenance. No peculiar or abnormal items were uncovered. Further, a visit was made to the the Battle Creek Flight Inspection Field Office, and a review of records was made. There were no anomalous conditions found even though they had increased the frequency of their flight checks to once every 30 days.

Attempts to contact pilots who had reported anomalous conditions were not successful in producing useful information.

On April 4, 1984, a Piper Aerostar Aircraft, piloted by a PDQ staff pilot, impacted tree tops near the middle marker as the flight proceeded on the approach to the airport. Damage to the gear re~ulted so that when the landing was attempted on the runway, the aircraft was further damaged and disabled. The pilot reported that she had observed a fly-down indication on the glide slope, had followed it, and impacted the tree tops.

Upon learning of the event, Ohio University proceeded with two courses of action. One was to arrange to interview the pilot, and the other was to set up continuous glide slope monitoring and recording facilities at the Pontiac airport. This began the second phase of the investigation.

On December 11, 1984, a Piper Navaho crashed during an ILS approach, killing the two men aboard. On December 13, 1984, an Ohio University engineer met with an FAA Regional engineer at Pontiac, inspected the ILS sites, inspected the accident site, flew the ILS, and collected and anno­tated the three strip chart records for the time of the accident. The recording system was calibrated. There was no indication of any anomalous performance of the ILS. In fact, the recordings showed reflected signals from three other aircraft which made approaches close in time (approximately 3 minute intervals) to the accident. Two eventually missed the approach and one was a practice approach. None completed a landing. The official weather was reported to be 400 overcast, 2 miles visibility in rain.

P-ILOT INTERVIEWS.

Although there had been many comments recorded by the FAA from pilots, these were generally recorded by ATC personnel, most of whom were not pilots nor familiar with the ILS glide slope. When pilots who had experience with flight on the Pontiac glide slope were contacted, no useful information could be obtained. Because of the trauma associated with the Aerostar accident which heightened awareness of details, and the oppor­tunity to interview only 8 days later the pilot who was not injured, great care was taken to obtain a maximum amount of information which related to glide slope operation.

The Aerostar accident occurred at 0730 EST when the weather was reported to be indefinite ceiling 100 feet obscured, visibility 1/2 mile in fog. Interestingly, an aircraft landed just a few minutes ahead of the Aerostar approach time and one just shortly after. The Aerostar clipped the tops of the trees, ingesting brush into the nacelles and damaging the leading edge

-27- .

of the right wing and the right landing gear scissors. Because of the latter, after reaching the runway and with what might have otherwise been a safe landing, the right landing gear collapsed causing the aircraft to swerve off the runway. The principal damage was due to this landing.

The pilot reported she had a problem with the glide slope indication because it suddenly showed a rapid fly-down indication, and she indicated that corporate pilots and her chief pilot had also had this experience. She appeared to accept fault for the impact because she said that she had responded to a spurious signal. When she arrived for the interview, she had a book with her [16] and said that if she had studied this book, she

_ would not have had the accident because the book points out that you have to be alert for peculiar things that happen to the ILS. When questioned in detail concerning this, she could not identify the specific reference and she concluded that it had to do with glide slope characteristics being dependent on ground plane topography.

The approach was reported to have started well, with an intercept at the outer marker at the published altitude of 2700 feet MSL. The aircraft was stabilized on the approach with deviations no greater than ± one dot (75 microamperes, 1/2 scale). She knew the weather conditions were low because the aircraft which had just landed reported that the ceiling was "right at minimums." She said that she should have known that there was a message in that report. She admitted that she did not have a lot of experience with low-ceiling conditions.

As the aircraft approached the middle marker (she did not remember hearing the middle marker tones), the pilot observed the glide slope needle moving rapidly downward. She followed this indication, and observed the needle to move slightly upward just before impact with the trees. She did not know her altitude when the indicator began to move downward. The point of impact was below the decision height for the airport.

The glide slope needle movement was rapid, but not a snap or a step. It was rapid enough, however, that she feels that she should have recognized it as an improper signal. It went full scale fly-down, and then began to recover. She was watching the instruments when she hit the trees; she does not remember for certain what the indication was at the moment of impact, but she believes that the needle was recovering and moving upwards.

During the time of the approach, the air was smooth and the wind was from the east at 16 knots. The pilot reported that she did not have to make any power adjustments to hold the 120 to 130 knots during the approach. The aircraft was lightly loaded.

At no time was the pilot able to provide a cross check between the glide slope indication and visual indications with the ground. When she heard the impact with the tree tops, she looked up and proceeded to follow the approach lights to the airport. After acquiring visual contact, she did not further observe the glide slope needle.

The information provided indicated that there was no complaint concerning the localizer. The needle was kept centered during the approach. The

-28-

impact point in the trees was approximately 100 feet north of the middle marker.

The pilot stated that she did not want the impression left that the acci­dent was not her fault. She said more than once that she should have moni­tored the altimeter; she accepts the responsibility for not doing that. She feels that the ILS's have imperfections, and that she should have known this and flown accordingly. She felt she "dropped her guard."

In response to a question concerning her attitude now towards the Pontiac glide slope, she said she is now aware that there can be problems and she will now monitor altitude and not be trapped. She said she feels this way about any glide slope because of the accident. The implication seemed to be that there are bad glide slopes, and in response to a question as to where there might be bad glide slopes, both she and her chief pilot said that they know Pontiac to be a bad one because "the tower tells them this." The reference is to the cautionary announcement that was provided each pilot prior to beginning an approach.

A standing invitation was given to the PDQ chief pilot, assistant chief pilot and officials of the company to provide any information concerning anomalous indications produced by the Pontiac glide slope.

Within 3 weeks following this accident, Ohio pniversity placed into opera­tion three recording systems for obtaining·continuous data on glide slope path and below path clearance performance.

On February 6, 1985, during a routine check for comments, a Vice President of PDQ (a non-pilot) advised that their president had reported to him obtaining a fly-down indication while making an approach, much the same as their pilot who had flown into the trees. The President of the company was later interviewed by phone concerning his experience.

The president of the company advised that he had received the fly-down indication at Pontiac while making an approach with a ceiling of 600 feet. He went on to say that he had also experienced this condition at Columbus, Ohio, and because of this repeat observation, had taken the aircraft to the Electrosonics Corporation (an avionics repair station at Port Columbus) for an equipment check.

With his permission, a follow-up with Electrosonics was made and interviews were conducted with both of their service managers. A malfunction had been found with the King KI 525A Horizontal Situation Indicator (HSI) which was manifested with intermittent operation of the glide slope indicator. Bench tests showed that after 20 minutes of operation, the glide slope needle evidenced a floating condition due to a thermal intermittent from heat generated by the instrument.

This KI 525A indicator was replaced in the aircraft, and the defective unit returned to the factory. Inquiries of the factory personnel produced no useful information. From comments made by King personnel, the problem is not a chronic one, but rare.

-29-

Importantly, the aircraft being flown was Aerostar N3645T, the same aircraft that had struck the treetops on April 4, 1984. Also, in an interview later with a PDQ training pilot, the report was obtained that the peculiar condition with glide slope indications had been observed before on training flights. All indica~ions from talking with PDQ maintenance personnel are that the particular KI 525A HSI had not been changed over the period of the year, so all anomalous glide slope indications from PDQ can be traced to that specific indicator.

PILOT REPORTS OF INTEREST.

On Sunday evening, May 13, 1984, three pilot reports were obtained by the tower that the glide slope was not being received~ Status monitors in the tower showed normal operation. The maintenance man was called; he checked his status indicators and went to the glide slope site, finding everything normal. Chart records obtained by Ohio University during that time period showed conclusively that there was no loss of signal.

GROUND-BASED DATA COLLECTION SETUP.

Because of the interest in the performance of the Pontiac glide slope even down to periods of time as short at one and two seconds, continuous recording of the glide slope signal was begun on April 20, 1984, in three locations. These were 1) the near-field monitor indication recorded at the glide slope transmitter building; 2) the below-path clearance at the middle marker, 25 feet above the ground, and 3) the below-path clearance at the middle marker 35 feet above the ground. Each of these signals is directly related to the signal the pilot is receiving.

The signal samples in the case of the far-field are obtained through the use of airborne-type receivers. These r~ceivers are Narco model UGR-2 which have had a modification in the audio signal processing area. The modification has been accomplished to allow an extra large range of DDM values to allow recording in the area well below the on-course of the glide slope. Large range (by a factor of 5 beyond the pilot's scale indication) with good linearity-has been provided.

Figure 2-24 shows the far-field monitoring site established by Ohio University near the Pontiac middle marker. A 40-foot tower supports two bent dipole glide slope antennas at the 25-foot and 35-foot levels. Signals from each of these antennas are fed to separate receiving and recording systems. These systems are shown in figure 2-25. Direct current-regulated power supplies delivering 13.6 volts to the airborne-type receivers are operated from the commercial power serving the middle marker and the general area near the airport.

The recorders used are the Heath Schlumbarger SR 206 which employ colored felt-type marker pens to produce the analog strip chart record. The chart speed used for the three recorders was 6 inches per hour. The chart width is approximately 12 inches with 20 microamperes or 80 percent of·alarm per inch.

-30-

The two monitoring systems at the middle marker were housed in the 2 1/2 ton truck shown in figure 2-26. This truck provided an excellent shelter for the equipment, especially after special insulation was added and the compartment reduced in size by internal paritioning. It proved to be espe­cially valuable in protecting the equipment when vandals demolished all of the breakable items on the truck, but could not enter the bed area.

The recorder in the near-field was simply connected across the path monitor meter terminals with the pen responding ± one inch to 140 percent of tolerance.

Approximately 145 rolls of strip chart da~a were taken over the 12-month period. Some observers had felt that there were more complaints concerning the glide path during certain seasonal periods, henca, the desire to have one year of continuous data. Recording terminated on April 10, 1986, at 1830Z. All recordings were showing nominal values.

Data were remarkably complete in that only less than 1% or 242 hours was lost due to recorder pen malfunctions. The total recording time was. 8520 hours per location or 25,560 hours total for the 355 day period. Fortunately, pen failures were at the far-field station and at no time did both pens fail simultaneously. In otQer words, there was at least one recording of far-field information being made at all times.

FLIGHT DATA.

Special flights were made on the glide slope at every opportunity. Some flights were tracked using an optical theodolite, Warren-Knight Model 83. The theodolite was positioned consistent with the u.s. Flight Inspection Manual, 0 AP 8200.1, Section 217.25. No abnormal conditions were found on any of the flights.

TRACKING OF AIRCRAFT.

On several occasions, when visits were made to the Pontiac Airport to collect chart recordings, fly the glide slope, and interview personnel, a theodolite was taken to track aircraft as targets of opportunity to deter­mine if any deviations were evident in their approach tracks. The pilots were not told that they were being tracked. The hypothesis was that there might be some peculiar air currents that would cause the aircraft to deviate regularly from the glide slope. No evidence was obtained to sup­port the hypothesis.

SNOW EFFECTS.

Because the glide slope at Pontiac is an image type, it can be affected by snow. When reports were received that a heavy cover of snow existed at Pontiac, an Ohio University team went to make flight measurements to deter­mine path angle. This was accomplished, with a path angle of 2.99 degrees measured. A close investigation of the ground plane snow cover revealed that an average of only 6 inches existed over the ground plane; hence, the path angle reading is consistent with that expected from a long study of ground plane snow effects [17, 18, 19].

-31-

ANALYSIS OF GROUND DATA.

Every minute of operation of the analog strip chart data from each of the three recording sites was scrutinized for abnormalities. None was found of significance. Appendices A, B, and C contain tables of the deviations that were noted.

Appendix D shows the record from the upper antenna of the far-field monitor and the path monitor from the station for the period to include the time of the aircraft accident on December 12, 1984, Greenwich time. It is obvious from the record that no change in glide slope operation had taken place.

RECOMMENDATIONS.

The data and information obtained during this investigation are quite conclusive. The only items which bear further attention are as follows:

First, the various comments that have been received suggest that if there is any real, physical problem, it would be in the radio frequency inter-

. ference (RFI) area. The possibility that a spurious signal is being received which contaminates the glide slope signal exists, but the probabi­lity is believed very small. No rigorous attempt was made to search out RFI. No reports were obtained from pilots. If there were any RFI problems, they would have to be infrequent, _19_~<!li~~g_,__i!Jld specific in character so that it would affect a glide slope receiver. The fact that Ohio University operated two fixed, airborne-type receivers continuously for one year without RFI problems is strong evidence that it does not exist. To produce final convincing evidence, a spectrum analyzer and spe­cial RFI-sensitive recordings should be implemented.

Second, methods of obtaining pilot report-type data should be refined and made more rigorous. Pilot reports, of course, involve the human factor, meaning that psychology plays a part. An inexperienced, nervous pilot will generally give a different report than will a seasoned veteran. Yet, when reports are documented, some of this type of information is missing. One is weighted equally with another. Also, the effect of desire-for con­sistency is frequently present. When a solicitation is made by ATC, for example, and a pilot reports a problem with a facility, it is not unusual for the next pilot who reports having heard the report of the first, to be somewhat consistent even when the first report is incorrect. A study should be made by the FAA to develop a method to minimize the subjectivity in pilot reports. ·

Finally, a considerable amount of glide slope signal exists in the back­course region. Because there is a published back-course localizer approach, and even though there is a notation to disregard the glide slope signal, it is undesirable to have it present because it is distracting and misleading. A glide path structure is evident that leads to the stop end of the runway. The glide slope antenna installation should be modified to minimize the radiation to the rear.

-32-

THEORETICAL INVESTIGATION OF NULL REFERENCE, SIDEBAND REFERENCE AND CAPTURE EFFECT GLIDE SLOPE SIGNAL SCATTERING FOR CRITICAL AREA DETERMINATION.

CONCLUSIONS.

The following conclusions are based on the theoretical work performed to examine the effects of scatterer aircraft on the performance of null reference, sideband reference, and capture effect glide slopes.

1. The null reference glide slope has significantly larger critical areas for Category I and Category II/III operations than either the sideband reference or capture effect system.

2. Critical area size is strongly dependent on the overall size of the scatterer aircraft. No specific formula is sufficient to define fully this relationship, but Aero Commander (R-500) and smaller size aircraft have very little effect on the various types glide slope systems.

3. When the fuselage of the scatterer aircraft is perpendicular to runway centerline, the largest contributor to path structure perturbations is the tail section of the aircraft.

4. The use of single dipole antennas as radiating elements causes the critical areas to be significantly larger than when FA-8976 antennas are used. The use of GRN-27 radiating antennas pro­vides only a very slight improvement over the directional array.

5. A limited study of the critical areas for the endfire glide slope indicates that the computations require approximately 100 hours of CPU time on an IBM 4381 for each scatterer orientation. This is so excessive as to be impractical at this time.

INTRODUCTION AND BACKGROUND.

The term critical area is defined for purposes of this study as that area in which the presence of a scatterer aircraft will cause aberrations in the radiated ILS glide slope signal which exceed tolerances specified in U.S. Flight Inspection Manual 8200.1.

These aberrations are normally evideat to a pilot flying an instrument approach as undesired movements in the glide slope CDI indicator. Short term irregularities are usually only an inconvenience, although they may cause some autopilots to disengage. Long term aberrations are very unde­sirable, and can in an extreme case, produce unsafe conditions.

An examination of nearly any airport layout will reveal areas where parked, holding, or taxiing aircraft will be in remarkably close proximity to the glide slope antenna array. It is intuitively obvious that any large con­ducting object placed near an antenna transmitting electromagnetic

-33-

radiation will have an effect on the transmitting system and its radiation patterns. It is therefore necessary to determine what restrictions or pro­hibitions should be implemented with respect to the area surrounding ILS transmitting arrays in order to insure that the signals serving user aircraft on approach remain adequate and safe.

Numerous experimental and theoretical investigations, both in the U.S. and abroad, have attempted to examine the effects of aircraft near both the localizer and glide slope transmitting arrays. In 1974, Ohio University published the first comprehensive work [20] on mathematical modeling of large aircraft near the ILS transmitting antennas. This work was validated with on-site measurements using the Boeing 747 and the Lockheed C-SA. The math model was also checked against data collected [21] at Heathrow Airport, London, England. Additional studies at Ohio University [22, 23, 24] serv~d to refine the technique of mathematical modeling. Other reports published by Ohio University have provided both experimental and theoreti­cal data to establish the minimum critical area required for the GRN-27 dual frequency localizer [25] and the 8~element and 14-element single­frequency localizers [26] operating under Category III tolerances. Both of these studies deal with the scattering effects of wide-body aircraft such as the B-747. More recently, Ohio University has published a report [27] which establishes critical area requirements for the 8-element and 14-element log-periodic, single-frequency localizers for aircraft sizes ranging from the Douglas DC-9 through Boeing-747. The most recent study of glide slope critical areas [28] deals with general·aviation aircraft impact on the null-reference glide slope.

While all of these previous studies have provided valuable information towards establishing minimum requirements for critical areas, there are many gaps in the subject matter which have never been addressed. The pur­pose of this study is to expand on previous work and provide a comprehen­sive determination of critical areas for the null-reference, sideband reference and capture-effect glide slope arrays relative to a wide range of scatterer sizes.

OBJECTIVE OF WORK.

The broad objective of the work presented in this report is to identify areas of the earth's surface near the glide slope transmitting array where the presence of an aircraft will cause unacceptable derogation of the glide slope radiated signal. The identification process consists of using a mathematical model, which has been validated through the use of flight

-measurements, to predict the values of perturbations in the glide slope CDI signal. These predictions will be used to define an appropriate critical area that will allow pilots and air traffic controllers to take appropriate holding actions to prevent out-of-tolerance path anomalies from occurring in space. The manifestation of this will result in hold lines painted on the taxiways accompanied by standardized signs indicating the bold posi­tions.

Within the broad purpose of this work there .are three specific objectives to be achieved. These are:

-34-

1. Obtaining the typical relationship by which the critical area can be estimated from the size of the interfering aircraft.

2. Determining the critical area size relationship to the type of glide slope transmitting array.

3. Determining the critical area size relationship to the type of radiating elements used in the transmitting array.

APPROACH TO SOLUTION.

The basic approach to the solution of this problem is to use the math model to produce calculations of glide slope CDI perturbations caused by aircraft in the vicinity of the glide slope transmitting array.

A previous study [29] has demonstrated reasonably good agreement between experimental and calculational results for the glide slope. Because of the expense and difficulty in collecting a large number of samples experimen­tally, a calculational approach has been found to be the most effective and efficient method of systematically study1ng glide slope critical areas. This approach is further justified by past field measurements which have established the validity of the mathematical model.

There are some difficulties even with this approach. For example, there are an infinite number of locations where a scatterer aircraft can be placed for the simulated approach. To make the problem tractable, the area used for scatterer location is broken into a grid, and a simulated glide slope approach is calculated with a scatterer present at each of the grid locations in turn.

An additional difficulty is choosing a sampling interval along the simu­lated approach that will provide sufficient data to analyze the approach, and yet keep the number of necessary calculations at a reasonable level. It should be noted that the current implementation of the Physical Optics Mathematical Model is designed to deal with a maximum of 500 points.

The sampling interval has been chosen to be 50 feet. This means that an imaginary glide slope receiver is placed at a position along the approach, all the calculations necessary to determine glide slope CDI at that point are performed, and the receiver position is advanced 50 feet. This process is started at a distance of 24,300 feet from the threshold, and is halted at the threshold. This gives 487 data points along the approach. Each data point is assumed to be valid for a 50 foot segment of the approach, or 25 feet each side of the computed point.

A f~nal difficulty lies in determining a suitable filter which can be applied to the calculated data in order to make it compatible with standard flight inspection techniques. The time constant of the deviation recording system for flight inspection work is specified by ICAO [30] as being 50/V seconds, where V is the aircraft speed in knots. The speed of the aircraft is assumed to be 200 ft/sec for all simulations in this study. This con­verts to 118 knots, and yields a recommended time constant of .42 second.

-35-

This is normally achieved in ILS receivers by means of a large capacitor across the meter loads. Each receiver is designed with the capability of driving a certain load impedance, and resistors are usually connected in parallel with the meter movements as needed to achieve the desired impe­dance. For example, a typical ILS receiver might be designed to drive a load of 333 ohms. Since 333 ohms represents the parallel combination of three 1000 ohm resistors, the receiver can drive up to three separate meter movements, with 1000 ohm resistors substituted for unused meters.

The formula for the time constant of this circuit is given by

where,

TC = R x C

TC is the time constant in microseconds, R is the resistance in ohms, C is the capacitance in microfarads.

Using a 1200 ~F capacitor with the 333 ohm load in the example above will give a time constant of .40 second. This is in very good agreement with the time constant specified by ICAO for flight inspection work.

While a low-pass filter of this sort is easily implemented using physical components such as capacitors and resistors, ~t is considerably more dif­ficult to derive a mathematical formula which wi~l apply these filter characteristics to calculated data. The basic formula for such a filter has been derived [31] and can be expressed as

FX(i)=(SP*(X(i)+X(i-1))-(FX(i-1)*(SP-2.0*TC)))/(SP+2.0*TC)

where,

i = receiver position number, X = input (unfiltered) data,

FX = Filtered value of X, SP = Sampling Period of data in seconds, TC = Time Constant desired for filter.

In this study, calculations for CDI are made in SO foot increments along the approach path. The aircraft velocity is 200 ft/sec, so the sampling period is .25 second. The desired time constant is .42 second. Substitution of the input data values for X will yield the filtered output data. Since this formula represents a specified standard, all calculations will be based on use of this filter.

Since an aircraft is nominally on runway centerline during an approach, the glide slope perturbations are examined only in terms of path structure. The glide slope CDI values are analyzed for the entire approach to deter­mine if they exceed any of the structure

1tolerances specified in u.s.

Flight Inspection Manual 8200.1,Section 217.5. Path structure analysis is a computerized process. Programs and EXEC's have been written which apply

-36-

standard flight check tolerances to each simulated approach for both CAT I and CAT II/III operation. The peak CDI perturbations are predicted with respect to specific ILS Zones, viz, 1, 2, and 3 [32]. Since different tolerances sometimes are applied for different zones, this distinction is considered desirable.

In order to determine the typical relationship by which the critical area can be estimated from the size of the interfering aircraft, 5 different sizes of aircraft are included. In order of overall increasing scatterer size, they are:

1. Rockwell Aero Commander R-500 2. Convair C-440 3. Douglas DC-9 4. Boeing B-727 5. Boeing B-747

The orientation of the scatterer aircraft involves 4 discrete cases:

1. perpendicular to runway centerline, tail away from the runway. 2. perpendicular to runway centerline, tail towards the runway. 3. parallel to runway centerline, tail away from the glide slope

mast. 4. parallel to runway centerline, tail towards the glide slope

mast.

MATHEMATICAL MODEL DESCRIPTION.

This effort has been completed through the use of the same model as . Longworth [33, 34] and McFarland [35], which is an updated version of the 1974 model used by Rondini [36] and earlier by Chin et.al. [37].

The model operates using the physical optics principles by considering the aircraft as a target or reflector. The target can be satisfactorily modeled by considering it as a collection of flat plates whose profile is that of the specific aircraft. The plates are assumed to be perfectly conducting and located with a specified orientation at a specific location in an area through which the glide slope signals are propagating.

The plates are broken into incrementally small areas, and currents flowing in the incremental plates become source currents for the scattered signals. An integration of the contributions produced by the incremental plates is accomplished, and summed with the direct radiation from the localizer.

An in-depth description [38] of the physical optics approach to modeling was published by Ohio University in 1983. A user's guide [39] to the math model is also available.

CALCULATION WORK.

The final product of this work is intended to be a set of maps delineating those locations where a certain size and orientation of interfering

-37-

aircraft will cause the glide slope to exceed certain tolerances specified in u.s. Flight Inspection Manual 8200.1, Section 217.5.

The scope of this particuliar work is larger than any previous study of critical areas. The data is analyzed with respect to two tolerance cate­gories for each of 5 sizes of scatterers. Each of the three specific objectives listed previously doubles the number of cases to be considered. Since most of these specific objectives require the determination of trends and relationships, it is intuitive that processing larger quantities of data will make these trends and relationships more obvious. Several steps have been taken to allow processing the largest bulk of data possible.

Computer programs and EXEC's have been written allowing the calculations to be performed on one computer while the results are spooled to a second com­puter for examination and processing into files which can be plotted. This arrangement requires no operator intervention, and allows the first com­puter to generate data almost continously. It also drastically reduces the data storage requirements, since only the results are spooled to the second computer.

In order to obtain the objectives of this study, approximately 140,000 simulated approaches have been calculated and analyzed. This consumed a total of about 1000 hours of CPU time on an IBM Model 4381 computer. Approximately 1000 plots were produced, with 138 included in this report.

While the general approach to the solution has been outlined, there are many variables yet to be defined. The math model requires a data file which is used as an input to the model for purposes of defining these variables. Since many of these are somewhat arbitrary, they are presented here in hopes that future work can be ~tandardized. Tables 2-1 and 2-2 summarize the values assigned.

The final definition of variables required to specify fully all required inputs to the math model is the actual radiating pattern of the individual array elements. The 3 types of radiating elements used in the simulation are:

1. Half-wave dipole; 2. Directional Array (FA-8976); and 3. 2-lambda Array (GRN-27).

The equations used to represent the radiation patterns at these element types may be stated as follows:

Half-wave dipole: F=SQRT(l-R(2)**2) ••••••••••••••••••••••••••••••••••••••• Equation 1

FA-8976: F=ABS(SINC(X)) •••••••••••••••••••••••••••••••••••••••••• Equation 2 X=4.0*PI*ATAN2(R(2),SQRT(R(1)**2+R(3)**2))

GRN-27: F=ABS(SINC(X))••••••••••••••••••••••••••••••••••••••••••Equation 3 X=2.*PI*ATAN2(R(2),SQRT(R(1)**2+R(3)**2))

-38-

GLIDE SLOPE ANTENNA ELEVATION DISTANCE FROM TYPE OF ARRAY POSITION (FT) CENTERLINE(FT)

NULL REFERENCE LOWER 14.13 400.37 UPPER 28.26 399.62

SIDEBAND REFERENCE LOWER 7.10 400.25 UPPER 21.10 399.75

CAPTURE EFFECT LOWER 14.10 400.75 MIDDLE 28.30 400.00 UPPER 42.40 398.75

Table 2-1. Antenna position variables used as input data for calculations.

-39-

Glide Slope Frequency••••••••••••••••••••••••••••••••••• 332.0 MHz

Distance from Mast to Threshold •••••••••••••••••••••••••• 1050 Ft

Distance from Threshold to ILS Point c . .................. 950 Ft

Distance from Threshold to ILS Point B. • • • • • • • • • • • • • • • • • • 3500 Ft

Distance from Threshold to ILS Point A. • • • • • • • • • • • • • • • • • • 24300 Ft

Starting Distance of Simulation •••••••••••••••••••••••••• 24300 Ft

Sampling Rate of Simulation •••••••••••••••••••••••••••••••• so Ft

Elevation Angle of Approach •••••••••••••••••••••••••••••• 3.00 Deg

Azimuth Angle of Approach•••••••••••••••••••••••••••••••• o.oo Deg

Speed of Approach•••••••••••••••••••••••••••••••••••••• 200 Ft/Sec

Table 2-2. Basic assumptions and variables used as input data for calculations.

-40-

where

F. Relative Field Intensity Level PI = Arcos(-1.0) R(l) = Normalized value of X in coordinate system R(2) = Normalized value of y in coordinate system R(3) = Normalized value of z in coordinate system

Figures 2-27, 2-28 and 2-29 demonstrate plots of these radiation patterns. It should be noted that there are some combinations of array type and radiating elements which are seldom, if ever, used. All possible com­binations are included in this report for the sake of completeness.

Five siz~ groupings of aircraft are used as scatterers. This provides a large range of sizes, and makes interpreting the data easier. The dimen­sions of the scattering plates used to simulate 5 types of aircraft are shown in figure 2-30.

Figure 2-31 represents the locations and dimensions of the area in which the aircraft were placed. The left boundary of the plot represents the runway centerline, the glide slope mast is at the bottom center. The first position for which a simulated approach is calculated is on runway cen­terline, 50 feet forward of the glide slope mast. As each simulation is completed, the filter formula is applied. The CDI values for the entire approach as well as the peak values for each ILS zone are then stored in computer memory. The program then increments the location of the scat­terers to the next position, and a new simulation begins. The computer program continues this operation until approaches have been calculated for all 408 points on the grid.

All 408 approaches are then analyzed to determine if they exceed any of the structure requirements specified for each of the ILS categories. All tolerance limits applied the path structure are obtained from the u.s. Flight Inspection Manual OA P 8200.1. Table 2-3 summarizes the structure tolerances for the two categories of ILS operation.

A plot, similar to figure 2-31, is produced for both ILS categories with a dot at every location where the presence of a particular size and orien­tation of scatterer causes the path structure to exceed tolerance limits. These maps serve to define the critical area for each case studied. The reference point on the aircraft for these maps is the center of the base of the fuselage as derived from figure 2-30.

In addition to the critical area maps, a contour map is prepared for each ILS zone. These maps represent the peak CDI that occurred within that par­ticular zone for each of the 408 approaches. Contour maps also are based on filtered data as described earlier. Since each of the 408 approaches is represented by a specific point on the grid of figure 2-31, the peak CDI value for each point forms a scalar field which allows a contour map to be drawn. From an examination of these contour maps, it is possible to deter­mine what an aircraft placed in a certain location with a certain orien­tation will produce in terms of maximum path perturbation, as well as the

-41-

CAT I Zone 2: ± 30 JJA from actual path angle STRUCTURE Zone 3: ± 30 JJA from graphical average path

CAT I Not Applicable ALIGNMENT

CAT II/III Zone 2: ± 30 JJA at ILS Point A, then a linear decrease STRUCTURE to ± 20 JJA at ILS Point B

Zone 3: ± 20 JJA from graphical average path

CAT II/III ± 37.5 JJA about the commissioned angle at ILS Point B; ALIGNMENT expanding linearly to ± 48.75 JJA about the

commissioned angle at ILS Point C; expanding linearly to ± 75 JJA about the commissioned angle at threshold.

Table 2-3. Summary of tolerances applied to all calculated CDI values.

-42-

ILS zone in which the perturbation occurs. By knowing the tolerance limit for a particular ILS zone, it is possible to determine the general outlines for critical area directly from the contour maps. The task is complicated by the fact that the tolerance limits in some zones have tapered brackets and that U.S Flight Inspection Manual OA P 8200.1 makes exceptions for brief out-of-tolerance conditions. These conditions cannot readily be determined from the contour maps, but are automatically taken into con­sideration for the critical area maps. The contour maps also are based on a reference point in the center of the base of the fuselage for each scat­terer.

Figure 2-32 is a sample of the CDI information that is calculated for each of the 408 grid locations for a particular scatterer. The scatterer in this instance is a C-440 with the center of the fuselage located 500 feet forward of the glide slope mast, and 450 feet from runway centerline. The C-440 fuselage is perpendicular to the runway centerline, with the tail away from the runway. The CDI trace on this plot demonstrates what the pilot of an aircraft on approach would see under these conditions. The perturbation reaches a peak value of 29 ~A at a distance from the array of about 13,500 feet. Figure 2-33 is a sample contour map based on the same conditions as figure 2-32. It provides the peak CDI perturbation in ILS Zone 2 for 408 positions of a perpendicular C-440. By locating the point on the contour map that corresponds to the position (500, 450) of the C-440 used to calculate figure 2-32, one can see that the closest contour line is 30 ~A.

The following critical area maps and contour plots are arranged in order of overall increasing scatterer size, which is as follows:

1. Aero Commander (R-500) 2. Convair 440 (C-440) 3. DC-9 4. B-727 5. B-747

Two orientations are presented for each aircraft size: 1) fuselage perpen­dicular to the runway centerline with the tail away from .the runway and 2) fuselage parallel to the runway centerline with the tail towards the glide slope mast. The actual orientation is indicated on each plot. In each case, the critical area maps are presented first. In some cases, critical area maps may be missing for certain ILS categories. This indicates that there are no locations causing out-of-tolerance conditions for that cate­gory. A list of locations for which there are no out-of-tolerance points is shown in table 2-4. There are also some instances of contour maps not being present for certain ILS zones. This means that the largest pertur­bation in that zone is less than 10 ~A, which is the minimum value of the contours. A list of these contour maps is given in table 2-5.

The contour lines on each contour map have been chosen such that the entire range of structure tolerance limits are represented. Since the minimum value involved in analyzing the structure of a glide slope approach is 20 ~A, the minimum contour shown is 10 ~A. The maximum contour is 60 ~A, with

-43-

AIRCRAFT ARRAY* ORIENTATION ILS CATEGORIES

R-500 NR perpendicular, tail away from runway I

R-500 NR parallel, tail towards the mast I ,II/III

DC-9 NR perpendicular, tail away from runway I

R-500 SBR perpendicular, tail away from runway !,II/III

R-500 SBR parallel, tail towards the mast !,II/III

DC-9 SBR perpendicular, tail away from runway I

B-727 SBR perpendicular, tail away from runway I

R-500 CE perpendicular, tail away from runway I ,II/III

R-500 CE parallel, tail towards the mast I ,II/III

DC-9 CE perpendicular, tail away from runway I

B-727 CE perpendicular, tail away from runway I

B-747 CE perpendicular, tail away from runway I

* NR = null reference SBR = sideband reference CE capture ef~ect

Table 2-4. ILS categories for which no critical area maps are presented. No out-of-tolerance locations identified for X ) SO feet and Y ) 0 feet.

-44-

AIRCRAFT ARRAY* ORIENTATION REGIONs**

R-500 NR parallel, tail towards the mast X, y

R-500 SBR parallel, tail towards the mast X

R-500 CE parallel, tail towards the mast X, y

* NR null reference = SBR = sideband reference CE = capture effect --- -- -~-

** X ILS Zone 2 = y = ILS Point B to ILS Point c

Table 2-5. Regions for which no contour maps are presented. Maximum contours are less than 10 ~A.

-45-

contour increments of 10 ~A. More contour lines than this tend to clutter the plots. The computer programs which calculate the contour lines have the capability to automatically scale the contour intervals to fit the data, but then each contour map is based on different values. This makes it very difficult to compare maps, so the maps in this report are all based on an increment of 10 ~A between contours.

Grid positions for the scatterer aircraft as indicated in figure 2-31 are measured with respect to the center of the base of the rectangle repre­senting the main fuselage of each aircraft. If it is desired to determine critical area sizes or contour lines with respect to another reference point on the aircraft, it is necessary to refer to figure 2-30 for aircraft dimensions. Once the distance from the center of the main fuselage section to the new reference point is established, the contour maps and critical area maps can easily be interpreted from this new reference point.

As stated earlier, several specific orientations of scatterer aircraft are included in the calculations. The results, which are summarized in tables 2-6 through 2-14, reflect the distance (to the next 25 foot interval) from the runway centerline to the nearest point on the aircraft longitudinal axis. Therefore, the aircraft fuselage lies completely outside the criti­cal area boundaries shown in these tables. It should be emphasized that many of the aircraft orientations which show no identified out-of-tolerance locations in these tables may still cause_ very large CDI deviations. These deviations may be of short duration, or occur in an area of the approach where there are no specifications for path structure.

These tables may be used to examine the relationship between critical area size and size of the reflector aircraft. A comparison of similar orien­tations for different size aircraft shows, in general, a clear trend of increasing critical area size as aircraft size increases. Determining the exact nature of the relationship. of critical area to scatterer size, however, is complicated by several facts. First, the overall size of the scatterer is not the only factor. In the case of a perpendicular scat­terer, the tail section is the primary scatterer. The relative proportions of tail and fuselage vary even among aircraft of very similar size.

As a further complication, the data represents only perturbations along runway centerline at a normal elevation angle of approach. This is a reasonable limitation when examining the effects the scatterers will have on an aircraft making an approach, but is not optimum when determining relationships to scatterer size. Some scatterers may have their maxim~m impact along something other than a nominal approach.

Finally, while the resolution of the results demonstrated in tables 2-6 through 2-14 is 25 feet for x and 25 feet for y, the grid increments as shown in figure 2-31 are 50 feet for x and 50 feet for y. This masks small differences in critical area size. In spite of these limitations, the trend of increasing critical area size with increasing scatterer size is readily apparent.

In order to reduce the number of figures presented to a reasonable number, plots are presented only for the two most typical orientations. These are:

-46-

CATEGORY I CATEGORY II/III AIRCRAFT AND ORIENTATION X y* X y~

R-500, perpendicular, tail away from runway 100 400 400 400 R-500, perpendicular, tail towards runway 500 400 R-500, parallel, tail away from mast R-500, parallel, tail towards mast

C-440, perpendicular, tail away from runway 1675 425 1875 425 C-440, perpendicular, tail towards runway 1550 425 1800 425 C-440, parallel, tail away from mast 525 450 625 550 C-440, parallel, tail towards mast 425 450 675 550

oc...:9, perpendicular, tail away from runway 1650 400 2500 400 DC-9, perpendicular, tail towards runway 1400 400 2200 400 DC-9, parallel, tail away from mast 550 450 650 600 DC-9, parallel, tail towards mast 500 450 700 550

B-727, perpendicular, tail away from runway 2400 375 2950 375 B-727, perpendicular, tail towards runway 2100 375 2700 425 B-727 , parallel, tail away from mast 725 500 875 700 B-727, parallel, tail towards mast 675 500 825 750

B-747, perpendicular, tail away from runway 3000 450 3500 450 B-747, perpendicular, ti:lil towards runway 3100 450 3600 450 B-747, parallel, tail away from mast 850 550 1000 750 B-747, parallel, tail towards mast 850 550 1000 750

* x = distance in front of glide slope mast to nearest point on aircraft longitudinal axis

y = distance perpendicular to runway centerline to nearest point on aircraft longitudinal axis

no identified out-of-tolerance locations for x ) SO' and y ) O'

Table 2-6. Critical area vs. aircraft orientatio~ for null reference glide slope, dipole antennas.

-47-

~··

CATEGORY I CATEGORY II/III AIRCRAFT AND ORIENTATION X y* X y*

R-SOO, perpendicular, tail away from runway R-SOO, perpendicular, tail towards runway R-SOO, parallel, tail away from mast R-SOO, parallel, tail towards mast

C-440, perpendicular, tail away from runway 200 42S 2SO 42S C-440, perpendicular, tail towards runway C-440, parallel, tail away from mast 12S 400 32S 4SO C-440, parallel, tail towards mast 17 s 400 32S 4SO

DC-9, perpendicular, tail away from runway sso 3SO DC-9, perpendicular, tail towards runway 2SO 400 DC-9, parallel, tail away from mast 2SO 4SO 3SO 4SO DC-9, parallel, tail towards mast 2SO 400 400 4SO

B-727, perpendicular, tail away from runway 300 22S B-727, perpendicular, tail towards runway so 37S B-727, parallel, tail away from mast 37S 4SO 47S soo B-727, parallel, tail towards mast 37S 4SO 47S soo

B-747, perpendicular, ·tail away from runway 4SO 1SO 600 200 B-747, perpendicular, tail towards runway 700 4SO B-747, parallel, tail away from mast 700 300 800 4SO B-747, parallel, tail towards mast sso 4SO 6SO sso

* x = distance in front of glide slope mast to nearest point on aircraft longitudinal axis

y = distance perpendicular to runway centerline to nearest point on aircraft longitudinal axis

= no identified out-of-tolerance locations for x ) SO' and y > O'

Table 2-7. Critical area vs. aircraft orientation for sideband reference glide slope, dipole antennas.

-48-

CATEGORY I CATEGORY II/III AIRCRAFT AND ORJENTATION X y* x y*

R-500, perpendicular, tail away from runway R-500, perpendicular, tail towards runway R-500, parallel, tail away from mast R-500, parallel, tail towards mast

C-440, per-pendicular, tail away from runway 350 425 750 425 C-440, perpendicular, tail towards runway 800 425 C-440, parallel, tail away from mast 225 450 475 500 C-440, parallel, tail towards mast 225 450 475 550

DC-9, perpendicular, tail away from runway 650 300 800 400 DC-9, perpendicular, tail towards runway 600 400 DC-9, parallel, tail away from mast 400 450 500 550 DC-9, parallel, tail towards mast 400 450 500 550

B-727, perpendicular, tail away from runway 850 375 B-727, perpendicular, tail towards runway 550 225 750 325 B-727, parallel, tail away from mast 525 450 625 650 B-727, parallel, tail towards mast 525 450 675 700

B-747, perpendicular, tail away from runway 250 150 1050 350 B-747, perpendicular, tail towards runway 1050 350 1250 400 B-747, parallel, tail away from mast 700 700 800 750 B-747, parallel, tail towards mast 650 700 750 700

* x = distance in front of glide slope mast to nearest point on aircraft longitudinal axis

y = distance perpendicular to runway centerline to nearest point on aircraft longitudinal axis

= no identified out-of-tolerance locations for x ;> 50' and y ) O'

Table 2-8. Critical area vs. aircraft orientation for capture effect glide slope, dipole antennas.

-49-

CATEGORY I CATEGORY II/III AIRCRAFT AND ORIENTATION X y* X y*

R-500, perpendicular, tail away from runway 150 400 R-500, perpendicular, tail towards runway R-500, parallel, tail away from mast R-500, parallel, tail towards mast

C-440, perpendicular, tail away from runway 1675 425 1875 425 C-440, perpendicular, tail towards runway 1600 425 1800 425 C-440, parallel, tail away from mast 425 450 675 500 C-440, parallel, tail towards mast 425 450 675 450

DC-9, perpendicular, tail away from runway 1650 350 1900 400 DC-9, perpendicular, tail towards runway 1400 400 2000 400 DC-9, parallel, tail away from mast 500 450 700 500 DC-9, parallel, tail towards mast 450 450 700 500

B-727, perpendicular, tail away from runway 2075 375 2700 375 B-727, perpendicular, tail towards runway 2100 375 2700 425 B-727, parallel, tail away from mast 675 500 875 500 B-727, parallel, tail towards mast 675 500 825 550

B-747 j perpendicular, tail away from runway 3050 400 3200 450 B-747, perpendicular, tail towards runway 3050 450 3200 450 B-747, parallel, tail away from mast 850 500 1100 550 B-747, parallel, tail towards mast 800 500 950 550

* x = distance in front of glide slope mast to nearest point on aircraft longitudinal axis

y = distance perpendicular to runway centerline to nearest point on aircraft longitudinal axis

= no identified out-of-tolerance locations for x ) SO' and y > O'

Table 2-9. Critical area vs. aircraft orientation for null reference glide slope, directional antennas (FA-8976).

-50-

CATEGORY I CATEGORY II/III AIRCRAFT AND ORIENTATION X y* X y*

R-500, perpendicular, tail away from runway R-500, perpendicular, tail towards runway R-500, parallel, tail away from mast R-500, parallel, tail towards mast

C-440, perpendicular, tail away from runway 100 425 250 425 C-440, perpendicular, tail towards runway 350 425 C-440, parallel, tail away from mast 125 400 325 450 C-440, parallel, tail towards mast 175 400 375 400

DC-9, perpendicular, tail away from runway 250 350 DC-9, perpendicular, tail towards runway 300 400 DC-9, parallel, tail away from mast 200 400 400 450 DC-9, parallel, tail towards mast 250 400 400 450

B-727, perpendicular, tail away from runway 300 225 B-727, perpendicular, tail towards runway 300 325 B-727, parallel, tail away from mast 375 450 475 450 B-727, parallel, tail towards mast 425 450 475 450

B-747, perpendicular, tail away from runway 400 150 600 150 B-747, perpendicular, tail towards runway 700 400 B-747, parallel, tail away from mast 600 450 700 450 B-747, parallel, tail towards mast 550 450 650 450

* x = distance in front of glide slope mast to nearest point on aircraft longitudinal axis

y = distance perpendicular to runway centerline to nearest point on aircraft longitudinal axis

= no identified out-of-tolerance locations for x ) SO' and y > O'

Table 2-10. Critical area vs. aircraft orientation for sideband reference glide slope, directional antennas (FA-8976).

-51-

CATEGORY I CATEGORY II/III AIRCRAFT AND ORIENTATION X y* X y*

R-500, perpendicular, tail away from runway R-500, perpendicular, tail towards runway R-500, parallel, tail away from mast R-500, parallel, tail towards mast

C-440, perpendicular, tail away from runway 350 425 750 425 C-440, perpendicular, tail towards runway 800 425 C-440, parallel, tail away from mast 375 400 475 450 C-440, parallel, tail towards mast 375 400 475 450

DC-9, perpendicular, tail away from runway 750 350 DC-9, perpendicular, tail towards runway 550 400 DC-9, parallel, tail away from mast 400 450 500 500 DC-9, parallel, tail towards mast 450 450 550 500

B-727, perpendicular, tail away from runway 700 375 B-727, perpendicular, tail towards runway . 550 225 750 375 B-727, parallel, tail away from mast 525 450 675 550 B-727, parallel, tail towards mast 525 450 675 500

B-747, perpendicular, tail away from runway 1050 350 B-747, perpendicular, tail towards runway 1100 350 1250 400 B-747, parallel, tail away from mast 700 550 850 700 B-747, parallel, tail towards mast 700 500 800 700

* x = distance in front of glide slope mast to nearest point on aircraft longitudinal axis

y = distance perpendicular to runway centerline to nearest point on aircraft longitudinal axis

= no iden.tified out-of-tolerance locations for x :> SO' and y > O'

Table 2-11. Critical area vs. aircraft orientation for capture effect glide slope, directional antennas (FA-8976).

-52-

CATEGORY I CATEGORY II/III AIRCRAFT AND ORIENTATION X y* X y*

R-500, perpendicular, tail away from runway 150 400 R-500, perpendicular, tail towards runway 500 400 R-500, parallel, tail away from mast R-500, parallel, tail towards mast

C-440, perpendicular, tail away from runway 1700 425 1900 425 C-440, perpendicular, tail towards runway 1550 425 1800 425 C-440, parallel, tail away from mast 475 450 625 450 C-440, parallel, tail towards mast 525 450 675 450

DC-9, perpendicular, tail away from runway 1700 350 2550 400 DC-9, perpendicular, tail towards runway 1550 400 2200 400 DC-9, parallel, tail away from mast 550 450 750 500 DC-9, parallel, tail towards mast 500 450 700 500

B-727, perpendicular, tail away from runway 2075 375 2950 375 B-727, perpendicular, tail towards runway 2100 375 2950 425 B-727, parallel, tail away from mast 775 500 925 500 B-727, parallel, tail towards mast 675 500 825 500

B-747, perpendicular, tail away from runway 3050 400 3550 450 B-747, perpendicular, tail towards runway 3050 450 3600+ 550 B-747, parallel, tail away from mast 850 500 1000 550 B-747, parallel, tail towards mast 800 500 950 550

* x = distance in front of glide slope mast to nearest point on aircraft longitudinal axis

y = distance perpendicular to runway centerline to nearest point on aircraft longitudinal axis

= no identified out-of-tolerance locations for x ) SO' and y ) 0'

Table 2-12. Critical area vs. aircraft orientation for null reference gl~de slope, 2-lamdba antennas ( GRN-27).

-53-

CATEGORY I CATEGORY II/III AIRCRAFT AND ORIENTATION X y* X y*

R-500, perpendicular, tail away from runway R-500, perpendicular, tail towards runway R-500, parallel, tail away from mast ----R-500, parallel, tail towards mast 50 400

C-440, perpendicular, tail away from runway 200 425 250 425 C-440, perpendicular, tail towards ruJ1way 350 425 C-440, parallel, tail away from mast 275 400 325 400 C-440, parallel, tail towards mast Not Available

DC-9, perpendicular, tail away from runway 250 350 DC-9, perpendicular, tail towards runway 250 400 DC-9, parallel, tail away from mast 250 400 400 450 DC-9, parallel, tail towards mast 250 400 450 450

B-727, perpendic~lar, tail away from runway 300 175 B-727, perpendicular, tail towards runway 550 375 B-727, parallel, tail away from mast 425 450 475 450 B-727, parallel, tail towards mast 425 400 475 450

B-747, perpendicular, tail away from runway 800 150 B-747, perpendicular, tail towards runway 450 400 900 400 B-747, parallel, tail away from mast 600 450 700 450 B-747, parallel, tail towards mast 550 450 650 450

* x = distance in front of glide slope mast to nearest point on aircraft longitudinal axis

y = distance perpendicular to runway centerline to nearest point on aircraft longitudinal axis

-- = no identified out-of-tolerance locations for x ~ SO' and y > O'

Table 2-13. Critical area vs. aircraft orientation for sideband reference glide slope, 2-lamdba antennas (GRN-27).

-54-

CATEGORY I CATEGORY II/III AIRCRAFT AND ORIENTATION X y* X .y*

R-500, perpendicular, tail away from runway R-500, perpendicular, tail towards runway R-500, parallel, tail away from mast R-500, parallel, tail towards mast

C-440, perpendicular, tail away from runway 750 425 C-440, perpendicular, tail towards runway 650 375 800 425 C-440, parallel, tail away from mast 225 400 475 450 C-440, parallel, tail towards mast 425 400 575 450

DC-9, perpendicular·, tail away from runway 750 400 DC-9, perpendicular, tail towards runway 400 400 800 400 DC-9, parallel, tail away from mast 400 450 500 450 DC-9, parallel, tail towards mast 450 450 550 450

B-727, perpendicular, tail away from runway 700 375 _ __Ii--722 r--perpend-i.cular, tail towards runway 600 225 700 375

B-727, parallel, tail away from mast 575 450 675 500 B-727, parallel, tail towards mast 525 450 675 450

B-747, perpendicular, tail away from runway 950 300 B-747, perpendicular, tail towards runway 1100 400 1250 400 B-747, parallel, tail away from mast 750 500 800 500 B-747, parallel, tail towards mast 700 500 800 500

* x = distance in front of glide slope mast to nearest point on aircraft longitudinal axis

y = distance perpendicular to runway centerline to nearest point on aircraft longituqinal axis

= no identified out-of-tolerance locations for x ~ SO' and y ) O'

Table 2-14. Critical area vs. aircraft orientation for capture effect glide slope, 2-lamdba anten­nas (GRN-27).

-55-

1. Perpendicular orientation with tail away from the runway. 2. Parallel orientation with tail towards the mast.

These two orientations are intended to represent aircraft entering the run­way, as well as aircraft taxiing parallel to the runway. Figures 2-34 through 2-162 illustrate the critical area boundaries and show the contour maps of the peak perturbation based on the center of the fuselage of the scatterer for the null reference, sideband reference and capture effect glide slopes. Figures 2-34 through 2-162 are based on directional radiating elements (FA-8976). Figures 2-163 through 2-167 provide a com­parison of critical area and contour maps when dipole radiating elements are used while figures 2-168 through 2-172 demonstrate the effects of 2-lambda radiating elements (GRN-27). The dipole and GRN-27 plots may be compared directly to the FA-8976 element results shown in figures 2-64 through 2-68.

RECOMMENDATIONS.

The following recommendations are based principally on the theoretical calculations just completed which has identified areas in the vicinity of various types of glide slope arrays where the presence of various sizes and orientations of scatterers will cause the glide path structure to exceed tolerances specified in u.s. Flight Inspection Manual 8200.1, Section 217.5.

1. The user should identify the inherent noise present in the glide path structure of interest. The results in Chis study are based on this noise being zero. Since this is not true for practical sites, the results are not directly applicable for glide slope structures having high noise levels already. The perturbations produced by the scatterer aircraft must be added to the base level of inherent noise.

2. Since the inherent noise can increase the size of the critical area, efforts should be made to reduce this noise by fundamental site improvement.

3. At sites where inherent noise is a problem, the contour maps should be used to identify the contours which will cause the path structure to exceed tolerance limits when the contour is added to the inherent noise.

4. Assuming inherent noise fs not a problem, the critical area maps should be used to identify the boundaries which should be pro­tected from the presence of scatterer aircraft during instrument approaches.

S. Critical area size increases so drastically with very large aircraft that consideration should be given to specifying critical area in more than one category. Smaller airports which are unlikely to serve jumbo jets can use other than worst case critical areas.

-56-

6. Critical area size for a given airport should also be based on the ILS category of operation. Critical area limits based on CAT II/ III tolerances should not be specified for airports having only CAT I operations.

7. Critical area size is very dependent on the type of glide slope array. Critical area determination should be based on the type of array at each site.

8. Since critical area size is greatly increased if dipole antennas are used as radiating elements, consideration should be given to replacing existing dipoles with FA-8976 antennas at problem sites. In most cases, use of GRN-27 radiating elements provides no advantage over FA-8976 elements.

-57-

DETERMINATION OF AIRCRAFT DELIVERY HEIGHTS AT THRESHOLD FOR COUPLED ILS APPROACHES TO RUNWAY lSR AT DALLAS-FORT WORTH REGIONAL AIRPORT.

SUMMARY AND CONCLUSIONS.

Measurements to validate Order 8240.47 [40] with respect to the present siting of the glide slope serving runway 18R at the Dallas-Fort Worth Regional Airport, Dallas, Texas, have been completed. This was accomplished by observing the wheel crossing heights (WCH) of aircraft flying coupled ILS approaches.to runway 18R. Measurements were divided into the following three categories:

1. Full Autoland

2. Uncoupled at the threshold

3. Uncoupled at the CAT II minimum (100')

The data show that aircraft making coupled approaches to runway 18R are being delivered at heights within the tolerances set forth in FAA Order 8260.34 [41]. The arithmetic mean of the measured antenna crossing heights is 53.7 feet for all 143 approaches decoupled at 100 feet of altitude or below.

The application of FAA Order 8240.47 to the flight recordings made in the previous work by Ohio University [42] produces reference datum heights (RDH) of 44 feet and 46 feet. These same recordings produce achieved reference datum heights (ARDH) of 56 feet and 58 feet. In addition, FAA flight checks show values of 47 feet and 44 feet for RDH, and 51 feet and 45 feet for ARDH. Thus, the values of RDH fall well below the tolerance limits of Order 8240.47, although three out of the four ARDH measurements are within the tolerance limits, i.e. within the window from SO to 60 feet.

The difference in the values of RDH and antenna crossing height can be understood by examining the flight recording in figure 2-173. In ILS Zone 2 there is a steady decrease in path angle from ILS Point A to ILS Point B. Thus, the best fit straight line for the data points taken in this region produces a low RDH when extended to the threshold (see figure 2-174). Another possible reason for the difference in antenna crossing height and RDH is that the autopilot is flying the aircraft based on continously received glide path information, rather than calculating the best fit straight line from discrete samples of the ILS Zone 2 glide path or memorizing a histo~y.

The present siting of the glide slope at 18R does not produce a path in space for which application of Order 8240.47 yields an RDH within tolerance, although the path is within the tolerance limits of the u.s. Flight Inspection Manual, OAP-8200.1. The glide slope does produce an ARDH within the tolerances of the Order.

-58-

INTRODUCTION AND BACKGROUND.

This task is specifically concerned with validating FAA Order 8240.47, "Determination of Instrument Landing System (ILS) Glidepath Angle, Reference Datum Heights, and Ground Point of Intercept," with respect to the glide slope serving runway 18R at the Dallas-Fort Worth Regional Airport.

Measurements by the FAA and Ohio University have shown that the path meets the requirements of the U.S. Flight Inspection Manual OAP-8200.1, but does not provide a sufficiently high reference datum height (RDH) to meet the threshold crossing requirements of FAA Order 8260.34. Specifically, RDH values of 44 feet and 47 feet obtained by the FAA, and 44 feet and 46 feet measured by Ohio University fall significantly below the prescribed 55 ± 5 feet given by FAA Order 8260.34. In contrast, these same measurements pro­duced 3 out of 4 values of achieved reference datum heights (ARDH) within the 55 ± 5 feet tolerance. Specifically, FAA measurements for ARDH were 51 feet and 45 feet, while Ohio University data show values of 56 feet and 58 feet.

DATA COLLECTION.

Equipment. Data were gathered for this study using a theodolite, Warren-Knight model WK-83. Other equipment used include a stadia rod, a measuring tape, and three VHF radios. Two radios were used for com­munication between observers and one was used by the observer at the threshold to monitor .the ground control frequency.

Procedure. The measurements described in this report were taken between June 1, 1984, and June 15, 1984. The data collection principally involved visual determination of the wheel crossing heigbts of aircraft crossing the threshold of 18R prior to landing. This was accomplished with two obser­vers, one at the threshold and one at the theodolite. The theodolite posi­tion was surveyed as indicated in figure 2-175. This position near the runway was chosen to minimize slew in the azimuth plane, thus reducing the probability of tracking errors. The difference in elevation between the threshold and the theodolite position was measured as shown in figure 2-176. After the theodolite was leveled, the distance (h-1) and the eyepiece height were measured. From this information the difference in elevation was calculated to be 4.25 feet.

Prior to the testing, a meeting was held with officials from the FAA and Alr Transport Association of America (ATA) to brief them on the procedure. A request was made for the ATA to communicate to the pilots that the eva­luation would be taking place and to request their cooperation. Arrangements were made with Airport Operations for airdrome access. An ATIS message was prepared to alert the pilots when the measurements were being made. The message read as follows:

"Evaluation of glide slope to runway 18R in progress. Request coupled approach to threshold if practical. Advise ground control of point where uncoupled. Men and equipment 35 feet west of runway."

-59-

The ATIS message alone did not produce the desired results, perhaps because the ATIS is typically monitored by the flight engineer rather than the pilot. It was found that a much more effective procedure was to have the approach controller request the pilot to fly a coupled approach and report on ground control frequency at what altitude the approach was decoupled. It should be noted that if the approach controllers did not allow for flexibility in intercepting the ILS, the auto pilot couplers, in general, could not satisfactorily intercept and track the ILS path. In fact, the efficient data collection periods turned out to be during times when lowered visibility required IMC ILS approaches. Unfortunately for this work, these conditions only existed for two of the fourteen days of data collection.

The theodolite operator tracked the lowest part of the aircrafts' main landing gear on approach. The observer at the threshold would call "mark" over the communication frequency when the main gear crossed the threshold. On this cue the theodolite operator would stop tracking and record the tail numbe~ of the aircraft, the time, and the elevation angle as read from the theodolite. The observer at the threshold was responsible for recording the date, time, type of aircraft, carrier, and the point of decoupling.

The wheel crossing height (WCH) was calculated using the expression

WCH = d tan(eEL) + (h-1)

where the variables are defined per figure 2-174. The eyepiece height was measured before and after each testing session. The antenna crossing heights listed in the results were obtained by adding the glidepath-to­wheel heights to the WCH. The values used for glidepath-to-wheel height were taken from FAA Order 8260.34.

DATA PRESENTATION AND ANALYSIS.

The empirical data are given in appendix E. Note that the data have not been chosen selectively, but that rather all samples in which a report on decoupling point information was received are included.

A total of 740 aircraft were tracked, with decoupling point information received for 143 approaches, thus resulting in a sampling efficiency of 19 percent. This low efficiency is mostly ~ttributable to the large number of operations at DFW. In addition to these quantitative data, there were many pilots who reported qualitative information about the ILS. This information, which may be of interest, is included in appendix F.

-60-

A basic statistical analysis performed on the data yields the following results:

Antenna Crossing Heights Standard Number (feet) Deviation of

Mean Minimum Maximum (ft) Samples

Auto land 53.7 46.1 62.1 4.2 33

D/C at 50 ft. 53.4 38.7 60.9 5.6 38

D/C at 100 ft. 53.8 36.8 69.6 7.5 72

The mean TCH for all 143 samples is 53.7 feet.

SPECIFIC FINDINGS.

The data show that aircraft making coupled ILS approaches to Runway 18R are being delivered at an acceptable height with respect to FAA Order 8260.34. The mean antenna crossing height values are consi~te~~_amQ~g_~~~ three categories of data. The standard deviation from the mean antenna crossing height increases with an increase in the decoupling altitude. This is reasonable to expect since as the pilot decouples farther from the threshold, more time is available for pilot controlled deviations from the electronic path as the approach becomes more of a visual type. A sen­sitivity analysis of the method used to measure the WCH is discussed in appendix G.

RECOMMENDATIONS.

Based on the data presented in this report, previous research, and on extensive experience, the recommendations are as follows:

1. Continue the operation of the glide slope on 18R with no change in position.

2. Improve the reflecting ground plane in front of the glide slope antennas by more uniform grading of the surfaces.

3. Examine the capability of mathematical models to predict the RDH and ARDH from detailed topographic information.

4. Examine the relationship of RDH and ARDH to the topography of the ground plane, and determine the sensitive areas of reflecting ground in terms of producing the RDH and the ARDH values, i.e. perform a sensitivity analysis.

5. Determine some limits of terrain .variation that exist to produce acceptable RDH and ARDH, but at tolerance values set forth in Order 8240.47.

-61-

6. Correlate the information presented in this report with calcula­tions for and measurements made at other airports.

7. Data on airborne equipment as it relates to antenna crossing heights, should be acquired and analyzed. The plan is to perform this task in the very near future.

-62-

SITING EVALUATION WITH RESPECT TO REFERENCE DATUM HEIGHTS FOR THE GLIDE SLOPE SERVING RUNWAY 18R AT THE DALLAS-FORT WORTH REGIONAL AIRPORT.

SUMMARY AND CONCLUSIONS.

The newly installed glide slope serving Runway 18R at the Dallas-Fort Worth Regional Airport has produced an out-of-tolerance condition with respect to the reference datum heights. The reference datum height (RDH) has been found below and outside tolerable levels, although the achieved reference datum height (ARDH) produced by the glide slope is within the prescribed tolerance limits of 55 ± 5 feet. The glidepath parameters have been found to be within the tolerance limits of the u.s. Flight Inspection Manual 0 AP 8200.1. Empirical data previously obtained by Ohio University have shown that user aircraft are being delivered at acceptable heights with respect to the t,olerances in FAA Order 8260.34.

The purpose of this siting evaluation is to determine if some modifications can be made to the site at Dallas that will en~ble it to satisfy all required tolerances of FAA Order 8240.47, specifically the RDH tolerance. The problem is mathematically modeled to determine the effects of the imperfect reflecting ground on the signal in space, and therefore on the quantities RDH and ARDH.

The relationships of RDH and ARDH to terrain imperfections are determined by developing a terrain profile which closely approximates· the actual reflecting terrain in front of the glide slope antennas and examining the sensitivity of the reference datum heights to variations in the original terrain profile. This enables one to determine which grading changes will allow the glide slope serving Runway 18R to meet the requirements of Order 8240.47.

At the DFW 18R site, there are three bowl-shaped depressions located within the first Fresnel zone. The results of math modeling of the site at Dallas indicate that all three of these depressions contribute to the below­tolerance RDH value. Specifically, the depression centered between the Air Freight Road and the North Emergency Road causes the steady decrease in path angle in ILS Zone 2. The RDH is most sensitive to this depression. The depression located between Taxiway W-18 and W-19 causes the fly-down signal near ILS Point B. The depression between Taxiway W-18 and the North Emergency Road contributes to the below tolerance RDH, but the RDH is not as sensitive to this depression as to the other two. Modeling results also indicate that only slight differences in the RDH are produced by grading the terrain between the glide slope antennas and Taxiway W-19 for minimum upslopes to the taxiway.

The conclusion is that decreasing the depths of the bowl-shaped depressions between the Air Freight Road and the glide slope antennas in the first Fresnel zone will produce an in-tolerance RDH. The RDH is most sensitive to the depth of the depression between the North Emergency Road and the Air Freight Road, and more sensitive to the depression between Taxiway W-18 and W-19 than to the depth of the depression between Taxiway W-18 and the North Emergency Road.

-63-

INTRODUCTION AND BACKGROUND.

The motivation for this work is the failure of the newly installed glide slope serving Runway 18R at the Dallas-Fort Worth Regional Airport to pass commissioning flight checks. An out-of-tolerance condition with respect to FAA Order 8240.47, "Determination of Instrument Landing System (ILS) Glidepath Angle, Reference Datum Heights, and Ground Point of Intercept."

Order 8240.47 prescribes that all new glide slope sites be commissioned based on the height at the runway threshold of an extension of a best-fit straight line analysis of the glidepath. The regression analysis to deter­mine the best-fit straight line is made on a set of evenly spaced data points over a selected sampling interval. In particular, the reference datum height (RDH) is computed in ILS Zone 2 and the achieved reference datum height (ARDH) is computed from 6000 feet to ILS Point C.

Measurements by the FAA and Ohio University [43] have shown that the path meets the requirements of the U.S. Flight Inspection Manual 0 AP 8200.1 [44], but does not produce a sufficiently high RDH to meet the threshold crossing requirements of FAA Order 8260.34 [45]. Specifically, RDH values of 44 and 47 feet, obtained by the FAA, and values of 44 and 46 feet, obtained by Ohio University, fall significantly below the required 55 ± 5 feet established by Order 8240.47. In contrast, these same flight measure­ments produced 3 of 4 values for ARDH within the 55 ± 5 feet tolerance. Specifically, FAA measurements produced ARDH values of 51 and 45 feet, while analysis of Ohio University flight recordings produced ARDH values of 55 and 56 feet.

A digitized flight recording of· the differential amplifier output for a low app~oach to Runway l~R is shown in figure 2-177. The steadily decreasing path angle in ILS Zone 2 causes the below tolerance RDH.

The first phase of this work consisted of validating Order 8240.47 with respect to the present siting of the glide slope serving DFW Runway 18R. This was accomplished by observing the actual heights of aircraft flying coupled ILS approaches to Runway 18R. Measurements were taken of the wheel crossing heights of the aircraft, from which the height of the glide slope antennas was calculated. This ailowed direct comparison with the RDH and ARDH values.

The empirical data indicate that aircraft flying coupled ILS approaches to Runway 18R are being delivered to the threshold at antenna heights averaging 53.7 feet [46]. This value is comfortably within the tolerances for threshold crossing heights set forth on Order 8260.34.

Thus, the present siting of the glide slope at DFW 18R does not produce a path in space for which the application of Order 8240.47 produces an RDH within the given tolerance, although the path is within the tolerance limits of the U.S. Flight Inspection Manual, OA P 8200.1. The glidepath does produce an ARDH that is within the tolerances set forth in Order 8240.47.

-64-

METHOD OF ANALYSIS.

The purpose of this siting evaluation is to determine if some modifications can be made to the site at Dallas that will enable it to satisfy all required tolerances of FAA Order 8240.47, specifically the RDH tolerance. The problem is analyzed by applying math models to determine the effects of the imperfect reflecting ground on the signal in space, and therefore on the quantities RDH and ARDH. This approach is taken since the glide slope system is a null-reference type: the path in space cannot be altered by electrical adjustments in the system, as is possible with the capture effect glide slope system (47, 48]. The relationships of RDH and ARDH to terrain imperfections are determined by developing a terrain profile which closely approximates the actual reflecting terrain in front of the glide slope antennas and examining the sensitivity of "the reference datum heights to variations in the original- terrain profile. In this manner the grading changes which will allow the glide slope serving Runway 18R to meet the requirements of Order 8240.47 are determined.

A topographic map of the glide slope critical area terrain, showing the multiple terrain irregularities, is given in figure 2-178. Of particular significance are the three bowl-shaped depressions located within the first Fresnel zone. This analysis consists basically of varying the depth of the three depressions parametrically and observing the sensitivity of RDH and ARDH.

The terrain profile which represents the actual reflecti.ng terrain is listed in figure 2-179. The coordinates listed in the terrain profile form a set of reflecting plates which approximate the actual ground. The math model used for this work was the Ohio University OUGS3D glide slope model, which was developed and verified in 1982. This model uses the Geometric Theory of Diffraction to calculate the complex electromagnetic fields in space, taking into account the effects of imperfections in the earth forming the ground plane. Another short program was written to calculate the reference datum heights from the differential amplifier information. The math model produces, in essence, a differential trace for a given terrain profile and system parameters.

At this site the terrain varies transverse to the runway, and therefore a three-dimensional terrain profile is necessary. The three-dimensional data base is obtained from the topographic map and input to the computer in matrix form. Data points input to this matrix are the three-dimensional coordinates of the plate corners, as shown in figure 2-179. These coor­dfnates are plotted isometrically in figure 2-180. The z-axis is chosen as the elevation coordinate and is referenced to the base of the antenna mast. The y-axis is positive outward along the centerline toward the approach path. The computer model selects the appropriate terrain data based upon the size of the first Fresnel zone for all receiver positions where the fields are to be computed.

The simulated differential trace for the three-dimensional terrain profile described above is plotted in figure 2-181. A plot of the model prediction of figure 2-181 versus the measured data shown in figure 2-177 is given in

-65-

figure 2-182. The maximum error of the modeled data with respect to the measured data in ILS Zone 2 is seen to be about 10 microamps.

SPECIFIC RESULTS AND FINDINGS.

The results of the parametric sensitivity analysis are displayed in table 2-15. Figures 2-183 through 2-187 illustrate the effects of varying the depth of the depression between Taxiways W-18 and W-19 (see figure 2-178), and clearly show this depression is the cause of the drop in path angle near ILS Point B. The sensitivities of RDH and ARDH to the depth of this depression are plotted in figures 2-188 and 2-189, respectively. Note that the elevations listed are the lowest elevation, or bottom of the depression, and are measured relative to the base of the antenna mast (597.85' MSL). The analysis shows that the first depression contributes to the out-of-tolerance condition.

The plots in figures 2-190 through 2-193 show the effects of raising the height of the second depression, between Taxiway W-18 and the North Emergency Road. The sensitivities of RDH and ARDH are plotted in figures 2-194 and 2-195, respectively, which show that the second depression contributes to the out-of-tolerance RDH quantity, but that the RDH value obtained is not as sensitive to this depression as to the one between the taxiways.

The results of raising the depth of the third depression, between the North Emergency Road, and the Air Freight Road are shown in figures 2-196 through 2-200. The reference datum height sensitivities, plotted in figures 2-201 and 2-202, show that the quantity RDH is more sensitive to the height of this terrain than to the heights of the other two depressions.

In addition to the results described above, a profile simulating the effects of grading the terrain between the glide slope antennas and Taxiway for a smooth slope up to the taxiway was tested. This terrain profile is plotted isometrically in figure 2-203. The simulated differential amplifier trace for this profile is shown in figure 2-204. The antenna heights were adjusted to produce a three degree path angle for this part of the analysis. For reference all profiles contained in this report which are based on the terrain modification described above are given an 'A' subscript. Examination of figure 2-204 shows that this modification does not produce any marked effects on the path in space, other than raising the path angle.

A parametric sensitivity analysis identical to the one performed on the original terrain profile produced the results which are presented in table 2-16. Comparison of table 2-15 with table 2-16 shows that the modification to the terrain between the glide slope antennas and Taxiway W-19 changes the RDH by less than two feet. The resulting CDI traces and sensitivity plots are shown in figure 2-205 through figure 2-224. The sensitivity plots show that the trends established by decreasing the depth o.f the terrain depressions are the same as those obtained for the original sen­sitivity analysis.

-66-

DEPRESSION PROFILE ELEVATION RDH (FT)

- 18R - 45.1

1 11 598.85 45.3 (Between- 12 599.85 48.0 Taxiways) 13 600.85 47.9

14 601.85 47.9 15 602.85 47.9

2 21 599.85 45.4 (Between 22 600.85 45.6 Taxiway W-18 23 601.85 46.0 & N. Emergency 24 602.85 46.6 Road)

3 31 598_.85 47.0 (Between N. 32 599.85 48.3 Emergency Road 33 600.85 49.2 & Air Freight 34 601.85 50.6 Road) 35 602.85 49.9

Table 2-15. Results of sensitivity analysis for three terrain depressions.

-67-

ARDH (FT)

55.5

55.7 56.4 56.9 57.5 57.7

54.8 54.4 54.1 53.7

55.1 55.2 55.1 55.2 55.6

DEPRESSION PROFILE ELEVATION RDH ARDH (FT) (FT)

- 18RA - 45.6 54.8

1 llA 598.85 46.2 54.8 (Between- 12A 599.85 48.0 55.0 Taxiways) 13A 600.85 48.2 56.4

14A 601.&5 48.3 57.1 15A 602.85 48.4 57.3

2 21A 599.85 46.1 54.3 (Between 22A 600.85 46.5 53.8 Taxiway W-18 23A 601.85 46.9 53.3 & N. Emergenc'1 24A 602.85 47.5 52.9 Road)

3 31A 598.85 47.7 54.7 (Between N. 32A 599.85 49.0 54.6 Emergency Road 33A 600.85 50.4 54.5 & Air Freight 34A 601.85 51.7 54.5 Road) 35A 602.85 50.8 55.0

Table 2-16. Results of sensitivity analysis for modification to Terrain Between Antennas and Taxiway W-19.

-6-8-

A corrected terrain profile was used to examine the cumulative effects of gradipg changes to the first and third depressions, and the effects of raising the terrain forming the second depression. The simulated results of decreasing the depth of the first depression to 2 feet above the base of the mast and the depth of the third depression to 4 feet above the base of the mast are shown in figure 2-225. Analysis of the sensitivity of RDH to the depth of the second depression (between Taxiway W-18 and the North Emergency Road) showed that the RDH at this site changes by only 2 feet when the elevation of the depression is raised by 3.15 feet.

The results for the corrected terrain profile indicate that an RDH within the tolerances of FAA Order 8240.47 can be produced by the grading changes described above.

A sensitivity analysis for the corrected terrain profile was also performed with the siting modification to the area between the antennas and Taxiway W-19 simulated. The results indicate that this change produces an RDH only 5 inches greater than before the modification.

CONCLUSIONS.

Conclusions based on the results of the analysis described in this document and on empirical data collected at the site are as follows:

1. The below-tolerance RDH value produced by the glide slope serving DFW Runway 18R is caused by the cumulative effects of three bowl­shaped depressions that are located within the first Fresnel zone.

2. The steady decrease in path angle in ILS Zone 2 is caused by the depression located approximately 1600 feet from the glide slope antennas and between the Air Freight Road and the North Emergency Road.

3. The fly-down in the path near ILS Point B is caused by the terrain depression centered approximately 900 feet in front of the glide slope antennas and between Taxiways W-18 and W-19.

4. The RDH is more sensitive to the terrain depression at 1600 feet from the antennas than to the other two bowl-shaped depressions.

5. The RDH is more sensitive to the terrain depression centered at 900 feet from the antennas than to the depression centered at approximately 1242 feet from the antennas. ·

6. Decreasing the depth of each of the bowl-shaped depressions results in an increased RDH value.

7. Changes in the depth of the depressions will not produce an out­of-tolerance ARDH.

-69-

RECOMMENDATIONS.

Based on the results presented in this paper, empirical data collected, and years of experience with siting glide slopes, the recommendations for the site at Runway 18R are:

1. If continued operation of the null-reference glide slope system at Runway 18R is desired, then the existing terrain must be graded to allow the RDH value produced to satisfy the required Category II tolerance of 55 ± 5 feet. The grading should be performed according to the following criteria:

a. The terrain between the North Emergency Road and the Air Freight Road should be graded for smooth slopes to an eleva­tion of at least 601.85 feet MSL.

b. The terrain between Taxiway W-18 and W-19 should be filled to an elevation of at least 599.85 feet MSL.

2. The dynamic nature of the fill dirt forming the reflecting ground for the glide slope should be stabilized by grading for minimum slopes and by planting a quality sod to prevent soil erosion.

3. Consideration should be given to installing a capture effect glide slope system that is known to be more tolerant of irregular reflecting ground.

4. The relationships of RDH and ARDH to ground plane imperfections should be examined in general by performing a sensitivity analysis to determine the most critical terrain regions.

5. The tolerances for RDH and ARDH should be re-examined since empirical data indicate that user aircraft are not delivered to the threshold at heights corresponding to the measured RDH at this site.

6. The data base established thus far with respect to application of FAA Order 8240.47 should be supplemented by correlating the infor­mation obtained with data from other sites having varying topo­graphic profiles to insure universal application.

-70-

ANALYSIS OF EFFECTS ON REFERENCE DATUM HEIGHTS OF A PROPOSED TAXIWAY ADDITION FOR RUNWAY 9R AT THE CHICAGO O'HARE INTERNATIONAL AIRPORT.

SUMMARY AND CONCLUSIONS.

A computer simulation to determine the effects on the reference datum heights of a proposed taxiway addition to Runway 9R at the Chicago O'Hare International Airport was completed. The following conclusions are pro­vided:

1. The proposed siting modification will permit the ground plane to ~erve adequately a null reference glide slope. Reference is made to FAA 8240.47 RDH/ARDH Category I tolerances.

2. The proposed taxiway is located in the glide slope critical area, and this area must be protected from taxiing jet aircraft, parti­cularly during Instrument Meteorological Conditions (IMC).

3. The proposed access road to the glide slope hut will cross in front of the transmitting antennas, and will invite problems by allowing vehicles to be present possibly sending the monitors

.into alarm.

INTRODUCTION.

The addition of a taxiway to Runway 9R at the Chicago O'Hare International Airport has been proposed. The proposed work includes a modification to the glide slope critical area terrain. A topographic map of the glide slope critical area terrain, including the proposed siting modification, is shown in figure 2-226. The proposed taxiway crosses directly between the approach path and the glide slope transmitting antennas.

The purpose of this work is to determine the effects on the reference datum height (RDH) and achieved reference datum height (ARDH) that will be caused by the proposed modification to the terrain. The RDH and ARDH are relati­vely new, measured quantities to which tolerances are applied [49].

ANALYSIS AND RESULTS.

The problem is analyzed using the OUGS glide slope model which includes the effects of non-ideal terrain. The OUGS model was developed and validated at Ohio University during the past 10 years. This model is based on the geometrical theory of diffraction (GTD), and includes the effects of diffraction, reflection, and blockage of the electromagnetic waves by irre­gular terrain.

Inspection of the topographic map in figure 2-226 reveals that the reflecting terrain is basically smooth and flat, except for the elevated region where the proposed taxiway is located. A good approximation to the actual terrain is the two-dimensional terrain profile shown in figure 2-227. This terrain profile is entered as part of the input data to the mathematical model. In addition, data pertaining to the facility

-71-

(frequency, antenna locations, antenna current distribution, etc.) are input to the model. An input file for a low approach is shown in figure 2-228.

The terrain slopes up longitudinally from the antenna mast. To produce a 3.0 degree path angle, the antenna heights must be adjusted to compensate for this upslope. The longitudinal gradient is approximately:

tan -l (4.5/1240) = 0.2°

The antenna heights ar~ selected to produce a path that is 2.8 degrees above an ideal ground [50]. The heights used in this analysis are 30.30 feet for the upper antenna and 15.15 feet for the lower antenna.

The simulated differential amplifier recording for the terrain profile shown in figure 2-227 is shown in figure 2-229. The elevated region of terrain where the proposed taxiway is located causes the path angle to decrease steadily throughout ILS Zone 2. This results in the aiming point, established by the regression analysis of FAA 8240.47, being closer to the threshold than if the site were ideal, and therefore the RDH is lower than the ideal value. At this site, however, the antennas are located far enough back from the threshold to compensate for the decrease in RDH from the ideal value.

The reference datum height tolerances for Category I operation are deter­mined from FAA Order 8260.34. This order prescribes maximum and minimum values of wheel crossing heights (WCH) at the runway threshold. For aircraft in height group 4 (B-747/767, 1-1011, DC-10, A-300), the minimum and maximum allowable RDH for Category I operation are 35 feet and 75 feet, respectively. The maximum recommended RDH is 60 feet. The predicted RDH of 57.7 feet satisfies required tolerances for Category I operation.

Figure 2-230 shows the simulated differential amplifier recording with Category I structure tolerances applied. Clearly, all structure tolerances are satisfied.

RECOMMENDATIONS.

Based on the data presented in this report and previous experience, the recommendations are:

1. Complete the proposed taxiway addition to Runway 9R to facilitate operational objectives. The proposed terrain contours will form a path in space that satisfies RDH tolerances for Category I ILS operation.

2. If the proposed taxiway addition is completed, the glide slope critical areas must be protected, since the taxiway crosses directly in front of the glide slope antennas.

3. The access road to the glide slope hut should be located behind the transmitting antennas, not in front as the topographic map of the proposed modification shows.

-72-

EFFECTS OF IRREGULAR PATH-FORMING TERRAIN ON GLIDE SLOPE REFERENCE DATUM HEIGHTS.

SUMMARY AND CONCLUSIONS.

The effects on the reference datum heights caused by various types of irre­gular terrain in the path-forming region have been determined using mathe­matical models. Also, the effects of parametric variations in the location, extent, and elevation of the irregular terrain region are iden­tified. Note that the terms "irregular terrain" and "terrain irregularity" are used synonymously with the terms "irregular terrain" and "rough terrain" used in FAA Order 6750.16B, "Siting Criteria for Instrument Landing Systems" [51]. Types of irregular terrain examined are simple ele­vated and depressed regions within the first 5000 feet from the antennas. The dimensions and location of these elemental terrain features are system­atically varied and the response of RDH and ARDH recorded. The results are furnished as graphs of RDH and ARDH versus distance to a region of irregu­lar terrain and size of terrain variation. An album of responses to repre­sentative profiles of irregular terrain results. This album may be used to evaluate a glide slope site with respect to FAA Order 8240.47. The siting engineer can use the graphs to predict the amount of terrain roughness that can be tolerated. Examination of results for various terrain irregulari­ties leads to the following conclusions:

1. RDH is more sensitive to irregular signal-forming terrain than ARDH.

2. RDH and ARDH vary significantly from the ideal when elevations are present in the terrain as compared to depressions. The larger the elevation the greater the change in RDH/ARDH.

3. The criteria established in this work for limits on terrain deviations are less stringent than existing siting criteria. If all existing criteria are met, so will be the tolerances for RDH/ ARDH.

4. RtiH is sensitive to terrain at greater distances from the antennas than ARDH.

S. RDH and ARDH are relatively insensitive to depressions in the reflecting terrain that are less than 100 feet in extent.

6. when the terrain has a uniform, longitudinal slope, the RDH/ARDH are simply changed by an amount produced by the slope. This amount is equal to the threshold elevation above that value should a flat plane have existed.

7. Both RDH and ARDH increase as the distance from runway cen­terline to the glide slope antenna mast increases. For allowable positions of the antenna mast, the RDH and ARDH change from nominal values by less than 3 feet.

-73-

8. RDH and ARDH increase as the distance between the antenna mast and threshold increases.

9. RDH and ARDH terrain sensitive areas extend farther from the antennas as the path angle is lowered.

10. The reference datum heights are not necessarily only a function of the terrain. Reflectors such as vertical surfaces can produce change.

11. The location of the flight measurement reference system in accordance with Order 8240.47, which allows the reference to be positioned at locations other than the antenna mast, can some­times allow in-tolerance values for RDH to be obtained when the location specified in OAP 8200.1, 217.5 produces out-of­tolerance values.

12. RDH·and ARDH are insensitive to small, uniform, lateral terrain slop~s.

INTRODUCTION AND BACKGROUND.

Order 8240.47 prescribes that glide slopes shall be commissioned based on the height at the runway threshold of an extension of a best-fit straight line through path angle"samples in certain regions of the approach. The reference datum height (RDH) is obtained from path angle samples in ILS Zone 2, while the achieved reference datum height (ARDH) is determined from samples of the glidepath from 6000 feet (from threshold) to ILS Point c.

Tolerances are applied to the reference datum heights based on the category of ILS operation at the site and the types of aircraft that are expected to use the facility. The tolerances for the different aircraft are determined by minimum and maximum values of wheel crossing heights. These limits are set forth in FAA Order 8260.34, "Glide Slope Threshold Crossing Requirements" [52].

Application has produced varied results. At Greenville, SC, a site with a history of marginal performance, application of the new standard provided flight check results indicating a better performance and obviated the need for costly relocation. At the Dallas-Fort Worth Regional Airport applica­tion prevented commissioning of a facility despite the fact that all tolerances in effect prior to establishment of the new order were met.

One objective of this work is to provide basic data to allow the siting engineer to identify limits on terrain irregularities for RDH and ARDH cri­teria to be met. The RDH and ARDH are relatively newly designated measured variables to which tolerance values must be applied. While Threshold Crossing Height (TCH) values are determined strictly from geometrical con­siderations, the RDH/ARDH values are derived from measured data produced by the radiating glide slope system. For desired RDH/ARDH values to be obtained knowledge of the effects of ground plane variations must exist. The results of this work are intended to provide some of this general knowledge.

-74-

Types of irregular terrain considered in this work are shown in figure 2-231. The results are compiled in plots of RDH/ARDH versus distance to and size of the irregular terrain region. An example is shown in figure 2-232.

ANALYTICAL METHOD.

The path formed in space by a properly adjusted image-type glide slope system is a function only of the topography provided no scattering objects (i.e. buildings, trees, aircraft) are present. For this reason the scope of this work is limited to determining the sensitivities of the reference datum heights to irregularities in the terrain serving as the ground plane at a site uncluttered by reflecting objects.

When this is the case, the reference datum heights are a function of the following parameters (see figure 2-233):

1. Coordinates of the glide slope transmitting antennas;

2. Coordinates of the flight measurement reference system;

3. Path angle determined by the sideband antenna height;

4. Reflection coefficient of the terrain serving as- the -ground--p-lane;-

5. Coordinates of the runway threshold;

6. Contours in the reflecting terrain.

In this work the location of the runway threshold and the terrain infor­mation are assumed to be given. The necessary terrain information is obtained from topographical maps of the glide slope site. To allow the effects of non-ideal terrain on the reference datum heights to be deter­mined, the following assumptions are made:

1. The antenna location is held constant at 1000 feet back from threshold and 400 feet off centerline;

2. The location of the flight measurement reference system is kept constant at the glide slope antenna mast and at the elevation of the runway abeam the mast.

3. The path angle is held constant at 3.0 degrees;

4. The simulated terrain is assumed to be perfectly conducting, and therefore has an ideal reflection coefficient for horizontally polarized, low-grazing angle type, electromagnetic waves.

When these assumptions are valid the reference datum heights are a function of the reflecting terrain only. For this reason the scope of this work is limited to determining the effects of representative terrain irregularities on the reference datum heights.

-75-

Each potential glide slope site in the field has a unique topography: therefore, an infinite number of possible ground plane imperfections exist. To bound this discussio~ categories of non-ideal terrain are identified. The categories are:

1. Irregular Terrain 2. Uniform Terrain Slopes

Irregular (rough) Terrain. This is the most detrimental and common siting deficiency encountered 1n siting a glide slope, and is given primary atten­tion in this work. This particular type of non-ideal terrain will be assumed to have ~he following characteristics (see figures 2-233 and 2-234):

1. The terrain is invariant in the direction perpendicular to the runway centerline (x-direction);

2. The extent, or width, in the direction of the runway centerline (y-direction) of the irregular terrain ranges from 10 feet to 500 feet. The terrain irregularities are large with respect to a wavelength.

3. The height, or depth, of the irregular terrain in elevation (z-direction) ranges from 2 feet to 50 feet.

4. The terrain serving as the ground plane is ideally flat in the horizontal plane, except for the extent of the irregular region, and extends a distance of 5000 feet from the antennas.

The effects of irregular terrain on RDH and ARDH are determined by using mathematical models to analyze parametrically the sensitivity of the reference datum heights to basic variations in the dimensions of a region of irregular terrain. This is accomplished by simulating a differential amplifier recording and from these data computing the reference datum heights for varying distances to a region of simulated irregular terrain. The result is a plot of the variation in the reference datum heights versus distance to the terrain discontinuity. These sensitivity plots are used as a guide in predicting whether a proposed site will serve adequately as the ground plane for an image-type glide slope system that will produce reference datum heights within required tolerances.

The mathematical model used in predicting the differential amplifier traces as a function of irregular terrain is the OUGS Geometrical Theory of Diffraction (GTD) model, which was developed and verified at Ohio University in 1982 [53]. The GTD model is used because it has been shown to be more accurate in predicting the effects of imperfect reflecting terrain than other glide slope models, such as those based on Physical Optics [54]. This glide slope model includes both diffracted and reflected ray contributions from the terrain to the electromagnetic fields in space.

Development of the GTD mo~el was motivated by the inability of models based on Physical Optics theory to predict accurately the diffraction and

-76-

blockage effects caused by upslopes. In contrast to the Physical Optics models, the GTD model includes shadowing and diffraction effects of the terrain between the transmitting antennas and the reflecting ground, and between the region of reflecting ground and the receiver.

The GTD approach to calculating the electromagnetic fields formed by an image-type antenna array over irregular terrain is divided into two parts: a geometrical process of determining which rays exist and where their reflection and/or diffraction point(s) lie, and a mathematical process of evaluating the magnitude and phase of the corresponding electric field at the receiver position by summing the contributions of all existing rays [55].

In this work the differential amplifier recording is obtained through simu­lation of an aircraft.flying a 3.0 degree approach with respect to the flight measurement reference system.

First, to bound the problem the shape and size of terrain irregularity are defined. The need is to choose the shape and size of terrain irregularity that are representative of the real world. The shapes of terrain irregu­larities initially considered for use in the sensitivity analyses are shown in figure 2-234. The names given to the types of irregular terrain are designated by the geometrical appearance of the cross-sectional shape, or profile; of the region of irregular terrain. Figure 2-234(a) shows a triangular-shaped terrain irregularity (more precisely an isosceles triangle). Figure 2-234(b) shows a rectangular-shaped irregularity. Figure 2-234(c) shows a trapezoidal-shaped terrain imperfection (more pre­cisely a symmetrical trapezoid with respect to the midpoint of its base). Simulated differential amplifier traces for these three types of irregular terrain, centered at 1500 feet, with a· width of 20 feet, and a height of 10 feet, are shown in figure 2-235. The results for the rectangular-shaped terrain irregularity correlate well with those for the trapezoidal-shaped irregularity. The triangular-shaped terrain irregularity, however, produ­ces a significantly different differential amplifier trace and corresponding RDH.

It is enlightening to compare sensitivity plots for the trapezoidal and triangular-shaped terrain irregularities. Since the differential amplifier traces for the rectangular and trapezoidal shapes are essentially iden­tical, only the trapezoidal- and triangular-shaped irregularities are con­sidered. The dimensional data pertaining to the terrain irregularity and the limits on distance to the region of rough terrain are input to a com­puter, which is used to produce a set of terrain profiles with the spe­cified irregularity in the ground plane located at varying distances from the antennas. An example of such a terrain profile is shown in figure 2-236. These terrain profiles are used as inputs to the GTD model. In addition to the terrain data, the input data to the math model include: . theodolite (reference) location, antenna location and current distribution, simulated flight pattern, and specification of the parameter of interest. A sample input file is shown in figure 2-237. The math model produces in essence a differential amplifier recording based on the input data. The reference datum heights are calculated based on the appropriate differen-

-77-

tial amplifier samples and required facility data. The result of the entire process is a plot of the RDH and ARDH versus distance to the terrain irregularity.

In summary, the math model is used to simulate a flight measurement for each location of the terrain irregularity; the reference datum heights are calculated from the simulated differential amplifier trace; and the end result is a plot of RDH and ARDH versus distance to the center of the terrain irregularity. A flow diagram of the process is shown in figure 2-238.

Sensitivity curves for the triangular- and trapezoidal-shaped terrain irre­gu~arities are plotted in figure 2-239 for a width of 10 feet and a height of 10 feet. The distance to the center of the irregularity is varied from 100 feet to 5000 feet in increments of 100 feet. The trapezoidal and triangular irregularities are seen to produce similar sensitivity curves. The magnitudes of the peak to peak variation are almost identical, while the sinusoidal variations in magnitude have different periods of oscilla­tion.

The sinusoidal nature of the reference datum height sensitivity curves is understood by examining figures 2-240 through 2-248, which are the simu­lated differential amplifier traces for each position of the trapezoidal­shaped terrain irregularity. The reference datum height sensitivity plots (figure 2-239) are calculated from these. As the distance to the irregu­larity increases, the differential amplifier traces are observed to 'tilt' around the three degree, or zero CDI, reference. This results in a change in the aiming point established by the regression analysis, and therefore in the RDH and ARDH. As the distance ~o the ground plane irregularity con­tinues to increase, the tilt of the differential amplifier trace shifts first in one direction and then reverses, causing_ the oscillations in the reference datum height sensitivity curves.

The sinusoidal variation in the sensitivity plots may lead one to believe that by choosing the distance from the antennas to the terrain irregularity appropriately, the RDH/ARDH tolerances may be accommodated. However, exa­mination of the sensitivity plots in figure 2-239 for the different shapes of terrain irregularities shows that the period of oscillation of the reference datum heights is dependent on the shape of the terrain irregu­larity, although the magnitudes of the sensitivity traces are nearly iden­tical. In practice the actual shape of the terrain profiles may not be known to great accuracy, and many different types of rough terrain will be encountered. In addition, the results given in this report are based on simulated terrain profiles consisting of perfectly flat, reflecting plates, while at most airports the terrain will deviate from these ideal con­ditions. Therefore, for a given region of irregular terrain _it is recom­mended that the peak values of the reference datum height sensitivity curves in the region of the rough terrain be considered as the worst case for a given height (depth), extent, and distance to the rough terrain, rather than obtaining a value from the sensitivity plot for a discrete location of the terrain roughness.

-78-

In general, application of these sensitivity curves to determine whether a proposed site is satisfactory or not will depend on variables that are held constant in this analysis. The location of the antennas is held constant at 1000 feet back from threshold and 400 feet off centerline, the path angle is held constant at three degrees, and the reference location is held constant. The following is intended to analyze the effects of changing these constants, and describe how the results produced by this work can be applied to such conditions.

The effects of antenna location and path angle on the reference datum heights at an ideal site are determined in appendix H. The results show that for the ideal site moving the antennas from 250 feet to 650 feet from centerline, which are the limits for lateral distance, causes the RDH to increase by 2 feet and the ARDH to increase by about 3 feet. This is understood by considering the curve formed by the intersection of the zero DDM conic and the localizer plane. As the conic moves away from the cen­terline, the curve becomes more hyperbolic and flares upward, and the resulting RDH and ARDH are moved accordingly upward. Since· the types of irregular terrain considered in this work are invariant in the direction perpendicular to the runway, the only response of the reference datum heights to the placement of the antennas at distances other than 400 feet off centerline will be the effect as described above.

When the antennas are placed at locations other than 1000 feet from threshold at a site with irregular terrain, the reference datum heights will be different than the values given in the sensitivity charts, because the distances from the reference (at the antenna mast) to the threshold and from the antennas to the irregular terrain are changed. Figure 2-249 shows sensitivity curves for antenna locations of 800, 1000, and 1200 feet back from threshold and 400 feet off centerline. Inspection of figure 2-249 shows that the curves are identical except for a constant difference of 10 feet between each. This is the same value as that obtained for an ideal site with the antennas located identically. The conclusion is that the effects of antenna locations other than at 1000 feet from threshold can be analyzed using the sensitivity curves by adjusting the value of deviation obtained from the curve by the same amount that the different antenna loca­tion would produce in the RDH or ARDH at an ideal site.

The results of this work are not limited by the arbitrary choice of antenna location. The results of the sensitivity analyses are given in deviation from the ideal values. Therefore, the sensitivity curves give an indica­tion, based on terrain considerations, of how much the reference datum heights can be expected to deviate from the values that would be produced by an ideal site.

The effects of varying path angle, in the presence of rough terrain, pro­duce different sensitivity curves than the ones described in this report for a 3.0 degree angle. Sensitivity curves for path angles of 2.7, 3.0, and 3.3 degrees for a terrain irregularity are shown in figure 2-250. As would be expected, the lower value of path angle makes the refe~ence datum heights more dependent on terrain that is farther from the antennas. The results of this work are based on a 3.0 degree path angle, since it is the

-79-

most commonly found in practice. While it is recognized that the siting engineer cannot select the path angle, the effects of different path angles on the limits of terrain irregularities that can be allowed should be con­sidered when evaluating the site using the information provided by this work.

An album of sensitivity curves for a number of representative ground plane irregularities is included in the next section. The curves are used as a guide in predicting whether or not a proposed glide slope site will satisfy the required tolerances for RDH and ARDH. A method for using the sen­sitivity plots to evaluate a site with respect to FAA Order 8240.47 is pre­sented elsewhere. Terrain regions to which the reference datum heights are most sensitive to roughness are identified.

Uniform Terrain Slopes.

Lateral Slopes. A typical glide slope site has terrain that slopes down laterally from the runway centerline. This is generally the case since the runway must be provided with adequate water drainage. The base of the antenna mast is typically positioned below the elevation of the runway abeam the mast. Assuming the terrain slopes down at a uniform, or constant, rate to the glide slope antennas, the ideal zero DDM conic in space is tilted relative to the vertical by the same angle the terrain is tilted with respect to the horizontal (see figure 2-251). The horizontal of course is the zero elevation plane in a standard gravitational reference system. For small lateral gradients the path in space is nearly identical to the ideal path that would be produced if the ground were perfectly flat at the elevation of the runway [56]. This is understood by recognizing that the zero DDM path to the threshold is formed by intersection of the zero DDM conic and the localizer on-course (runway centerline), and is therefore raised by tilting the cone laterally relative to the vertical. This increase, however, is almost identically offset by the base of the antenna mast being located below the elevation of the centerline abeam the mast. The exact equation governing the effects of uniform, lateral terrain slopes on the path in space is determined in appendix I.

Longitudinal Slopes. To understand the effects of uniform, longitudinal terrain slopes on the reference datum heights, consider the ideal case illustrated in figure 2-252. If the terrain serving as the ground plane remains perfectly flat but is rotated with respect to the y-axis, the path in space is rotated by the same amount. This assumes that the angular displacement is small enough that phasing errors do not occur due to the transmitting antennas not being located at equal distances from the receiver. Since the only effect on the path in space is this angular rota­tion, the best-fit straight line determined by the regression analysis (used to determine the reference datum heights) is rotated by the same angle, as shown in figure 2-253. Since we have assumed the angle of rota­tion is small, the height of the best-fit straight line (RDH or ARDH) is therefore adjusted by the difference in elevation between the reference location and the threshold. This is similar to the computation of TCH for longitudinal slopes [57].

-80-

Note that in figure 2-252 the values of RDH, ARDH, and TCH are not exactly the same. This is caused by the hyperbolic shape of the ideal zero DDM path. The TCH is defined as the height at threshold of the asymptote ·of the zero DDM hyperbola. The RDH and ARDH are defined as the heights at threshold of best-fit straight lines through samples of the path in space. The hyperbolic shape of the path in space results in the slope of the best­fit straight line being slightly less than the slope of the asymptote to the hyperbola. The smaller slope produces a greater height of the best-fit straight line at threshold, and hence the RDH and ARDH will be slightly higher than the TCH. The ARDH ts sampled over a region of the approach that is closer to the threshold than RDH, and therefore the effect is more pronounced.

SPECIFIC FINDINGS.

Irregular (rough) Terrain. A number of sensitivity plots are presented in this section for representative shapes and sizes of irregular terrain. The plots are used to determine whether a specific terrain irregularity will adversely affect the reference datum heights produced.

The trapezoidal-shaped terrain irregularity is used throughout the rest of this report. Both upsloping and downsloping (with respect to the horizon­tal) regions of irregular terrain are considered. Figures 2-254 and 2-255 are sensitivity plots for upsloping irregular terrain regions extending for 10 and 20 feet, respectively. Each plot shows the effects of increasing the height of the irregular terrain from 2 to lO feet in increments .of 2 feet. Figures 2-256 and 2-257 show the effects of similar, but inverted, geometries. The depths range from -2 to -10 feet in increments of 2 feet. Inspection of these results leads to some important conclusions. As one would expect, the reference datum heights vary more widely when the height of a given terrain irregularity increases. Also, it is intuitive that the RDH should depend more on terrain at greater distances from the antennas than ARDH, since RDH is based on path angle samples in a region farther from the antennas. This is exactly the case as shown by the plots. The ARDH variation becomes small past 3000 feet, while the RDH still varies relatively widely. In addition, the reference datum heights are seen to be very insensitive to finite regions of depressed terrain less than 20 feet long. The RDH and ARDH are both insensitive to the extent of the elevated region. To examine this more closely, consider figure 2-258, which shows two terrain profiles for an upslope beginning at 1400 feet from the anten­nas, and rising to 10 feet by 1500 feet from the antennas. Figure 2-259 shows simulated differential amplifier recordings for these two terrain profiles. Inspection of this figure shows that the path formed in space is relatively independent of the terrain that is farther out than the elevated region.

To determine at what extent terrain depressions begin to affect the reference datum heights, sensitivity plots for increasing values of width are considered. Figures 2-260 through 2-263 show sets of sensitivity cur­ves for terrain depressions with widths of 100, 200, 300, and 400 feet, and depths of 5, 10, 15, and 20 feet. These figures indicate that at 200 feet the effects of the depression on reference datum heights become appre-

-81-

ciable. As figures 2-263 and 2-264 show, increasing the extent of the terrain depression more produces sensitivity curves that are similar to to those produced by upslopes in the reflecting terrain. As the depression becomes longer in extent, or wider, more diffracted energy is incident on the upsloping region at the far end of the terrain depression. Figure 2-264 shows a set of sensitivity curves for a terrain depression 50 feet deep, for increasing values of width from 100 to 500 feet. The figure clearly indicates that as the extent of the terrain depression increases, so does the variation of RDH and ARDH from the ideal. As is the case for elevated regions of the terrain, the RDH is more sensitive to the regions of irregular terrain than is the ARDH.

Uniform Terrain Slopes.

Lateral Slopes. The reference datum heights are very insensitive to uni­form, lateral terrain slopes. The sensitivities of RDH and ARDH to this type of reflecting terrain are shown in figures 2-245 and 2-246, showing that for all practical purposes the effects of uniform, lateral slopes on image type glide slope systems are negligible. Both RDH and ARDH increase slightly with increasing lateral terrain slope. For slopes as great as 3% (which is the practical limit produced by the the 0 AP 8200.1 [58] specifi­cation of maximum allowable path tilt) the reference datum heights differ from the ideal case by less than than 0.7 foot.

Longitudinal Slopes. The effects of uniform, longitudinal terrain slopes on the reference datum heights are simply additive, as is the case with the calculation of TCH. The RDH and ARDH are adjusted by the difference in elevation between the reference location and the threshold as specified in FAA Order 8240.47. Thus for a longitudinal upslope the reference datum heights are less than the values obtained for an ideal site. A longitudi­nal downslope produces a higher RDH and ARDH than the ideal site produces.

DISCUSSION.

In this section some information which has been obtained in field measure­ments, and is applicable to siting a glide slope consistent with FAA Order 8240.47, is presented. Two specific sites are considered.

Wheeling, WV (HLG). The glide slope serving Runway 3 at the Ohio County Airport, Wheeling, WV has a history of structure roughness. The approach plate for the glide slope indicates a TCH of 32 feet. In 1984 the facility was moved back from the threshold 200 feet more to its present location at 1250 feet from threshold and 200 feet off centerline.

When the recommissioning flight check was made FAA flight inspection found an out-of-tolerance condition with respect to the structure and the reference datum height. Flight measurements taken by Ohio University correlated with FAA data [59], and produced an RDH of 76 feet.

A level run at the site produced the recording shown in figure 2-265. A recording of the differential amplifier trace made on a low approach is shown in figure 2-266, showing marginal Category I structure.

-82-

A long row of tall trees running parallel to the centerline and about 250 feet off centerline on the same side as the glide slope antennas was suspected as the cause of the structure roughness. A picture of the tree line is shown in figure 2-267. A low approach along the tree line produced the recording shown in figure 2-268. The flare is due to the reference cone being established for a low approach on centerline, but more impor­tantly the structure appears to be excellent and certainly within Category I tolerances, suggesting that multipath from the line of trees is the cause of the long-period structure roughness.

On recommendation from Ohio University, FAA contracted to have the trees removed. A low approach recording after removal of the trees is shown in figure 2-269 [60]. Clearly all structure tolerances are met, and the RDH of 52 feet produced is also well within applicable tolerances, which for Height Group 1 aircraft is 30 to 60 feet.

The important lesson from this field experience is that reference datum heights can very well depend upon other environmental factors than imper-fect ground planes.

Greenville, SC (GMU). The glide slope serving Runway 18 at the Greenville, SC, Downtown Airport is an outstanding example of a runway pedestal site. The runway exists on an 18-foot pedestal with respect to the base of the antenna mast. In addition, the runway slopes down 8.7 feet to the threshold from abeam the mast.

FAA flight checks showed an out-of-tolerance flare downward at ILS Point B. Relocating the reference location in accordance with Order 8240.47 to 600 feet from threshold produced the low approach recording reproduced in figure 2-270, which still shows excessive flare [61]. The facility was made to perform within the tolerances by advancing the phase of the current in the upper antenna by 20 degrees [62]. The path angle, which had earlier been raised by 0.15 degree, also provided compensation. Figure 2-271 shows a low approach to the runway after the electrical modification.

The conclusions are that electrical modifications to the capture effect glide slope system can minimize the requirements for grading, and relo­cating the reference location in accordance with FAA Order 8240.47, pro­viding for the first time decoupling of the reference from the antenna mast, can allow better accommodation of the runway pedestal.

RECOMMENDATIONS.

Based on the data and results examined in the performance of this work, the recommendations are as follows:

1. The results of the sensitivity analyses should be used to eval­uate glide slope sites in terms of what deviations from the ideal values of reference datum heights can be expected.

2. At sites where elevated regions are present in the reflecting terrain, the use of a capture effect glide slope system should be considered.

-83-

3. Information on glidepath shaping methods for the capture effect glide slope system should be obtained and documented to facili­tate producing a path structure that satisfies the required reference datum height tolerances. Experimental data indicate that requirements for ground plane grading can be minimized by phasing the system consistent with the particular ground plane.

4. Other considerations in addition to the ground terrain should be taken into account when siting a glide slope, since data show that sources of multipath can cause out-of-tolerance reference datum heights.

-84-

PROPOSED METHODOLOGY FOR LANDING SYSTEMS MATHEMATICAL MODEL VALIDATION.

EXECUTIVE SUMMARY.

A methodology for the validation of landing systems with mathematical models is presented. It can be used to assess quantitatively the perfor­mance of and/or to determine the validity of a particular instrument landing system (ILS) model used for different types of terrain. In addi­tion the process can be used for estimating, with greater accuracy, a model's capability to predict the effects of terrain on a glide path at a virgin site.

The process is described in detail with a demonstration using data collected at five ILS glide slope sites, which represents four of the five categories of glide slope sites (with respect to terrain). The methodology uses RMS values of the differences between measured and modeled (or calcu­lated) data for a given site. The RMS value for each run is averaged from the same type of sites and are used as an indication of the level of per­formance. A rigorous statistical analysis could not be performed because of the limited amount of measurement data available. A larger data base would increase confidence in a model's prediction accuracy.

INTRODUCTION.

Purpose. The purpose of this project is to develop a mathematical model validation technique which can be used to assess quantitatively the perfor­mance of and/or to determine the validity of a particular instrument landing system (ILS) model for different types of terrain. In addition, the process is expanded to ascribe a probable accuracy to the prediction of the path in space of a virgin site. The accuracy of the prediction is a function of the·model used, type of site, and type of facility.

Background. The use of mathematical models to determine the performance of an ILS has long been a method of reducing the requirement for on-site eva­luations and/or extreme grading of sites to ensure adequate performance. This use of the models has resulted in a requirement to dete~mine the vali­dity of such use and to determine what level of confidence might be placed on the calculations. The types of facilities which it is desirable to ascribe a probable accuracy are null reference, sideband reference, and capture effect. These types of sites, based on reflecting ground terrain, are given in the ILS siting manual [63] (see figure 2-272).

Ideally, with a complete data base, the matrix shown in figure 2-273 would be complete with an average root-mean-square (RMS) value to quantify the accuracy of predicting the glide slope performance for each case. The matrix can be repeated for each model being evaluated.

METHODOLOGY.

A flowchart of the methodology process is shown in figure 2-274. This chart shows the steps involved from the initial data collection to obtaining the average Root Mean Square (RMS) values for various type sites

-85-

and various type facilities. Glide slope facilities were used; however, this procedure could be used on localizers as well. The assumptions made for this procedure are:

1. The true glide path scalloping due to terrain roug~ness does not change during consecutive flight data collections.

2. The portion of Zone 2 used in this procedure, 20,000 feet to 3,500 feet, is representative of the model's capability to predict the glide path in space for a given site.

3. The modeling engineer is experienced in choosing plates to repre­sent the ground terrain, and the terrain input data are therefore representative of the site under investigation.

4. Errors introduced by theodolite tracking, flight recording digi­tization, and terrain plate approximations are random and come from independent normal distributions.

The collection of field data is required, whether the model in question is used to calculate site performance, or if it is used in the validation pro­cess. Normally, the data required to predict a site are the location of the antenna system, the antenna current distribution, and sufficient terrain information to provide the necessary reflection data. Of these, the terrain data are generally the most difficult to obtain, since the eva­luation of quantity and quality is required and this may vary from site to site.

In addition to the data required to run the normal model, the validation model run requires the collection of aircraft position information and measured course deviation indicator (CDI) currents. The latter data sets are obtained through the use of a flight measurement package, in this case the Ohio University Mini-Lab Mark Ilia and a Radio Telemetering Theodolite (RTT). Furthermore, the recorded data sets are used also as the comparing reference.

Once the field data have been made· available, they must be manually entered into the computer. Antenna locations, antenna currents, and reference location information are typed into an input data file. Terrain infor­mation is usually obtained from contour maps provided by the facility. Then the reflection plate data are typed into the input data file based upon the modeling engineer's site evaluation.

Aircraft RTT-based position information and the measured CDI data recor­dings are digitized on magnetic tape and read into data files. The digi­tized data are then converted to CDI current values versus distance, and are sampled at SO-foot increments. The RTT and measured data files are edited and data associated with distances greater than 20,000 feet and less than 3,500 feet are eliminated. This leaves a constant 331 points for each model run and for each comparison with the measured data.

With a flight speed of 200 ft/sec, the sampling rate is four samples per second. Using the basic Nyquist criterion for sampled systems, a minimum

-86-

of two samples per period (2 HZ) is required to reproduce the highest fre­quency component. This is well above the system bandwidth of 0.5 Hz. The sampled measured CDI data are placed in data files for later use. The RTT and distance information is converted to rectangular coordinate position information to provide the aircraft location for the model run.

Having obtained the required input data files, the model is used to calcu­late the CDI current at the point in space representing the aircraft posi­tion as provided by the RTT data. This results in a model output data file which is a calculation of the measured CDI current. The time series analy­sis is now performed on both the model output data set and the measured CDI data set.

Before calculating the RMS of the difference between the two data sets, the bias is removed and the data are filtered. The justification for removing the bias is presented in appendix J. The filter used to remove the high frequency components of the data sets is·described by the following equation:

Y(N)= SR*(X(N)+X(N-l))-Y(N-l)*(SR-2*TC) ( SR+2*TC)

where SR = sampling -ra-te-------­TC = time constant X(N) = input data Y(N) = filtered output

The time constant used on these data is equal to .424 second which is the published standard of ICAO [64]. The sampling rate is the inverse of the sampling frequency and therefore is .25 second. The frequency response of this filter is presented in figure 2-275.

The RMS of the difference is calculated using the following equation:

RMS

where R(d) = M(d) =

N i

the measured CDI as a function of distance d the calculated CDI as a function of distance d the number of data samples the sample number

If the computer model and input data are accurate, the calculated CDI is very close to the measured CDI, and the RMS approaches zero.

The RMS values of runs from a given site are averaged with RMS values from other sites of the same type of site and facility (null reference, sideband reference and capture effect). The following formula is used in the calcu­lation:

-87-

w Avg. RMS of same type site = and same type facility

1 w I Avg. RMS for site P

P=1

where W p

= Number of sites of the same type site and glide slope facility Site number

The primary assumption is that the individual RMS values obtained for each run come from a normal distribution about some unknown mean for all runs for a given type of site ·and a given type of facility (i.e. null reference, sideband reference, and capture effect). The RMS values then can be grouped in an array, and the mean and standard deviation calculated using the following equations:

sample mean = ~ =

standard deviation

N 1 I xi N i=1

SD = / I

where x = the RMS value for a single run N = number of individual runs

As flight measurements from more sites become available, the average RMS value representing the model's capability to predict the glide path of that site should be included in the existing RMS data base, and a new average RMS value and standard deviation value calculated. The new average RMS and standard deviation will represent an improved estimate, since it includes more sites. If the accuracy of the facility data and terrain inputs are the same, the average site RMS values should be similar.

Then the average RMS can be used to compare the outputs of the individual mathematical models. The model with the lowest average RMS value is recom­mended for use in future work involving that type of site.

A confidence interval for the sample mean of the distribution of RMS values can be calculated using the following equation.

where

SD X = ~ ± t*

I n

X = actual mean of assumed normal distribution ~ = sample mean (average RMS)

SD = standard deviation n = number of samples (runs) t number associated with % confidence desired (table 2-17)

-88-

% C 0 N. F I D E N C E

R 0.90 0.95 0.975

1 3.076 6.314 12.706 2 1.6s6 2.92'21 4.303 3 1.636 2.3~3 3. 182 4 1.!533 2.132 2.776 :5 1.476 2.01:5 2.:571 6 1.440 1.943 2.447 7 1.41:5 1.6'3!5 2.36!5 a 1.397 1.660 2.306 9 1.383 1.633 2.262

10 1.372 1.612 2.2::9 11 1.363 1. 7'36 2.201 12 1.3:5$ 1.782 2.179 13 1.3:50 1.771 2. 160 14 1.34:5 1.761 2.145 1:5 1.341 1.7~3 2. 131 1S 1.337 1. 746 2.120 17 1.333 1.740 2. 11~ 18 1.330 1.734 2.101 19 1.323 1.72'3 2.0'33 20 1.32:5 1.72:5 2.066 21 1.323 1.721 2.080 2a 1.321 1. 717 2.a74 23 1.319 1.714 2.069 2'+ 1. 316 1. 711 2.064 2:5 1.316 1.708 2.060 as 1.31:5 1.706 2.0:56 27 1. 314 1.703 2.0~2 2S 1.313 1.701 2.048 2'3 1. 311 1.6'39 2.04:5 30 1.310 1.697 2.042

R=NUMBER OF SAMPLES -1

TABLE FROM REFERENCE 3

0.99

31.621 6.96S 4.~41

3 •. 747 3.36~ 3.143 2.996 2.8'36 2.621 2.764 2.718 2.681 2.6~0

2.S24 2.602 2.:583 2.:567 2.~~::

2.:539 2.:528 2.:518 2.:508 2.:500 2.492 2.4S:5 2.47'3 2.473 2.467 2.462 2.4:57

0.995

63.657 9.925 5.641 4.604 4.032 3.707 3.4'39 3.3:55 3.;::50 3.169 3. 106 3.'21:5!5 3.012 2.977 2.947 2.921 2.8'38 2.878 2.861 2.84:5 2.831 2.81'3 2.607 2.797 2.787 2.77'3 2.771 2.763 2.7~6 2.7:50

Table 2-17. "t" values for calculating confidence intervals.

-89-

Since the amount of data available is limited, the confidence interval will decrease as the number of RMS values increases [65].

This methodology provides information for estimating, with greater accuracy, a model's capability for predicting the effects of terrain on a glide path in space. Unfortunately, statistical methods for handling dependent string data do not exist, which prevents the determination of an absolute confidence level of a model's performance on a virgin site. However, with some practical experience and insight into the ILS, these values will help in making a decision as to what type of facility will operate within FAA tolerances at any virgin site.

,IMPLEMENTATION.

The implementation of the validation process is· now demonstrated using spe­cific examples. The first example uses the Tamiami (TMB) Type-1 site.

Again following the flowchart of figure 2-274, the field data required are collected. Figure 2-276 shows a sample terrain data profile that might be made available to field engineers. This particular data set would be con­sidered adequate but not completely accurate for this site. The modeling engineer then enters this terrain data into the appropriate input data file.

Antenna position information is obtained from measurements, and the antenna current values are measured using a probing technique adequate for the type of antenna in use. These data are now put into the model input file.

The only other information required to run the model calculation for the validation is the aircraft position information. As stated previously, this information is obtained through the use of the RTT and event marks placed on the measurements by the flight technician. A sample of the RTT information after digitization and processing is shown in figure 2-277. The RTT trace is an analog of the aircraft position as measured in microam­peres of current deviation based upon the RTT calibration. At TMB, the theodolite is calibrated for a 3.00 degree glide slope with a path width of 0.7 degree. To convert this to actual aircraft position, the RTT trace is digitized and then processed in a computer routine which converts the current measurements and the distance provided by event marks to distance and elevation information. In addition, the aircraft are assumed to be on an extension of the runway centerline. Table 2-18 shows the rectangular coordinates obtained for this particular run. This completes the required input data for running the model. -

The model is now used to compute the CDI currents measured by the aircraft at the positions computed above. The measured CDI currents are digitized and sampled at 50-foot intervals to provide data values at the points calculated by the model. Figure 2-278 shows the measured raw CDI trace directly from the aircraft recorder, and figure 2-279 shows the same trace following the digitization and sampling process. Figure 2-280 is the calculation provided by the model.

-90-

X y z o.o 20000.0000 '107 3.860 o.o 19950.0000 1070.5650 o.o 19900.0000 1067.6450 o.o 19850.0000 1064.9656 z o.o 19800.0000 1062.5852 o.o 19750.0000 1060.5123 o.o 19700.0000 1058.3230 o.o 19650.0000 1056.0377 o.o 19600.0000 1053.8012 o.o 19550.0000 1051.2359 o.o 19500.0000 1049.1078 o.o 19450.0000 1046.6707 y o.o 19400.0000 1043.7065 o.o 19350.0000 1041.6147 o.o 19300.0000 1039.2333 o.o 19250.0000 1037.1087 o.o 19200.0000 1035.0086 o.o 19150.0000 1032.4397 o.o 19100.0000 1029.9623 X o.o 19050.0000 1027.3648 o.o 19000.0000 1024.8514 o.o 18950.0000 1022.2867 o.o 18900.0000 1019.5927 o.o 18850.0000 1016.7711 o.o 18800.0000 1013.8987 o.o 18750.0000 1011.0188 o.o 18700.0000 1007.9039 o.o 18650.0000 1005.1530 o.o 18600.0000 1002.17 56 o.o 18550.0000 999.2256 o.o 18500.0000 996.5267 o.o 18450.0000 993.5531 o.o 18400.0000 991.1318 o.o 18350.0000 988.4659 o.o 18300.0000 985.5350 o.o 18250.0000 982.7092 o.o 18200.0000 979.9284 o.o 18150.0000 977.5110 o.o 18100.0000 975.1679 o.o 18050.0000 973.2592 o.o 18000.0000 970.8079 o.o 17950.0000 968.1107 o.o 17900.0000 965.4020 o.o 17850.0000 962.7285 o.o 17800.0000 959.8123 o.o 17750.0000 957.3422 o.o 17700.0000 954.4741 o.o 17650.0000 951.8047 o.o 17600.0000 949.2629 o.o 17550.0000 946.5698 o.o 17500.0000 943.7531 o.o 17450.0000 941.1836 o.o 17400.0000 938.7735 o.o 17350.0000 936.0796 o.o 17300.0000 933.3857 o.o 17250.0000 929.9191 o.o 17200.0000 926.5992 o.o 17150.0000 923.4312 o.o 17100.0000 920.4372 o.o 17050.0000 918.2525

Table 2-18. Rectangular coordinates converted from RTT CDI data.

-91-

Figure 2-281 shows both the measured raw CD! trace and the model output trace. Visually, the traces appear very similar except for an average 10 microampere bias error. Part of the bias is caused by the use of an incorrect theodolite reference angle during the RTT measurements (3.0 degrees versus 3.07 degrees actual path). The other part is from the measurements of the antenna heights above the average ground in the neigh­borhood of the mast. Since the range of 3,500 to 20,000 feet is con­sidered, this error does not affect the comparison process. In addition, the skewing between the traces has been tracked to a phase adjustment problem in the measured system. This improper adj~stment between the antenna elements introduces the angular displacement with distance as noted in the figure.

The bias is removed before statistical analysis is performed because the bias varies substantially from one site to another, thus preventing con­sistency between sites of the same type. Justification for this is stated in appendix K. The 331 CD! values are summed and the total is divided by 331 to obtain the bias. This bias is subtracted from each of the 331 CD! values. This process is done on both the calculated and measured data. The resulting traces are filtered through a first order low pass filter, described earlier, and the resulting traces are shown in figure 2-282. (Note, a different set of data are used for examples of the filtered, zero­mean, CD! traces.)

The RMS of the difference between the calculated and measured CD! data is calculated from the filtered, zero-mean data file (with bias removed) and is printed on the plots with the traces.

The RMS value of this run is averaged with other runs of the same site and combined with the existing average RMS value of the same type site and same type facility. The standard deviation of this set of RMS values is also calculated.

A confidence interval is calculated for the RMS value for the Type-1 site with a capture effect system. Using the 95% probability results in a con­fidence interval of 2.66 < ~ < 4.68 for the sample mean (averaged RMS).

Table 2-19 is the existing data base for a Type-1 site with a capture effect system. Included in this table is the average RMS value, standard deviation, and confidence interval of the average RMS value. Figures 2-283 and 2-284 are filtered zero-mean traces for the data included in table 2-19.

EXISTING DATA BASE.

This section presents the limited data base presently available for the math model validation procedure. For a Type-1 site three runs exist of Tamiami (TMB). The plots for these were shown in the previous section. The terrain profile is shown in figure 2-285. The corresponding table of RMS values was also shown in the previous section.

For a Type-3 site with a capture effect facility, three runs at both Lawton (LAW) and Kansas City (PAJ) exist. The respective terrain profiles are

-92-

SITE RUN tt RMS

TMB 1011 3.77

TMB 1012 4.21

TMB 1013 3.03

AVERAGE RMS = 3.67

STANDARD DEVIATION= 0.59

95% CONFIDENCE INTERVAL OF THE AVERAGE RMS = 2.66 <X< 4.68

Table 2-19. Type-1 site data base with capture effect facility.

-93-

presented in figures 2-286 and 2-287. Figures 2-288 through 2-293 are the filte~ed, zero-mean traces of the six runs and table 2-20 is the resulting RMS data set with the standard deviation and average RMS confidence inter­val.

For a Type-S site with a capture effect facility two runs exist for Springfield (SGH). The terrain profile is shown in figure 2-294 and filtered, zero-mean traces are presented in figures 2-295 and 2-296. Table 2-21 is the resulting data base.

For a Type-4 site with a capture effect facility two runs exist of Redbird (RDB). The terrain profile is shown in figure 2-297. Figures 2-298 and 2-299 are the zero-mean, filtered, CDI traces for RDB and table 2-22 1s the Type-4 data base.

SUMMARY AND CONCLUSIONS.

The purpose of this project is to develop a mathematical model validation technique which can be used to assess quantitatively the performance of and/or to determine the validity of a particular instrument landing system (ILS) model used for different types of terrain. In addition, the process can be used for estimating, with greater accuracy, a model's capability to predict the effects of terrain on a glide path at a virgin site. Mathematical programs and statistical techniques were developed which allow the modeling engineer to obtain these results.

Conclusions reached as a result of this study are:

1. The RMS values of the difference between the measured and calcu­lated CDI traces taken at 331 points in ILS Zone 2 for a given site are effective in quantifying the model's capability to pre­dict the glide path structure of that site.

2. The significance of the mean and standard deviation of the RMS values for the different types of sites increases as the data base increases.

3. The limited amount of measurement data available precludes having a high degree of confidence in a rigorous statistical analysis.

4. As was expected, the Type-1 site is predicted more accurately than the Type-3 site. This is shown in the magnitude of the respective average RMS values of 3.67 and 5.62.

5. The lack of randomness in the differences between the measured and calculated CDI traces prevents more sophisticated statistical ana­lysis to be performed on the data. This lack of randomness is due to dependence of one data point on the adjacent points. In other words, the CDI values do not change rapidly with distance.

-94-

SITE RUN# RMS

PAJ 1608 4.89

PAJ 1610 4.66

PAJ 1611 5.44

LAW 2603 5.·54

LAW 2604 7.28

LAW 2605 5.88

AVERAGE RMS = 5.615

STANDARD DEVIATION= 0.929

95% CONFIDENCE INTERVAL OF THE AVERAGE RMS : 4.65 < X < 6.36

Table 2-20. Type-3 site data base with capture effect facility.

-95-

SITE

SGH

SGH

RUN •

2319

2320

RMS

8.64

9.45

AVERAGE RMS = 9.05

STANDARD DEVIATION= 0.57

95% CONFIDENCE INTERVAL OF THE AVERAGE RMS = 6.05 < X < 1 1.59

Table 2-21. Type-S site data base with capture effect facility.

-96-

SITE

ROB ROB

RUN# RMS

2504 5.66

2507 3.90

AVERAGE RMS = 476

STANDARD DEVIATION = 1.24

95% CONFIDENCE INTERVAL OF THE AVERAGE RMS = -0.77 < X < 10.34

Table 2-22. Type-4 site data base with capture effect facility.

-97-

USERS'S MANUAL FOR LANDING SYSTEMS MATHEMATICAL MODEL VALIDATION PROCEDURE.

INTRODUCTION.

This manual describes the use of the ILS mathematical model validation pro­cedure developed at Ohio University. The validation procedure consists of statistically analyzing the field data collected on-site and the model out­put data created from the field data. The root mean square (RMS) of the differences between the measured and calculated CDI currents is calculated using 331 data points. These 331 data points are CDI current values taken at 50-foot increments beginning at 20,000 feet through 3,500 feet. There is an RMS value calculated for each run on a given glide slope site. The RMS values are averaged for sites of the same type of terrain and type of facility (i.e. null reference, sideband reference, and capture effect). · This averaged RMS value is useful for comparing one model's capability to another on the same site, and to quantify a model's capability to predict the glide path structure of a given type of site. The figures and data used to illustrate this manual were obtained from the Ohio University Tamiami test site.

DISCUSSION.

Preparation. Prior to beginning the validation procedure, field data be collected from the site to be modeled--and --a- --cef'-I'-B.in -profile of the slope site must be compiled from topographical maps of the facility. data are used to produce the model output applied in the validation.

must glide These

Field Data Collection. The collected field data consist of the location and specifications of the antenna system, the antenna current distribu­tions, terrain information to provide reflection data, and aircraft posi­tion information and raw course deviation indicator (CDI) currents.

The locations and heights of the glide slope antennas are measured relative to the runway ground point of intercept (GPI). The magnitudes and phases of the antenna currents should be measured using an appropriate probing technique or the ideal currents for the specific system may be used if a proper measurement technique is not available. In most cases, the currents are normalized to the middle antenna.

Obtaining the terrain data probably presents the most difficulty. Eval­uation of how much data is required varies from site to sit~. It includes site features such as soil type, ground cover, lakes, hills, ditches, and man-made structures. These data are considered when constructing the terrain profile and when executing the model. All of the above data are identical to that required for a normal model run [66].

Several glide slope approaches are flown, and the differential CDI (this is the difference between aircraft position and raw CDI) and raw CDI currents are recorded on a chart recorder (figure 2-300). The raw CDI for the examples used in this manual were recorded on the Ohio University Mini-Lab Mark III. The aircraft position is uplinked from a radio telemetering theodolite (RTT) which tracks ·the aircraft from the outer marker to the

-98-

threshold. Event marks, corresponding to landmarks whose distances from the threshold are known, are recorded as the glide slope is being flown. These are later used to create the distance versus CDI files used in the validation procedure.

Trace Digitization. The traces recorded during the glide slope runs are examined and those selected for the validation procedure are digitized. · The digitization process used at Ohio University utilizes a chart recorder, an analog-to-digital (a/d) data translator, and a Rockwell AIM-65 microcom­puter working with the main computer (IBM/370). The data are placed in the computer files and converted into a format usable by the validation programs.

Compilation of the Terrain Profile. The glide slope model requires a· terrain profile as a portion of the input. The terrain is approximated by a series of reflection plates. Terrain data collected by a field crew may assist in the selection of these plates. An effective plate selection method is to use the Rayleigh criteria (appendix K) to determine whether or not a terrain variation will affect the signal.

If a 2-D representation is selected, the terrain variations along a line outward from the antenna mast to the middle marker are considered. This profile is constructed by listing the x, y, and z c.oordinates of the terrain profile corners. Since the terrain is considered to be invariant in the x direction, the x coordinate is usually zero (figure 2-301). The 2-D terrain profile is included in the model input data file. Construction of this file is explained in detail in the next section.

For a more precise representation of the terrain, a 3-D terrain profile is constructed. This is a significantly more accurate model of the terrain since the terrain is not held invariant in the +x direction. The plate selection process remains the same as for a 2-D profile. Once the reflec­tion plates are selected, the x, y, and z coordinates of several points

·along the terrain profile corner are listed in order from the runway cen­terline. The first set of coordinates represents the terrain on the -x side of the centerline. The x coordinate is set at -10000.0 and the y and z coordinates are those where the profile corner intersects the centerline. The terrain on the -x side of the centerline is thus considered invariant in the x direction. The x, y, and z coordinates of the remaining points along the profile corner follow. For convenience, the points where the contour lines of the map and the profile corner intersect may be selected. This is repeated for each profile corner from the mast to the middle marker. When modeling using a 3-D profile, no profile information is con­tained in the model input data file. It is contained in a separate file which is called from the model.

The first line in a 3-D terrain profile file is formatted 2I5. The first variable is the number of elevation samples along the x-axis and the second variable the number of terrain profile corners. The x, y, and z coor­dinates of the 3-D profile data follow and are formatted 3F10.3 (figure 2-302). For further detail on constructing the profile, see appendix K.

-99-

Constructin the Model In ut Data File. The model input data are con­structed by running HEADER figure 2-303). HEADER creates the model input data file from the responses to its questions about the field data. The data for this input file again are identical to that for a regular model [67]. The difference is that, for validation purposes, the model makes its calculations at specified points and not at the points it creates. The input file consists of one comment line and five sections of numerical data.

The comment line is used for the identification of the particular valida­tion model run. This is followed by the terrain data in section 1. The number of corners in the 2-D terrain profile is entered. If a 3-D profile is being used, one or zero is entered. The 3-D profile is then accessed as a separate file.

The specific model is selected by entering:

0 for the Flat Earth Model 1 for the Geometric Theory of Diffraction (GTU) model 2 for the Physical Optics (PO) model

The Flat Earth model is best used for sites with extremely flat terrain such as the Tamiami Type-1 site. The other models are generally used for rougher terrain with the GTD model usually providing the greater accuracy for the roughest terrain.

The printout of extra information from the GTD and PO models is selected from:

0 for no messages 1 for full ray information for the GTD model and the

integral of the ground currents for the PO 2 for only indicating the existence of the. rays in the

GTD model.

If 1 or 2 are chosen, only a few receiver positions should be specified as a large amount of output for each receiver point will be produced.

If the Flat Earth model is chosen, the complex permittivity and the conduc­tivity of the ground are entered. For perfect conducting ground, the. permittivity and conductivity are set to zero. Typical values of permit­tivity and conductivity for other types of soil are listed in table 2-23 [ 68 ' 69 1 •

When a 2-D profile is being used, the x, y, and z coordinates of each terrain profile corner are entered in order going outward from the antenna mast to the middle marker. This section is skipped when using a 3-D pro­file.

Section 2 contains the theodolite data. The x, y, and z coordinates of the theodolite position with respect to ground point of intercept are entered.

-100-

Permittivity (Real Part,

Ground Conditions Assume '1m = 0) Conductivity

Very moist ground 30 5 X 10-3 to 1 X 10-2

Average ground 15 5 X 10-4 to 5 X 10-3

(Rural areas)

Very dry ground 3 5 x 10-5 to 1 X 10-4

(Urban, industrial areas)

Arctic land 15 5 X 10-4

Polar Ice 3 2. 5· x 10-5

Average ground with 18 5 X 10-4 to 5 x 10-3

vegetation .

Table 2-23. Permittivity and conductivity of selected ground conditions.

-101-

The reference path angle and reference path width are next en~ered. The angles are measured from the theodolite eyepiece relative to the horizon­tal. These angles define the receiver position relative to the theodolite eyepiece.

Section 3 contains the antenna data. The first 6 entries describe the glide slope antenna facility. The program prompts for the number of anten­nas in the array and the frequency of the system in MHz. The next three entries are for the depth of modulation in decimal, the A-ratio, and if applicable, the. depth of modulation of the clearance signal in decimal for­mat. The nominal values of these parameters are 0.4, 0.3 for a 3.0-degree path angle, and 0.9, respectively. The antenna orientation is entered next. It is set to zero if the axes of the dipole antenna are oriented in the x-z plane and .to one if the axes 'are oriented in the y-z plane.

The data for each antenna comprise the remainder of this section. The x, y, and z coordinates of the antenna location are entered. The lower antenna is considered first and the higher antennas_follow in succession. Next, the magnitudes and phases of the complex sideband-only sideband, the carrier sideband, and the clearance-carrier currents, in that order are entered. The nominal current magnitudes for the three common image type systems are as follows:

ISS res ICC NULL REFERENCE lower antenna o.o .1.0 o.o upper antenna 1.0 o.o o.o

SIDEBAND REFERENCE lower antenna -1.0 1.0 o.o upper antenna 1.0 o.o o.o

CAPTURE EFFECT lower antenna -0.5 1.0 0.484 middle antenna 1.0 -0.5 o.o upper antenna -0.5 o.o 0.484

Section 4 contains the pattern data describing the path flown by the aircraft and the receiver during the field data collection. The user is asked to select the type of pattern cut from the six available in the model. Since a hyperbolic cut is flown for the validation procedure, this pattern cut is selected from the offered menu. The number of positions where calculations are to be made is entered next. For the validation pro­cedure, calculations are made at 50-foot intervals from the initial receiver position of 20,000 feet to the final receiver position of 3,500 feet. This yields 331 positions which is considered sufficient for most validation runs. The print output option is set to zero. The user is next asked to input the x and y coordinates of the initial receiver position, the glide path angle, and the x and y coordinates of the final receiver position.

Section 5 contains the output control data which controls and selects the graphical output format when the GTD model is used. The user selects the output graphic from:

-102-

0 for no output graphics 1 for a line printer plot 2 for a Calcomp plot 3 for no output graphics 4 for both line printer and output plots

The x axis variable for these plots is the y distance in feet and the y axis variable is the CDI. These are selected from the menus offered by the program and entered. The maximum and minimum values of the x and y axes are entered next. These values are usually +100.0 and -100.0 for the x axis and 30000.0 and 0.0 for the y axis. The user is next asked to select the GTD rays to be examined in the model. If a ray is to be examined, a zero is entered. If not, a one is entered. After this information is entered, the program ends and the model input data file is stored on unit device 8.

Creating the Memo File. Information needed to run a validation for a par­ticular site is placed in a memo file and is referred to when running the validations (figure 2-304). This file should contain event mark infor­mation, the position of the theodolite, and the glide slope angle. Besides this basic information, any other bits of data deemed significant should be included.

The event mark information supplies the distance data for the distance vs. CDI files created in the validation procedure. This information consists of the distance of the first event mark from the GPI in feet, followed by the distance from the first event mark to the second, followed by the distance from the second event mark to the third, etc.

Example._ If the first event mark is at 31,000 feet and the second is at 25,000 feet and the third is at 20,000 feet, the event mark information would be the following:

31000.0 6000.0 5000.0

This procedure is followed for all event marks used in the glide slope run.

The position of the theodolite relative to the GPI and the glide slope angle are taken directly from the site data.

The preparations are now complete and the validation procedure may begin.

Validation Procedure. The validation procedure involves processing the data through several FORTRAN programs and making decisions based on the output of these programs. Figure 2-305 is the flowchart representation of the validation procedure.

The data must be converted from the data files created when the data are digitized to the distance vs. CDI format used in the validation procedure. The input data files contain two pieces of information per line. The first

-103-

number is a real three-digit number ranging from 000.0 to 999.0. This number indicates the location of the trace on the 10-inch flight recording. A 0.0 ~a CDI corresponds to 500.0 (i.e., the middle of the trace). At the chart edges, 000.0 is equal to -100 ~a CDI and 999.0 equals +100.0 ~a CDI. The second number on the line indicates the presence (set to 1) or the absence (set to 0) of an event mark. A "3" indicates end of data. RCHART is the FORTRAN program used by Ohio University to convert the output of the AIM 65 to a formatted data file on the IBM/370.

The input data are converted to the distance vs. CDI format by running DATACON. The input data are accessed on device 10 and the output is sent to device 8. DATACON also requires user input. The user is asked whether the data originally came from 10-inch or 6-inch charts. This is necessary since older recordings were made on a smaller recorder. The program then asks for the starting value. This is the first value of the event mark information contained in the site memo file. THE NUMBER MUST BE ENTERED WITH THE DECIMAL POINT. Failure to do so will cause the program to ter­minate prematurely. After a correct entry, the computer will send back a message indicating the number of points from that event mark to the next. When the computer prompt returns, enter the next event mark value. This process continues until the last event mark value is entered. The computer completes the data processing and creates a distance vs. CDI data file. Both the measured CDI and RTT input data files are processed through DATACON (figures 2-306 and 2-307).

DATACON creates a record for each data point recorded in the digitizing process. For most applications, this is more than enough data and would waste computer time; therefore, the data are interpolated to SO-foot incre­ments with no loss of accuracy. This is still above the system cutoff fre­quency if the glide slope validations were flown at 200 feet per second._ The program to perform the interpolation is INTERPOL. The output of DATACON is used as the input to INTERPOL. The input files to INTERPOL are accessed on device 2 and the output is put on device 6. No user input is required.

The interpolated output files are edited and the points above 20,000 feet and below 3,500 feet are discarded leaving 331 points of data. This is done because the extreme far-field and near-field data are not modeled accurately due to tracking errors. The edited files are plotted and stored for later use.

The RTT distance vs. CDI data is used as the input to CDITOREC from device 18. After requesting the x, y, and z coordinates of the theodolite and the glide slope angle, CDITOREC converts the RTT distance vs. CDI data to rec­tangular coordinates in the same system as the theodolite (table 2-24). The rectangular coordinates are outputted to device 8.

The model validation procedure requires that the ILS model be run for the glide slope site. This is done to determine what the CDI values should be for a given site under ideal conditions. The only difference between this and a regular model run is that the glide slope position information is not generated by the computer. It is, instead, supplied to the model from the rectangular coordinates generated by CDITOREC.

-104-

X y z o.o 20000.0000 1073.860 o.o 19950.0000 1070.5650 o.o 19900.0000 1067.6450 o.o 19850.0000 1064.9656 z o.o 19800.0000 1062.5852 o.o 19750.0000 1060.5123 o.o 19700.0000 1058.3230 o.o 19650.0000 1056.0377 o.o 19600.0000 1053.8012 o.o 19550.0000 1051.2359 o.o 19500.0000 1049.1078 o.o 19450.0000 1046.6707 y o.o 19400.0000 1043.7065 o.o 19350.0000 1041.6147 o.o 19300.0000 1039.2333 o.o 19250.0000 1037.1087 o.o 19200.0000 1035.0086 o.o 19150.0000 1032.4397 o.o 19100.0000 1029.9623 X o.o 19050.0000 1027.3648 o.o 19000.0000 1024.8514 o.o 18950.0000 1022.2867 o.o 18900.0000 1019.5927 o.o 18850.0000 1016.7711 o.o 18800.0000 1013.8987 o.o 18750.0000 1011.0188 o.o 18700.0000 1007.9039 o.o 18650.0000 1005.1530 o.o 18600.0000 1002.1756 o.o 18550.0000 999.2256 o.o 18500.0000 996.5267 o.o 18450.0000 993.5531 o.o 18400.0000 991.1318 o.o 18350.0000 988.4659 o.o 18300.0000 985.5350 o.o 18250.0000 982.7092 o.o 18200.0000 979.9284 o.o 18150.0000 977.5110 o.o 18100.0000 975.1679 o.o 18050.0000 973.2592 o.o 18000.0000 970.8079 o.o 17950.0000 968.1107 o.o 17900.0000 965.4020 o.o 17850.0000 962.7285 o.o 17800.0000 959.8123 o.o 17750.0000 957.3422 o.o 17700.0000 954.4741 o.o 17650.0000 951.8047 o.o 17600.0000 949.2629 o.o 17 550.0000 946.5698 o.o 17500.0000 943.7531 o.o 17450.0000 941.1836 o.o 17400.0000 938.7735 o.o 17350.0000 936.0796 o.o 17300.0000 933.3857 0.0· 17250.0000 929.9191 o.o 17200.0000 926.5992 o.o 17150.0000 923.4312 o.o 17100.0000 920.4372 o.o 17050.0000 918.2525

Table 2-24. Rectangular coordinates converted from RTT CDI data.

-105-

To create the validation model input file, the three periods at the end of the model input data file created by HEADER are replaced by a blank line and the rectangular coordinate information created by CDITOREC is appended to it (figure 2-308). The validation model input data file is accessed on device 5 by the computer. If a 3-D ·model is being used, the terrain pro­file must be on device 14. After the model has finished running, three output files are created. The files on devices 6 and 10 are general model output files and are not used. The file on device 8 is used in the valida­tion process only. It contains the distance vs. CDI information as calcu­lated by the model. This file is also plotted and stored (figure 2-309).

The two data files are then input to KRMS FORTRAN which filters the data, removes ~he bias (see appendix L), and calculates the RMS of the differen­ces of 331 data points (FREQRESP FORTRAN was used to determine the fre­quency response of the low pass filter used in KRMS FORTRAN). The zero-mean, filtered, CDI traces of the calculated and measured CDI are then plotted with the RMS value (figure 2-310). (Note, a different set of data is used for examples of the filtered, zero-mean, CDI traces.) The RMS value serves to quantify the model's capability to model the given site.

This RMS value is then grouped with others of the same type site and same type facility to attain a new average RMS and standard deviation values. The existing data base for a Type-1 site w~th a capture effect facility is shown in table 2-25.

-106-

-SITE RUN u RMS

TMB 1 011 3.77

TMB 1012 4.21

TMB 1013 3.03

AVERAGE RMS = 3.67

STANDARD DEVIATION= 0.59

95~ CONF I OENCE INTERVAL OF THE AVERAGE RMS = 2.56 < X < 4.68

Table 2-25. Type-1 site data base with capture effect facility.

-107-

CHAPTER III

EDUCATION SUPPORT

SEMINAR ON THE ENDFIRE GLIDE SLOPE.

A seminar on the Watts Endfire Glide Slope antenna system was conducted between June 5 and June 10, 1983 at the FAA Academy in Oklahoma City. The seminar covered installation, maintenance, and operation of. the slotted cable, endfire glide slope system. An overall outline of the material pre­sented is shown in appendix M. Lecture notes amounting to 190 pages were distributed to the attendees. Lecturers were Mr. Chester B. Watts, the inventor of the system, and Dr. Richard H. McFarland, who has worked with the endfire system for over 25 years.

LECTURE ON RELEVANT ISSUES CONCERNING CONSTRUCTION PRACTICES AS THEY MAY AFFECT PROPAGATION OF ELECTROMAGNETIC SIGNALS USED FOR AIR NAVIGATION.

Additionally, a lecture was prepared and given at the National Airspace Meeting held in Ft. Worth, Texas, on July 14, 1983. The subject was con­sideration of obstructions, and the type of construction necessary to mini­mize effects on navigation aids, particularly landing systems. The theme was that the subject involved complex considerations, and that these could best be handled by mathematical modeling to provide quantitative assessments.

-108-

CHAPTER IV

VERIFICATION OF STANDARDS AND TOLERANCES

INITIAL ASSESSMENT OF APPROPRIATENESS OF QUANTITATIVE VALUES USED TO QUALIFY GLIDE SLOPE STRUCTURES.

SUMMARY AND CONCLUSIONS.

The quantitative values established for tolerances on various parameters of the ILS glide slope are extremely important. They must not be too lax or there will be a negative effect on safety. If they are too stringent, then the facility may be more expensive to establish, more dirficult to maintain and there will be unnecessary outages. This means unavailability to the user and as a result, safety will be degraded simply by this absence.

Historical documents do not seem to be complete enough to allow a full knowledge at present of how each tolerance was determined initially. Current structure tolerances have been applied for nearly 40 years for Category I; however, for Category II there have been changes periodically over the past 15 years, especially to those tolerances applied in ILS Zone 3.

This work has been dedicated to exam1n1ng existing glide slope structure tolerances for appropriateness in providing for safety and availability. The approach for the first time has been through the use of sophisticated flight simulation. This work with structures is but a part of a larger picture because there are more than 10 parameters. which need be addressed that conceivably can significantly affect glide slope performance as seen by the aircraft. This specific work, therefore, is but an introduction to a more extensive study recommended for assessment of tolerances. Results indicate that current tolerances for Category II operation, viz,· 30 microamperes tapering to 20 at ILS point B, are good. While looser tolerances could be made acceptable by increased training efforts, many pilots are not afforded recurrent training; therefore, the existing tolerances are considered quite satisfactory.

Tolerances for Zone 3 should relate to how the user can be expected to use the glide slope in this region. Evidence indicates that airborne equip-menta vary with respect to how they process the signal. Some equipment desensitizes completely in Zone 3, so obviously it does not matter at all what quality of signal exists there. For some users, on the other hand, the signal is processed and readout of information is the same as if the aircraft were at the outer marker. Comment on appropriateness cannot be valid until a specification on user data processing is made.

Some specific conclusions derived from the flight simulation work are:

1. The structure tolerance of 30 microamperes tapering to 20 through ILS Zone 2 plus the structure and alignment tolerances for Zone 3 prescribed in OA P 8200.1 217.5 for Category II operations are appropriate. While an increase could be tolerated safely, an

-109-

increase in pilot training would be required, and it is reasonable to expect that general aviation pilots who do not receive recur­rent training might be adversely affected. The tolerances are restrictive enough based on the evidence that since there are no complaints by pilots, safe operation has existed for 40 years, and straight-line paths are considered by some to be represen­tative of the paths in space. The tolerances could be increased at least 20% but pilot training to accommodate this increase would be important to accomplish.

2. The spatial frequency of path roughness is critical to whether it is acceptable. Short period (i.e. high frequency) path or beam noise is only annoying and at the very most, causes consterna­tion and unhappiness. This is comparable to the control motion nois~ (CMN) identified in MLS work. Long-period perturbations in the path are more important because they can produce control actions whereby the aircraft actually follows the perturbed path. This places the aircraft usually in a position whereby more control pressures must be applied to effect the desired track to the airport.

3. The reversal criteria, more technically named the second deriva­tive constraint (rate of change of slope), is very important because it serves to limit the amount of control action that will be required in the aircraft, limit the amount of discomfort to passengers, limit additionally the displacement from a desired track in space, and enhance pilot satisfaction. Additional definition of this tolerance is desirable.

4. Indications are that pure desensitizing of airborne equipments to navigation information is not the best way to provide relief from beam noise. Presently, beam or path noise is bounded thus making it possible to provide filtering and not merely constraining the airborne equipment to ignore the signal. Certain known meteoro­logical conditions can produce a situation whereby the aircraft will miss the desired touchdown. Information from the JAWS pro­ject suggests there is a high enough probability of this occurring that protection should be provided in the airborne equipment. Microbursts of even small magnitudes could produce disaster because vertical positional information is degraded.

5. Moderate turbulence, which is the maximum that can reasonably be expected during conditions of low visibility in fog, does not have a significant, adverse effect on pilot performance. It does raise the work load of the pilot appreciably, may preclude the use of the autopilot, and will increase the probability of a missed approach. If the pilot is not capable of accepting a higher work load, then safety will be degraded.

6. Flight simulation using modern, motion-based simulators is an excellent means for examining appropriateness of landing system beam tolerances and, given enough experience, will allow for a determination of maximum allowable limits.

-110-

7. Although the horizontal and vertical axes are essentially indepen­dent in the autopilot/flight director computer systems, this is not true for the human. It is important to note that increased workload in tracking, say, the glide slope much of the time will produce degradation of pilot performance tracking the localizer. Flights in the simulator clearly showed that while the localizer remained a straight line, tracking of this course worsened as the glide slope roughness increased. In fact, the majority of missed approaches were due to deficiency in localizer tracking in spite of the fact that the localizer was always straight.

8. Pilot satisfaction is related directly to the ease in which a path can be flown and how it appears. A track which had the same magnitudes of displacement (and presumably the same adverse risk factor) was regarded as benign and quite acceptable, presenting no apparent problem.

9. The statistics show the aircraft track deviation to be essen­tially independent of the independent variables, such as whether raw data of flight director was used or what the magnitude was of the structure perturbation. Note should be made that only a 20% excess tolerance case was tested. Response to this was that the pilot was doing what would be expected, viz, tracking the course information. Obviously at some point his tracking capability would deteriorate and become dangerous. At 20% this point had not been reached.

10. The statistical analysis shows the pilot acts as a low pass filter and this, of course, is as would be expected intuitively.

11. Of the 41 statistically valid observations (cases) two were shown to have track deviations of 150 and 160 microamperes meaning that displacements of 0.7 degree were observed. This is considered noteworthy and further attention must be given to this kind of aberration.

12. Pilots appear to bias errors towards the high side which is undoubtedly due to their training that danger lurks on the low side rather than the high.

13. Rougher glide slopes (in excess of present tolerances) should have greater obstacle clearance values.

14. An autocorrelation analysis produced little of interest.

15. Statistical data did not seem to be normally distributed. The analysis of variance, which assumes a normal distribution for validity, casts doubt, therefore, on any conclusion concerning the interaction of the various independent variables on the maximum deviation.

16. Subject pilots performed better with the rough structures than did the autopilot.

-111-

17. The flight director was not useful to the pilot in helping him fly the rough path conditions. It tended to encourage over­controlling.

18. Subject pilots were used to quality ILS's and have been completely satisfied with the structures that have been provided at all air­ports. These structures are represented in the simulator by straight-line paths; because of this satisfaction these straight lines are considered representative of the real world.

19. Any ILS beam roughness that is allowed must be acceptable to flight director systems because pilots have complete trust in their operation. Only in very exceptional cases will the raw data be given preference, hence, if rougher paths are to be considered, the need exists for special experiences in training.

INTRODUCTION AND BACKGROUND.

This investigation is a beginning of an examination of tolerances which are routinely applied to the performance parameters of the instrument landing system glide slope and localizer. FAA Handbook 8200.1 Section 217 prescri­bes the procedures and tolerances which are applied to qualify systems for the aviation community to use.

The relevant numerical tolerance values are contained in Section 217.5, which the flight inspector uses to qualify systems. From preliminary research, it appears that the justification or basis for most of these numerics is lost in history. The question logically comes as to whether considering today's aircraft with performance ranging from supersonic to the basic light, general aviation aircraf:, are the tolerance values opti­mum?

Optimum must be defined with respect to certain references. Safety is the most important. Clearly, the perfect accident record with ILS operation over the past 40 yea~s answers the question as to whether the values are sufficiently conservative. The answer is yes. The following question, are they too conservative, is yet to be answered. When the tolerances are too tight, certain systems are removed from service and the user is deprived of any service. History has shown that accidents can occur when systems are not operating.

It is important, therefore, to insure that the tolerances are not too tight. What is too tight, is the key question. Optimum, therefore, must be regarded as those operating values which provide for safe operation, while still allowing for maximum availability of the system. This, of course, relates to maintainability.

Initially, the structure of the glide slope has been the item of interest. This is the quantity which is related most closely to the delivery of the aircraft at specific heights over the runway threshold. Motivation for this selection has been that at this point in history, there is con­siderable interest in the threshold crossing height, reference datum height, and achieved reference datum height.

-112-

APPROACH.

The approach being used initially to examine the appropriateness of the existing tolerances is through use of flight simulation. The remarkably fine and complete flight simulator equipment available today, made possible through the application of modern solid state technology, provides a tool for careful examination of tolerance values.

There are several techniques that could be employed to examine standards and tolerances values. The obvious one is the use of actual flight. This, of course, is expensive, difficult to execute with precise tolerance values applied, and very time consuming. Another is the compilation of existing and historical data and performing analyses and investigations concerning pilots who use certain, near-tolerance facilities. Additionally, there is computer simulation in which the human or autopilot is modeled and the complete flight activity is simulated. This technique is finding greater acceptance because it allows extensive and exhaustive examination of spe­cified conditions. At this time such a facility is not available, but plans are to consider this approach for future use.

Through the cooperation of Pan American World Airways a flight simulator housed at their International Flight Training Academy, Miami, Florida has been made available for this work. This simulator represents a Boeing

_ 727-200 aircraft and is complete with visual presentations. This equipment is so faithful in representing the total flight regime of the aircraft that it is used to qualify a pilot in the aircraft without his having any actual flight experience. It is FAA-certified for this training.

The plan has been to design three glide path structures which are at the edge of structure tolerance limits for Category II operation. These were programmed into the flight simulator to provide glide slope indications to the pilot that replaced those of the typical straight-line profiles used in the routine work. This, incidentally, was the first time that other than straight-line glide slopes were used in the Pan American simulation. From what information is available, these are the first less-than ideal profiles that have been used in training simulator work anywhere.

DATA BASE.

The Singer-Precision flight simulator is a full motion-based simulator which duplicates well a variety of flight conditions including IMC (Instrument Meteorological Conditions) of various kinds, including those used in "these experiments involving Category II, viz, ceiling of 100 feet and RVR (Runway Visual Range) of 1200 feet. While some tests were run with a Category II lighting configuration inciuding the centerline and touchdown zone lights, some included only threshold and runway lights.

Typically, the simulation provides glide slope information based on trigno­metric calculations assuming dimensions of 400 feet offset of the glide slope antennas with a 1000 feet set back from the threshold. This gives a threshold crossing height of 52 feet and a 56 feet height of the hyperbolic on-course.

-113-

Three different, non-ideal, glide slope structures were.provided on an on­call basis in the simulator. The three different paths shown in figures 4-1, 4-2, and 4-3 were designed and constructed with several objectives in mind. First, the intent was to produce magnitudes of roughness that were at the Category II tolerance limits. Second, there was an interest in trying to produce oscillations that might be near some natural frequency of the pilot and aircraft joint system. Finally, there was a desire to place the path at minimum elevation with a slope which would deliver the aircraft low and with a direction of even further low altitudes.

The astute reader may recognize that the path structures do not meet the criteria for reversals. This requirement on the first derivative of the curves certainly must be regarded as a an important consideration, but for these first tests, proved to be a limitation that would not let some of the other factors be examined for their inherent derogative potential.

Because of the flexibility in programming it was possible to examine the effects of a change in the magnitude of the irregularities in the path structure. Accordingly, a 25% out-of-tolerance condition in magnitude was produced and several pilots flew this condition.

Test Path 1 provided a uniform spatial frequency. The period was 4800 feet. The magnitude is at the Category II limits in ILS Zone 2.

Test Path 2 provides a slowly decreasing spatial frequency. The period extends from approximately 2500 feet at 3 1/2 mile range from threshold to 4800 feet at 3/4 mile. Further, approaching the threshold the path is made to move to progressive lower elevation values and continue so as to induce a pitch-down condition with the aircraft at the threshold.

Test Path 3 provides a very slowly changing path with a period of nearly 4 miles with the final delivery to the threshold low and with a trend to pitch down.

In all of the cases the magnitudes were set at Category II limits•

DATA COLLECTION.

The Pan American flight simulator is programmed routinely to produce straight-line glide slopes. For this work it was necessary to reprogram the desired glide slope structure. This was done by Pan American personnel during periods in December, 1983, and January, 1984, when the simulator was inactive.

Between January 23 - 27, 1984, approximately SO approaches were flown by various Pan American pilots. Most of these approaches were flown with a special dedication to this project work. Subject pilots ranged from the chief pilot to a first officer on the line. Parameters of interest were:

Character of structure roughness

Period of structure roughness

-114-

Magnitude of structure roughness

Weather minimums (VMC to 100 feet ceiling 1200 feet RVR)

Flight with and without aid of flight director

Runway lighting - with and without centerline and touchdown zone

Flight in smooth air and with moderate turbulence

The Pan American Simulator facility allowed for nine variables to be recorded in analog strip chart form. Selected for this project and acquired for each of the approaches were:

Bank angle

Pitch Angle

Lineal cross track error from localizer

Range from touchdown

Altitude

Airspeed

Heading

Glide Slope deviation in microamperes

Point where aircraft weight applied to wheels

Of particular interest were the glide slope and localizer variations of the flight track and where these variations occurred. Also, missed approaches were recorded.

The author rode on each of the simulated flights and was able to the make subjective evaluations and obtain comments from each of the subject pilots.

The procedure was to set the initial point of the flight approximately 2 miles outside the outer marker at 1500 feet above the field elevation. Approaches were made using Runway 1R at Dulles International or 9L at Miami International. Miami did not have the CAT II lighting. A standard instru­~ent approach was begun. The pilot was informed that the weather was at Category II minimums when those conditions were to be flown. The instruc­tions were to fly this approach as if it were being flown in routine line operation. Note should be made that Pan American practice is not to fly a Category II approach unless the couplers are available and to not go below 300 feet AGL if the flight director is not operative. In some of these tests these conditions were violated for purposes of collecting data.

-115-

The pilots were not informed of the details of the tests. Some suspected the tests involved wind shear. They were told only that it was a possible glide slope they might have to fly in practice and to do their best with the conditions at hand.

Initially, the subject pilots were given the three test cases of the glide slope in succession, first with the use of the flight director followed by the same three with only raw data available. Experience soon showed that Case 3 was not producing any unusual flight experiences, and this case was deleted from the repertory for the last one-half of the testing.

DATA ANALYSIS AND DISCUSSION.

The decision was made to review the 50 simulator flight records and extract significant items in the form of statistical data. The intent was to pro­vide the results of the simulator work in a more comprehensible form.

The maximum glide slope deviations were considered most important and how they related to range and whether the flight director was being used or not. The range where the maximum deviation occurred is believed signi­ficant and this information was extracted from the strip charts. In addi­tion, maximum heading deviation, maximum pitch excursion, maximum bank, and the touchdown point on the runway were extracted as items of interest, also. The compilation of the values is shown in appendix N.

The pilot variable was not considered. Even though individual differences in performance are recognized to exist, all pilots were type-rated and qualified in the Boeing 727-200 aircraft and were considered competent by the FAA and Pan American to fly the airplane. Both training captains, who spend most of their time in simulators, and line captains served as subject pilots.

One observation made early during the simulator work is that pilots are used to the straight-line glide slope. Knowing in advance that glide slope is straight allows the pilot of a heavy aircraft to use his vertical speed indicator as a principal indicator for descent adjustments to stay on the glide slope. Once established, the aircraft can be expected to remain on the glide slope. The glide slopes provided in this work were obviously not straight and this provoked reactions ranging from annoyance to mild hosti­lity. More than one pilot questioned the operation of the flight director •. The flight director tended to cause the pilot to overcontrol. The recor­dings made of the autopilot flight responses clearly indicated the the automatic systems would not do as well as a pilot who was conditioned to the fact that a non-ideal path existed. It was clear, also, that with repeated flights a pilot would do better and the comment by the training captains was that pilots could be trained to accommodate the rougher path structures but that was not being done now, obviously, with only straight­line paths available in the routine training. From comments it is apparent that pilots regard the real-world ILS paths essentially straight-line and consider the simulation routinely provided as representative.

Pilots would revert to raw data and seemed to prefer to process the rough course structure themselves. Considerable dependence was placed on sink

-116-

rate and they would perform averaging of the short period roughness using thts vertical rate information. With heavy aircraft this is appropriate, but with lighter aircraft more subject to small scale turbulence this is not always a reliable means. The pilots in the simulator learned or suspected that the path would return to a normal value and some accom­modated the roughness by simply waiting and making no changes to the aircraft descent rate. With the particular paths offered, this turned out to be a good technique.

The rough glide slope increased the workload of the pilots as might be expected. A consequence of this was the pilots did not do as well tracking the localizer even though it was a straight-line. This was particularly true when the flight director was not available. In fact, the missed approaches were almost without exception due to poor track on the localizer at the threshold.

The addition of moderate turbulence seemed only to add work load in contrast to the expected flight path deviations that might have occurred.

The radar altimeter read approximately 60 feet over the threshold when on the glide slope. The path was not programmed to be hyperbolic and this extra height could not have been accounted for by this geometry. It could not be determined why the radar altimeter which should have read a wheel height of approximat~ly 32 feet was so much in excess.

Below path clearance was checked with Test Case 3 and the raw data provided but 1/2 scale fly up command when a deliberate landing was made on the approach lights. Softening or desensitizing of the path indication by a factor approaching 6 undoubtedly accounted for much of this condition.

The impression report by several of the pilots,was that the flight director was overly sensitive and caused over controlling. The commands were very active as the portions of path roughness were flown. Pan American training emphasizes that the pilot should stay with the flight director commands which, in two of the cases, resulted in excess pitch motions of the aircraft.

The flights made on path structures whose magnitudes were 25% in excess of CAT II tolerances were slightly less desirable flight tracks with none becoming dangerous. Indications are that with training these paths with larger deviations could be accommodated. The pilots seem far more dissati­sified with the change from the straight-line to ~he paths at tolerance limits than they are moving 25% beyond.

Pilots are trained well enough not to let the aircraft get into an unsafe position or attitude unless false information is provided to them. In stressful situations attention is focused on the flight director and the information produced by this instrument is critical to. the safe operation. Probably the most dangerous situation is where a path would show no rough­ness but would simply lead to the wrong destination without any clues to the contrary. Glide slopes by their very nature cannot be constructed to give false courses to destinations other than the airport. Slow bends, of

-117-

course, can exist, and the protection must come from both magnitude and rate-of-change (ftrst derivative) tolerance limits. Flights conducted with magnitudes and rates slightly out of present limits. did not cause unsafe conditions but they were noticable and produced concern in the cockpit as to what was going on.

Each simulated approach was recorded on an eight-channel analog strip chart recorder. Critical, discrete items were taken from these records by hand. A sample of strip-chart data is shown is appendix 0 with the associated legend shown in appendix P. From these data, statistical data were derived.

First, an autocorrelation was run and the result was that the pilot was acting as a low-pass filter and this was already known from experience.

A determination was next made as to what factors, if any, had a significant effect on the maximum deviation of each run. This involved analysis of variance to determine the factors such as perturbation type, flight mode (raw data vis-a-vis flight director), or pilot status (training vs. line captain) which had an effect on the dependent variable, flight track deviation.

Case three data was discarded because there were too few flights. The dependent variable track deviation was divided into deviation above and below path.

The analysis may be described by the mathematica!" model such as

max-dev (above or below) = a + b (case) + c (range) + d (case * range) + E

where E represents a random error and a, b, c, d are constants.

If the dependent variable, max deviation, is not influenced by the inde­pendent variables, the hypothesis that

b = c = d = 0

cannot be rejected.

For the most part, there is no significant interaction at the 10% level. However, when considering below path deviations there were two cases which deserve further consideration. The deviations were 150 to 160 microamperes and this is considered a signficant amount of below-path excursion.

Consideration was given, in effect, to the possibility of a combination of two factors acting to provided maximum deviation. In the two-way analysis of variance for the perturbation case versus type of approach (flight director vs. raw data) there is evidence that the two factors together have an effect on maximum deviation but not separately.

In the two-way analysis of variance for the type of approach versus type of pilot (training vs. line pilot) there is evidence that the type of approach

-118-

has an effect at distances in excess of 3500 feet and when the first case of perturbation exists.

All other interactions were not significant, e.g., case vs. type of pilot, case vs. type of approach, type of approach vs. type of pilot, and case vs. range.

RECOMMENDATIONS.

Several recommendations are made based on the experience gained in this work. These are:

1. Flight simulation should be used to get a more extensive sampling of effects on track deviation produced by various types of glide slope structure perturbation.

2. Performance by other aircraft type should be examined.

3. Simulation used in flight t.raining should not be limited to straight-lines as is presently done.

4. Future simulation studies should acquire records of sink rate.

5. Simulation should be used to study effects of path angles greater than 3 degrees. The application would be specifically to allow consideration of keeping image glide slopes on the air during deep ground plane snow covers.

6. Measurements of effects of downdrafts on tracking of rough paths should be made.

7. Parameters, in addition to angle and structure roughness, should be studied. Examples of parameters are symmetry, rates of change of path oscillations, path width, and transverse structure of the glide slope.

8. An investigation should be made of the various types of couplers and their responses to selected perturbations determined.

9. In addition to the glide slope, the localizer should be examined in a similar manner.

10. Conditions which would allow expansion of tolerances should be identified specifically. An example is that given additional training, pilots_can handle safely rougher path structures.

-119-

CHAPTER V

DESIGN AND DEVELOPMENT OF GROUND-BASED TECHNIQUES FOR VALIDATION OF GROUND FACILITY PERFORMANCE

GROUND-BASED TECHNIQUES FOR VALIDATION OF SIDEBAND REFERENCE GLIDE SLOPE.

SUMMARY AND CONCLUSIONS.

Analytical and experimental work has been completed involving the sideband reference glide slope facilities at Tamiami, Florida (TMB); Morgantown, West Virginia (MGW); Clarksburg, West Virginia (CKB); and Bluefield, West Virginia (BLF). The data clearly demonstrate that there are areas in the near-field of a SBR glide slope where an 8' high probe will provide a good, reliable indication of system phasing for ground-check procedures. Typical values range from 50 to 100 ~A CDI change for each 10° of dephasing in either the upper/lower antenna phase, or CSB/SBO phase.

The probes, connected to a PIR (Portable ILS Receiver), provide for an excellent indication of overall facility performance. In fact, as demon­strated in the data section of this report, the probe at BLF would have detected an anomalous condition in which the far-field angle was far out­of-tolerance while the conventional SBR monitors were happy. In the pre­sence of 12 inches of snow, the BLF path angle lowered to 2.57° while the monitor reached only 20% of alarm limit. The CDI at the probe location changed by 129 ~A as the path shifted back up with melting and subsiding snow, while the conventional monitor changed by only 12-16 ~A.

Mathematical modeling of the near-field area for the purpose of phasing probes can be considered only a qualified success. The GTD (Geometric Theory of Diffraction) glide slope model generally shows the trends of CDI values in good fashion, but does not accurately predict absolute DDM values. For instance, when measuring at a probe height of 8', the 0 DDM points move further from the antenna mast as distance from the runway increases. This is clearly evident in the calculated data, but the actual locations of the 0 DDM points were usually off by 25' to 50'. In actual practice, the model is useful to eliminate undesirable areas subject to large, rapid DDM changes over small areas or locations requiring special equipment such as tall towers, while the desirable points remaining were located experimentally.

Experience shows it is difficult to position a truck or even a hand-held probe on the exact location repeatedly. Best results are obtained with a wooden mast, the base of which is anchored to the ground in a semi­permanent manner.

Despite the fact that the near-field probe provides an excellent indication of overall facility performance, it is not feasible to attempt to validate monitor performance for dephasing of the upper antenna by means of ground measurements in lieu of airborne measurements. If readings can be taken at the probe point in conjunction with flight checking at a given site, moni-

-120-

tor performance can be evaluated by ground measurements on later occasions. The actual response of ground check points varied by about a 2:1 ratio from site to site. Thus, dephasing the upper antenna and measuring the near­field alone does not provide sufficient knowledge about far-field changes. A SBR system at TMB which was maladjusted and then returned by flight checking to as near original conditions as possible did not display good repeatability at the probe locations with earlier data. If, on the other hand, a reference reading is taken at the probe emplacement when airborne measurements are made, it is possible to reset antenna phase to within 1°.

It is feasible to perform system phasing of a sideband reference glide slope on the ground once the system has been initially airborne phased; subject to the fact that w~thin the near-field area tested, CSB/SBO phasing and upper/lower antenna phasing give interactive results. If the SBO phase is advanced 5° and a reading taken in the near-field, it is not possible to determine from the measurement location whether the CSB/SBO phase is at fault or upper/lower antennna phase. If CSB/SBO phase can be determined by an independent method, such as a vector voltmeter reference reading after a flight check, then a properly located probe will give an excellent indica­tion of the upper/lower antenna phase and vice-versa.

Data for both a normal SBR system and a dephased SBR system indicate that the best results can be obtained in an area 400' to 500' off centerline. Calculations were performed for probe heights of 2' to 12' within this area, and no anomalous points or discontinuities were discovered. Measurements tended to confirm this, so a probe height of 8' was adopted as a standard.

The position of a satisfactory probe location may be found in the following manner:

Locate a line parallel to the runway, but 50 to 100 feet further from centerline than the antenna mast. Proceed out along this line until at least 400' from the mast and then look for 0 DDM using a PIR and probe antenna on an 8' to 10' pole. The only equipment needed is a standard PIR, the pole, and an antenna. It is suggested that the pole be fastened semi-permanently to some sort of base, as any variation in position will show up giving DDM changes.

Misphasing of 1° in either the CSB/SBO or upper/lower antenna causes 8-10 ~A of change at the best probe locations in TMB. This number is somewhat less at other sites and other locations, but seldom dropped below 5-6 ~/degree of dephasing. Calculations show that a change of about 34° in the upper/lower antenna phase is required to effect a 0.20° change in far­field path angle. It requires approximately a 38° change in either upper/lower antenna phase, or CSB/SBO phase to cause a far-field path width change of 0.20°. This means that the probe locations will indicate a 170 to 380 ~A change from nominal values before the far-field goes out-of-tol­erance.

-121-

INTRODUCTION AND STATEMENT OF THE PROBLEM.

Conventional monitoring techniques for the SBR glide slope are not always satisfactory. Previous studies and current field experience indicate that unnecessary shutdowns of SBR facilities are caused by the monitoring system showing an out-of-tolerance condition which often is not reflected in the far-field [71]. Ground plane anomalies, such as snow, may give spurious responses at close-in monitoring points such as those employed in the SBR system [72, 73, 74]. Flight checking the facility is very expensive and may not be available much of the time. Hence, a ground-based method of making phase verification measurements and an independent check on the con­ventional monitoring system is very desirable. The purposes of this study are:

1. To determine the feasibility of performing system phasing of the sideband reference glide slope (SBR) on the ground, rather than by airborne means once the facility has initially been airborne phased.

2. To determine the feasibility of validating monitor performance for dephasing of the upper antenna by means of ground measurements in lieu of airborne measurements.

APPROACH.

The approach that was used in the process of obtaining a solution to this problem was to start with a universe of possible points at which a SBR glide slope could be probed, and then to reject those which did not meet certain criteria. Specifically, the test plan consisted of 7 phases:

1. The Ohio University, Tamiami, ILS test site was configured as a SBR system and flight checked to ensure nominal values.

2. Noting the terrain limitations at representative SBR locations, the determination was made that the area under investigation should be limited to about 1000 feet in front of the mast and 550 feet perpendicular to the runway. Since the goal of the project was to identify locations suitable for routine field measurement, elevation was arbitrarily limited to a 20-foot height above prevailing terrain to preclude needing special equipment.

3. Plots of normal SBR DDM data were prepared for all points within this boundary on a 50-foot grid. Each point was measured at heights ranging from two feet to 12 feet at 2-foot intervals, for a total of 1,320 points. This process was repeated with the upper antenna advanced and retarded by 30°. Points were rejected at which DDM values were so large as to make further increases poten­tially caused by system dephasing .not easily detectable. Accept­able remaining points were further reduced by eliminating all those in which advancing and retarding the upper antenna did not cause the DDM to move in opposite directions from normal values.

4. The measured field data was compared to the values predicted by the math model. Since the near-perfect terrain at the TMB site

-122-

produces a very high quality signal, the DDM values measured at various points on the ground should be as close to theoretical values as could be reasonably expected. This served the dual purpose of demonstrating the efficacy with which the modeling process could be counted on to identify good locations for field measurements and simultaneously identify more unacceptable loca­tions.

S. A comparison was made of predicted and calculated data for the remaining suitable locations, and areas in which the best results were obtained were identified for further tests. Among the criteria used in the determination was the smoothness of the DDM plots in surrounding areas, the sensitivity of the probe locations to 30° dephasing, and the degree to which repeated measurements gave the same results. These areas were measured again while both the upper/lower antenna phasing and the CSB/SBO phasing was varied over a +60° to a -60° range. Semi-permanent emplacements were left at the two best locations for stability testing.

6. A similar process was performed at the other three SBR sites, namely Clarksburg, West Virginia (CKB), Bluefield, West Virginia (BLF), and Morgantown, West Virginia (MGW). The major difference was that the position of the semi-permanent emplacements was dictated primar~ly by the results of the TMB study, and most of the ground measurements were taken only as a way of proofing the math models for each site.

7. Normal measurements at each of the four sites were repeated over a period of 3 months to determine stability.

DATA COLLECTION AND DISCUSSION.

A major portion of the work involved making ground measurements of SBR systems at each of the four sites. All ground references and terrain ele­vation measurements for the math modeling process were obtained with a Warren Knight Model WK 83 theodolite. All ground DDM data collection was accomplished using either the Ohio University Mini~Lab Mark III or a Port­able ILS Receiver (PIR), Model FA-8766. Calibration of this equipment was confirmed with a IFR 401-L signal generator traceable to the National Bureau of Standards. All airborne data collection was flown in a Beech­craft Model 36 aircraft equipped with the Mark III Mini-Lab, which was also calibrated with IFR 401-L.

A total of 864 DDM plots were prepared for this project, a small sampling of which is included in this report. Those included were chosen as being most representative of the rest.

The terrain at each of the 4 SBR sites was surveyed in SO' grid increments away from centerline, with an elevation reading taken every SO' along these lines. The survey extended at least as far as threshold. Likewise, the distance off centerline for survey measurements was extended as far as practical •. These elevation measurements along with precisely measured

-123-

antenna heights were used as the basis for all the calculations per-formed by the model. In order to facilitate describing the ground measure­ment process, figure 5-1 illustrates the following coordinate system:

1. Y is the distance in front of the glide slope mast, parallel to the runway, with 0 at the base of the mast and Y increasing towards threshold.

2. X is the distance from runway centerline, perpendicular to the runway, with 0 on the centerline and X increasing towards the glide slope mast.

3. Z is the height of the pro~e above the terrain.

Figure 5-1 also shows typical measurement locations. As was explained earlier in this report, the calculations of DDM were limited to an eleva­tion of 20 feet. This was the reasonable limit to the height of a mast that two people could easily drag about in the field.

Figure 5-2 shows the plot of DDM vs. Y distance with X held constant at 50' from runway centerline. The bottom trace is at an elevation of 2', while the top trace is taken at an elevation of 20'. The remaining plots repre­sent 2' height increments between 2' and 20'. Note that the calculations ~aken-at-a-h~±ght uf 10' and below are biased strongly into the 150 Hz. As X is increased to 250', as is shown in figure 5-3 (all other conditions are the same as in figure 5-2), the traces are beginning to shift towards the 90 Hz. When X is increased still further to 500', as in figure 5-4, all the traces cross the 0 ~A line. This trend is also apparent in other plots not included, and can be summarized by saying that a plot of DDM vs. Y distance will show the curves getting more symetrical about the 0 ~A line as distance from runway centerline increases. It should also be noted in figure 5-4 that the patterns begin breaking up at the higher heights when Y is closer than about 300' to the mast, and there appears to be no advantage to working with a height of 20' as opposed to a height of 10' as X approaches 500'. Figure 5-5 makes this summary more apparent. In figure 5-5, all calculations are based on an elevation of 10'. The bottom trace is taken on centerline and the succeeding traces are taken in 50' incre­ments of X with the top trace being 550' off centerline. Since having a normal system DDM of 0 ~A gives the maximum possible room for potential DDM changes in either direction, the patterns that are the farthest from cen­terline seem to be the.best.

With the previous information in mind, figures 5-6, 5-7, and 5-8 examine the effects of dephasing the upper antenna. In all three figures, the short dashes represent advancing the phase of the upper antenna with respect to the lower, the longer dashes represent retarding the relative phase of the upper antenna, and the solid line indicates the normal phasing. In figure 5-6, taken at X=50', dephasing the upper antenna in either direction has an identical effect on the CDI values. This is an undesirable situation, since it would not be possible to determine the direction of the dephasing from looking at the probe data. Figure 5-7 represents the data collected at X=200'. In this area, a 30° phase retard

-124-

has a quite noticeable effect on a normal SBR system, but a 30° advance is hardly detectable. However, the retard trace is tending to drop to the opposite side of the normal trace as X is increased, and this trend con­tinues as X is increased still further. Figure 5-8, measured at X=500', is an excellent demonstration of this. At any Y greater than 350', the advance and retard phase changes cause large, reasonably symmetrical changes in the normal SBR. A close ·examination of the point at which a normal system crosses the 0 ~A line (Y=410') shows that a 30° advance in upper antenna phase cause a +210 ~ change, and a 30° retard causes a -260 ~A change.

Since both the normal SBR data and the dephased data seem to indicate that better results can be obtained in the area 400'-550' off centerline, addi­tional data is concentrated in this area. First, a full set of data was calculated for X=400' and X=500'. For each value of X, calculations were performed with CSB/SBO phasing advanced and retarded 30°, with upper/lower antenna phasing advanced and retarded 30°, and for a normal system. The entire calculation was repeated for every height between 2' and 12' inclu­sive. There were no anomalous points in any of the curves, and no discon­tinuities of any kind with respect to elevation changes, so a probe height of 8' was adopted for the measurement process •

. Figure 5-9 is the plot of the calculations made at X=500' and Z=8'. The center trace is for a normal SBR system, the upper trace is for upper antenna retarded 30°, and the lower trace is for upper antenna advanced by 30°. Figure 5-10 is the measured data for the same locations. The pre­dicted data has a steeper slope than the measured data. Some of this can probably be explained away by noting that the equipment used in the measurements such as the PIR does not have a flat response out to 400 ~A.

This does not explain the tilt between the measured and predicted data at lower CDI values. This tilt is present at all sites, and no explanation can be made at this time, other than to note that, in some cases, antenna offset can be adjusted in the models to remove it. In any case, the curves generally match in most respects; at least to the point that the general trends are well indicated. This particular plot was prepared for several other regions of the near-field, but the 500' line and beyond is more sen­sitive to phase changes.

Figures 5-11 and 5-12 demonstrate similar behavior of the CDI curves when the CSB/SBO phase is adjusted. Figure 5-11 is the predicted data while figure 5-12 is the measured data. In each case, the upper trace represents an advance in the SBO phase and the lower trace a retard in the SBO phase. The middle trace in each is the normal system. X=500' in both figures, and Z=8'. There is little difference in the behavior of the SBR system as measured at the probe location between a CSB/SBO phase adjust and a upper/lower antenna phase adjust. In either case, the area in question appears quite sensitive to phasing changes, and the curves are smooth and well defined. This indicated that the precise probe location is probably not important as long as it is in this general area, so the 0 ~A point was chosen as being easy to locate and allowing for best resolution without saturating the detection equipment. Consequently, the probe was moved along the X=500' line until 0 ~A was measured. The measured 0 ~A point was

-125-

located at Y=332', while the predicted 0 ~A point was at Y=39S'. This discrepancy is due to the tilt between the predicted and measured data as explained earlier.

In order to collect as much data as possible, all the 0 ~A points between X=3SO' and X=SSO' were located. The probe was placed at each position in turn while the upper/lower antenna phase was adjusted from a 60° advance to a 60° retard. This process was repeated for CSB/SBO phase. The results are shown in figure S-13 for upper/lower antenna phase and in figure S-14 for CSB/SBO phase. The position at X=350' is the least stable, and this is clearly reflected in figure S-14 where it can be seen that the 0 ~A point has shifted slightly into the 1SO Hz as the other points were being measured. This pattern held true on later stability checks; points further from the runway are more stable.

In summary, the data collected at the TMB test facility demonstrate that for 0 ~A points located at X ) 400 andY ) 300, dephasing of either the CSB/SBO or the upper/lower antenna of a SBR glide slope causes an easily detectable shifts in DDM. These points are in the general area of their predicted positions.

Upon completion of the data collection effort at TMB, work began with the other SBR sites; namely MGW, CKB, and BLF. Each site was surveyed in order to insure complete, detailed terrain information for input to the math model calculation, and DDM measurements were made at each point on the SO' by SO' grid described earlier. These values were converted to CDI ~A

values and plotted. A minor problem in comparing data between SBR facili­ties arises from the fact that the TMB facility has an offset of 4SO', far greater than BLF, MGW, or CKB. Consequently, measurements of these sites were concentrated in areas determined by the relative mast position; i.e. from a line 50' on the runway side of the mast to a point 100' on the far side of the mast. Therefore, the plots of measured and calculated data for BLF, MGT and CKB are limited to X values of 200' to 3SO'.

The predicted data for a normal SBR system at MGW are presented in figure S-1S. The range of X is 200' to 3SO'. The measured data for the same points are shown in figure 5-16. In general, the math model performed poorly at this site. A comparison of the two figures shows that the measured data are shifted nearly 200' closer to the mast than predicted. Except for this shift, however,. the general shape and trends of the measured data reflect a reasonable likeness to the predicted values.

Following the procedure established at TMB, the 0 ~A point predicted for each of the four values of X was modeled with the upper/lower antenna phase advanced and retarded by 60°. The predicted response to these conditions is shown on figure S-17. The 0 ~A point along each value of X was measured under the same dephasing conditions. These data are plotted in figure S-18, and a comparison with figure S-17 shows very good correlation between the two. The predicted values for the CSB/SBO dephasing is presented in figure S-19. The corresponding measured data is shown in figure S-20. A comparison between predicted and measured values for these two sets of data indicates that the area in question has good response to SBR dephasing.

-126-

CKB results are similar to those of MGW. Figure 5-21 and figure 5-22 show the predicted and measured values, respectively, for a normal SBR system at CKB. The measured data are shifted away from the mast by an average of about 75'. This is somewhat better than MGW, but still not as good as de­sired. The predicted and measured data for the upper/lower antenna de­phased conditions is demonstrated by figures 5-23 and 5-24. For CKB, the correlation between these values is not as good as expected, but the measured plots still seem to indicate good response at the chosen locations for probes. Figure 5-25 shows the predicted values for CSB/SBO phase changes, but no measured data was collected for this condition.

Finally, the predicted and measured values for BLF are given in figures 5-26 and 5-27. BLF demonstrates better correlation between measured and calculated values than CKB and MGW. The measured data are shifted away from the· mast by only 30'-40', when compared to calculated values. The predicted data for upper/lower antenna dephasing are shown in figure 5-28 and measured vales for the same data is shown in figure 5-29. Four posi­tions for the probe were used in the calculations, but only two were imple­mented for measurements. Likewise, figure 5-30 shows the predicted results of CSB/SBO dephasing and figure 5-31 shows the measured results. As in earlier cases, the calculations do not agree as closely as desired with field measurements, but the measurements do demonstrate that the probe will definitely detect the phase errors in question.

After completing the dephasing measurements, the next item of concern involved repeatability and stability of the measurements. Experiments at the TMB site showed that repeatedly positioning the truck at exactly the same spot was difficult, and that the presence of the truck affected the readings in some locations. Consequently, most field measurements were made with a wooden mast. Figure 5-32 is a photograph of one.of the wooden masts employed in the measurement process. A 12' version was constructed for the TMB measurements. As the most likely positions were determined, the wooden mast was left in position and anchored to the ground. This removed the possibility of positioning error. This meant that any day-to­day changes in the data were either due-to changes in the SBR itself (expansion/con~raction of the cables, phasers etc.) or problems with measurement stability. It is not possible to determine completely the contribution of each. Repeated measurements over a 2 week period at TMB showed that the 4 positions farthest from the runway were the most stable; the maximum deviation recorded was only 5 ~A. The probe location nearest the runway was the worst, showing a deviation of 28 ~A over the same period, while the second nearest had a deviation of 18 ~A.

Since 4 of the 6 probe positions stayed very near 0 ~A, then the dif­ficulties at the other two positions were probably related to inherent dif­ficulty with measurements in those locations rather than a change in the SBR itself• Data such as that earlier shown in figure 5-4 demonstrated irregularities in the plots as Y decreased towards 300'. Since the 0 ~A points along the cuts nearest the runway are getting close to this 300' area, the conclusion can be made that the points further out are more desirable from a stability standpoint.

-127-

A final stability check at MGW, CKB, and BLF was conducted approximately 6 weeks after ground measurements were completed. The measurements at MGW repeated within 15 ~A. The measurements at CKB had changed by about 50 ~A from their original values due to changes in the transmitting equipment, but were stable at the new readings. One probe at BLF repeated exactly, while the other had shifted 44 ~ for unknown reasons.

One major area of concern left unexamined at this point was the effects of snow on the probe locations. On the night of February 2, 1984, BLF received approximately 12 inches of snow. After checking with BLF main­tenance personnel to determine that the SBR was operational, Ohio Uniyersity personnel departed for BLF the next morning with the intent of checking the probe to see how it had been affected by the heavy snow and

.measuring the far-field. The far-field angle was 2.57° with a path width of 0.79°. A check of the SBR monitors showed that the integral and width monitors were essentialy nominal while the path monitor registered about 12-16 ~A during this period. This was 20% of alarm values. Measurements made at the probe emplacements showed 0.100 DDM into the 90 Hz, or -86~A.

By the middle of the next day, much of the the snow had melted and settled to an average depth of 5 to 6 inches. A flight check under ~hese con­ditions yielded a path angle of 2.98° with a width of 0.60°. The SBR moni­tors were all nominal. A check of the probe location with the PIR gave 0.043 DDM into the 150 Hz, or +43 ~A. In summary, the probe changed by a total of 129 ~ as the SBR far-field driftea out of tolerance and back, while the conventional monitor changed only 16 ~A.

The collected data are shown in tabular form in table S-1.

RECOMMENDATIONS.

From information obtained from both analytical and experimental phases of the work effort, it is recommended that the probe be given a trial under actual field conditions to determine its usefulness to maintenance person­nel. Using procedures described in this report, suitable probe positions should be located at several additional SBR sites. Regular and periodic DDM readings should be made and recorded along with conventional SBR moni­tor readings. Special efforts should be made to correlate probe data with flight check information. If probe readings deviate significantly from conventional monitor readings, it is recommended that a flight check be conducted on the facility. This will serve to determine very quickly whether the probe will have any practical utility.

As a specialized application of this recommendation, one of the sites cho­sen should be BLF. This will allow further investigation of the lowered path angle in the presence of deep snow. It should be noted that collect­ing data on this problem is somewhat difficult, since the vagaries of the weather are a factor. The snowfall should be sudden and deep, so that the change in path angle is clearly delineated. The weather immediately after snowfall must be good enough for optical tracking, and additionally the phenomena is probably a function of terrain and may be a function of the type of snow· as well. It was somewhat of a coincidence that all these fac-

-128-

CAPABILITIES

Will detect phase changes of ~ 1° in either CSB/ SBO or upper/lower antenna phase.

CDI changes 170-380 ~A before far-field goes out of tolerance.

CDI moves in opposite directions from 0 ~A as phase of CSB/SBO or upper/lower antenna is advanced or retarded.

LIMITATIONS

Cannot determine whether CSB/SBO phase or upper/ lower antenna phase is incorrect.

The amount of change at a selected probe location for a given amount of dephasing varies by a 2:1 ratio from site to site.

Math modeling predictions do not pinpoint the best measurement locations.

EQUIPMENT

Requires only an antenna, a PIR, and an 8' pole. Pole must be accurately positioned and anchored in place.

BEST PROBE LOCATION

Readings are less stable towards runway, more stable in front of and beyond the mast.

No apparent advantage to measuring at heights above 8'.

More sensitive to dephasing as distance in front of mast increases, until high DDM values begin to saturate equipment.

Table 2-26. Data Summary.

-129-

tors occurred at BLF, and there is no assurance at all that they will re­occur in the near future. To increase the odds in favor of further investigation of this phenomena, an effort should be made to identify and install probes at other SBR sites that have truncated ground planes and that are subject to heavy snowfall. Since this condition may only persist for a few hours, it is additionally recommended that some of these sites be instrumented such that data from the probe and from the standard monitors can be acquired at a remote location. This will allow continuous data collection under adverse weather conditions and will pinpoint the most desirable times to conduct flight checking of the facility.

-130-

CHAPTER VI

INVESTIGATORS AND ACKNOWLEDGEMENTS

This final report is the consolidation of work performed and reported by the staff at the Avionics Engineering Center, Department of Electrical and Computer Engineering, College of Engineering and Technology, Ohio University, Athens, Ohio. Contributions are from all levels ranging from undergraduate interns to faculty and staff. A number of FAA personnel at headquarters, at regional offices, and in the field have contributed in a variety of ways.

The following is a list of principal contributors and others who served to bring this work to a successful completion.

Dr. Richard H. McFarland has served as project director. He has also served as pilot for data collection and principal data analyst on all tasks. He authored the material on "Evaluation of Glide Slope Performance of ILS Serving Runway 18R at Dallas-Fort Worth Regional Airport"; "Determination of Aircraft Delivery Heights at Threshold for Coupled ILS Approaches to Runway 18R at Dallas-Fort Worth Regional Airport" (co-authored); "Siting Evaluation with Respect to Reference Datum Heights for the Glide Slope Serving Runway 18R at the Dallas-Fort Worth Regional Airport" (co-authored); "Effects of Irregular Path-Forming Terrain on Glide Slope Reference Datum Heights" (co-authored); and "Investigation of Pontiac Glide Slope Performance." He also authored "Analysis of Structure Roughness of Wheeling, West Virginia Glide Slope" (co-authored); "Electrical Modification to Capture Effect Glide Slope Facility to Eliminate Requirement for Site Relocation"; "Results from Experimental Approach Investigating Improvement of Tri-City Glide Slope Performance to Category II Standards"; "Recommendations for Restoring the Parkersburg Glide Slope to Service"; and "Results of Optimization Work on Dekalb-Peachtree Waveguide Glide Slope." He conducted a seminar on the endfire glide slope system, and presented a lecture entitled "Relevant Issues Concerning Construction Practices as They May Affect Propagation of Electromagnetic Signals Used for Air Navigation."

Mr. Walter D. Phipps has served as technical editor for this final report. He has served as project engineer for the critical areas study and the ground-based performance validation work. He authored "Ground-Based Techniques for Validation of Sideband Reference Glide Slope Performance," and "Theoretical Investigation of Null Reference, Sideband Reference, and Capture Effect Glide Slope Signal Scattering for Critical Area Determination."

Mr. Larry D. Brady served as project engineer for the work done on reference datum heights (RDH). He authored "Analysis of Effects on Reference Datum Heights of a Proposed Taxiway Addition for Runway 9R at the Chicago O'Hare International Airport," and co-authored "Determination of Aircraft Delivery Heights at Threshold for Coupled ILS Approaches ·to Runway 18R at Dallas-Fort Worth Regional Airport"; "Siting Evaluation with Respect

-131-

to Reference Datum Heights for the Glide Slope Serving Runway 18R at the Dallas-Fort Worth Regional Airport"; and "Effects of Irregular Path-Forming Terrain on Glide Slope Reference Datum Heights." He also authored "Fortran Programs for Computation of Glide Slope Reference Datum Heights"; "Results of Flight Measurements Performed on Category II Glide Slope Serving Runway 22 at Tri-City Airport, Bristol, Tennessee"; and co-authored "Analysis of Structure Roughness of Wheeling, West Virginia Glide Slope."

Mr. Jeff Dennis authored "Proposed Methodology for Landing Systems Mathematical Model Validation" and "User's Manual for Landing Systems Mathematical Model Validation Procedure."

Mr. Joe D. Longworth authored "Investigation·of Anomalous Performance of the Lambert-St. Louis International Airport Runway 30R ILS Category II Glide Slope."

Mr. Sammy Puleo served as data collection specialist for most of the air­borne work. He also was responsible for calibrations and references for the measurement equipment.

Dr. Robert w. Lilley has coordinated the efforts involving the Ohio University IBM 4341 and IBM 4381 computers.

Report production was furnished primarily by Ms. Alicya Shade with assistance from Ms. Janine Muklewicz and Christine Wagner. Drafting assistance was provided by Mr. Edgar Espinoza.

FAA personnel in Headquarters and the Regions, have been most helpful and supportive in the conduct of this work. For this, the staff is very grate­ful.

-132-

CHAPTER VII

BIBLIOGRAPHY

[1] U.S. Standard Flight Inspection Manual, FAA Handbook OA P 8200.1, Department of Transportation, Federal Aviation Administration.

[2] Department of Transportation, Federal Aviation Administration, FAA Order 8240.47, titled, "Determination of Instrument Landing System (ILS) Glidepath Angle, Reference Datum Heights, and Ground Point of Intercept," dated May 10, 1983.

[3] Department of Transportation, Federal Aviation Administration, FAA Order 8260.34, titled, "Glide Slope Threshold Crossing Height Requirements," dated October 26, 1983.

[4] U.S. Standard Flight Inspection Manual, FAA Handbook OA P-8200.1.

[5] "Instrument Landing System Improvement program," Third Interim Report, Report No. SRDS RD FAA-RD-72-71, Contract FA69WA-2066, Avionics Research Group, Ohio University, Athens, OH., June 1972, PP• 3-13 •

[6] "Radio Theodolite Placement Criteria for Glide Path Measurement," Final Report, Report No. SRDS RD-69-4, Ohio University, Athens, OH., January 1969.

[7] International Standards and Recommended Practices, Aeronautical Communications Annex 10 to the Convention on International Civil Aviation, Third edition of Volume !-July 1972.

[8] Leister, Harry J., "A Procedure for RTT Position Improvement Using Linear Regression Analysis of Glide Slope Structure," Tracals Report 82/665-277, U.S. Air Force Communications Systems Command, Special Report, May 30, 1982.

[9] "Engineering and Technical Services to Improve Reliability and Maintainability of the Instrument Landing System (ILS) Components," Final Report, Report No. FAA-R-6750.3, Contract DOT-FA79WA-4340, Avionics Engineering Center, Ohio University, Athens, OH., December 1980, pp. 113-130.

[10] "Instrument Landing System Improvement Program," Third Interim Report, Report No. SRDS RD FAA-RD-72-71, Contract FA69WA-2066, Avionics Research Group, Ohio University, Athens, OH., June 1972, PP• 3-13.

[ 11] McFarland, R. H., "Improved Accommodation of Runway Pedestal by Application of Linear Regression to Glide Slope," Precis 32, Avionics Engineering Center, Ohio University, Athens, OH., November 18, 1984 •.

-133-

[12] "Engineering and Technical Services to Improve Reliability and Maintainability of the Instrument Landing System (ILS) Components," Final Report, Report No; FAA-R-6750.3, Contract DOT-FA79WA-4340, Avionics Engineering Center, Ohio University, Athens, OR., December 1980, pp. 113-130.

[13] McFarland, R. H., "Experimental Examination of the Shreveport, Runway-14 Category-II ILS Glide Slope," Report No. EER-52-1, FAA Contract DTFA07-81-P-01055, Avionics Engineering Center, Ohio University, Athens, OR., June 1981.

[14] "Instrument Landing System Improvement Program," Third Interim Report SRDS Report No. RD 72-71, FAA Contract FA69WA-2066, Avionics Research Group, Ohio University, Athens, Ohio; June, 1972.

[15] McFarland, R. H., "Measurements of the Butte, Montana Capture-Effect Glide Slope for Mathematic Model Validation," Technical Memorandum N-7, Avionics Engineering Center, Ohio University, Athens, Ohio; 1980.

[16] Howard, LouT., Flying the ILS, Lgoma Press, Amityville, NY; December, 1976.

[17] Mroz, Mark, "Glide Slope Facility Snow Data Compilation," Technical Memorandum N-1, Avionics Engineering Center, Ohio University, Athens, Ohio; September, 1977.

[18] McFarland, Richard H., and Battistelli, J.J., "Investigations to Provide Improved Glide Slope Performance During Periods of Ground-Plane Snow Cover," Vols. I and II, Report No. FAA-RD-74-69, Contract DOT-FA69WA-2066-MOD 17 (EER 5-17 AND 5-18), Avionics Engineering Center, Ohio University, Athens, Ohio; April, 1974.

[19] Mitchell, Larry H., and McFarland, Richard H., "The Performance of the Null-Reference Glide-Slope System in the Presence of Deep Snow 1975-1976," Report No. FAA-RD-77-24 (EER 29-1), Contract No. DOT FA76WA-3764, Avionics Engineering Center, Ohio University, Athens, Ohio; January, 1977.

[20] Rondini, R. A. and McFarland, R. H., "Experimental Validation of Boeing 747 ILS Signal Scattering Calculations for Critical Area Determination," Final Report, FAA-RD-74-57 (EER 18-1), Avionics Engineering Center, Ohio University, Athens, Ohio; January, 1974.

[21] Reiffer, ·D. R., "Instrument Landing System for CAT III Operation, Multipath Interference Effects Due to Ground Movement of a Boeing 747 Aircraft on Heathrow Airport, London," TELS 34/2/069; January, 1971.

[22] Gorman, James T., "The Effects of Scattering from a 747 Aircraft Fuselage on the Operation of the Glide Path Portion of an Instrument Landing System," Master's thesis (supported by Federal Aviation

-134-

Administration), Avionics Engineering Center, Ohio University, Athens, Ohio; March 20, 1971.

[23] Kwon, YoungS., "The Effects of Reflection from Boeing 747 on Image Glide-Path Systems," Master's thesis (supported by Federal Aviation Administration), Avionics Engineering Center, Ohio University, Athens, Ohio; June 10, 1972.

[24] Rondini, Robert A., "A Study of Diffraction of Electromagnetic Waves Around Large Stationary Aircraft and Its Effects on Instrument Landing System Guidance Signals," Dissertation (supported by Federal Aviation Administration), Avionics Engineering Center, Ohio University, Athens, Ohio; June, 1976.

[25] Longworth, Joe D., "Instrument Landing System Critical Area Studies: Phase I, Theoretical and Experimental Investigation of Boeing 747 Dual-Frequency Localizer Signal Scattering for CAT III Critical Area Determination," (EER 59-3), Avionics Engineering Center, Ohio University, Athens, Ohio; November, 1983.

[26] Longworth, Joe D., "Instrument Landing System Critical Area Studies: Phase II, Theoretical Investigation of Wide-Body Aircraft Single-Frequency 8- and 14-Element Localizer Signal Scattering for Category III Critical Area Determination," (EER 59-3), Avionics Engineering Center, Ohio University, Athens, Ohio; November, 1983.

[ 27] Phipps, Walter D., "Theoretical Investigation of Single-Frequency 8-Element Localizer Signal Scatterer for Critical Area Determination," Report No. DOT/FAA/PM-85/4, U.S. Department of Transportation, Federal Aviation Administration; May,- 1985.

[28] McFarland, R. H., "Instrument Landing Critical Area Studies: Phase III, Critical Area Determination for Null-Reference Glide Slope Considering General Aviation Aircraft," (EER 59-3) Avionics Engineering Center, Ohio University, Athens, Ohio; September, 1983.

[29] Op. cit. McFarland, pp. 131, 150.

[30] "Manual on Testing of Radio Navigation Aids," ICAO Document 8071, Vol. 1, p. 62A, (3rd Edition, January, 1974).

[31] Op. cit. Phipps, PP• 167.

[32] U.S. Standard Flight Inspection Manual, FAA Handbook 0 AP 8200.1, Section 217.2, Figure 217-1A (Change 32).

[33] Op. cit., Longworth, Phase I.

[34] Op. cit., Longworth, Phase II.

[35] Op. cit., McFarland.

-135-

[36] Op. cit., Rondini.

[37] Chin, G., et. a!., "Users Manual for ILSLOC: Simulation for Derogation Effects on the Localizer Portion of the Instrument Landing System," Report No. FAA-RD-73-76, Department of Transpor­tation, Transportation Systems Center, Cambridge, MA; August, 1973.

[38] Op. cit., McFarland, pp. 90, 91.

[39] Op. cit., Chin, et. al.

[40] "Determination of Instrument Landing System (ILS) Glidepath Angle, Reference Datum Heights, and Ground Point of Intercept," Department of Transportation, Federal Aviation Administration, FAA Order 8240.47, May 10, 1983.

[41] "Glide Slope Threshold Crossing Height Requirements," Department of Transportation, Federal Aviation Administration, FAA Order 8260.34, October 26, 1983.

[42] McFarland, Richard H., "Evaluation of Glide Slope Performance of ILS Serving Runway 18R at Dallas-F-ort Worth Regional Airport," Technical Memorandum G~2, FAA Contract DTFA01-83-C-20025, Avionics Engineering Center, Ohio University, Athens, Ohio; January, 1984.

[43] McFarland, Richard H., "Evaluation of Glide Slope Performance of ILS Serving Runway 18R at Dallas-Fort Worth Regional Airport," Technical Memorandum G-2, Contract DTFA01-83-C-20025, Avionics Engineering Center, Ohio University, Athens, Ohio; January, 1984.

[44] U.S. Standard Flight Inspection Manual, FAA Handbook 0 AP-8200.1, Section 217.

[45] "Glide Slope Threshold Crossing Height Requirements," Department of Transportation, Federal Aviation Administration, FAA Order 8260.34; October 26, 1983.

[46] Brady, Larry D. and Richard H. McFarland, "Determination of Aircraft Delivery Heights at Threshold for Coupled ILS Approaches to Runway 18R at Dallas-Fort Worth Regional Airport," Technical Memorandum G-5, Contract DTFA01-83-C-20025, Avionics Engineering Center, Ohio Univer­sity, Athens, Ohio; July, 1984.

[47] "Engineering and Technical Services to Improve Reliability and Maintainability of the Instrument Landing System (ILS) Components," FAA-R-6750.3/EER-50-1, Avionics Engineering Center, Ohio University, Athens, Ohio, pp. 38-113; December, 1980.

[48] McFarland, Richard H., "Electrical Modification to Capture Effect Glide Slope Facility to Eliminate Requirement for Site Relocation," Precis 33, Avionics Engineering Center, Ohio University, Athens, Ohio; November 19, 1984.

-136-

[49] "Determination of Instrument Landing System (ILS) Glidepath Angle, Reference Datum Heights, and Ground Point of Intercept," Department of Transportation, Federal Aviation Administration, FAA Order 8240.47; May 5, 1983.

[50] "Glide Slope Threshold Crossing Height Requirements," Department of Transportation, Federal Aviation Administration, FAA Order 8260.34; October 26, 1983.

[51] "Siting Criteria for Instrument Landing Systems," Department of Transportation, Federal Aviation Administration, FAA Order 6750.16B.

[52] "Glide Slope Threshold Crossing Height Requirements," Department of Transportation, Federal Aviation Administration, FAA Order 8260.34; October 26, 1983.

[53] Ungvichian, Vichate, David Hartwig, and Joe Longworth, "User's Manual for the Ohio University Geometric Theory of Diffraction Glide Slope Model," Technical Report, EER 47-7, Avionics Engineering Center, Ohio University, Athens, Ohio; February, 1982.

[54] Luebbers, R.J., L.H. Mitchell, v. Ungvichian, and M. Mroz, "The Ohio University ILS Modeling Center," Technical Memorandum S-46, Avionics Engineering Center, Ohio University, Athens, Ohio; November, 1977.

[55] Ungvichian, v., "UTD Terrain Reflection Model with Application to ILS Glide Slope," Ph.D. Dissertation, Department of Electrical Engineering, Ohio University, Athens, Ohio; June, 1981.

[56] McFarland, ~ichard H., "Radio Theodolite Placement for Glide Slope Path Measurements," Final Report, Order No. Wl-9-00987-1, Project No. 320-101-04N, SRDS No. RD-69-4; January, 1969.

[57] "Siting Criteria for Instrument Landing Systems," op. cit. p. 82.

[58] U.S. Standard Flight Inspection Manual, FAA Handbook 0 AP 8200.1, Section 217.

[59] McFarland, R.H. and L.D. Brady, "Analysis of Structure Roughness of Wheeling, West Virginia Glide Slope," Precis No. 30, Avionics Engineering Center, Ohio University, Athens, Ohio; September 18, 1984.

[60] McFarland, R.H., "Results of Removal of Trees from Glide Slope Critical Area," Precis No. 34, Avionics Engineering Center, Ohio University, Athens, Ohio; December 21, 1984.

[ 61] McFarland, R.H., "Electrical Modification to Capture Effe-ct Glide Slope Facility to Eliminate Requirement for Site Relocation," Precis No. 33, Avionics Engineering.Center, Ohio University, Athens, Ohio; November 19, 1984.

-137-

[62]

[63]

[64]

[65]

[66]

"Engineering and Technical Services to Improve Reliability and Maintainability of the Instrument Landing System (ILS) Components," FAA-R-6750.3/EER 50-1, Avionics Engineering Center, Ohio University, Athens, Ohio, p. 38-113; December, 1980.

"Siting Criteria for Instrument Landing Systems," FAA Document 6750.16A; August 19, 1973.

"Manual on Testing of Radio Navigation Aids," ICAO Document 8071, Vol. 1, 3rd Edition, p. 62A; January, 1974.

Hogg, R. v. and T. A. Tanis, "Probability and Statistical Inference," Macmillan Publishing Company, Inc., New York; 1977.

Ohio University, "Proposed Methodology for Landing Systems Mathema­tical Model Validation;" Report No. DOT/FAA/CT-TN85/33, Avionics Engineering Center, Ohio University, Athens, Ohio; June, 1985.

[67] Ungvichian, Vichate, David Hartwig and Joe Longworth, "User's Manual for the Ohio University Geometric Theory of Diffraction Glide Slope Model," OU/AEC/EER 47-7, Avionics Engineering Center, Ohio University, Athens, Ohio; February, 1982.

[68] Fink, D. G. (Ed.), "Electronic Engineer's Handbook," McGraw-Hill, New York, 1975, pp. 18-79.

[69] Dielectric Materials and Applications, Part V, T~chnology Press, M.I.T., Cambridge, Mass. 1954.

[70] U.S. Standard Flight Inspection Manual, FAA Handbook 0 AP 8200.1, Section 217.5.

[71] Chamberlin, Kent, "Capture-Effect and Sideband-Reference Glide-Slope Performance in the Presence of Deep Snow," FAA-R-6750.1, Avionics Engineering Center, Ohio University, Athens, Ohio; July, 1978.

[72] Croxford, Raymond, R. H. McFa:dand, "Seminar on the ILS Sideband­Reference Glide Slope," Avionics Engineering Center, Ohio University, Athens, Ohio; December, 1978.

[73] McFarland, Richard H., "Anomalous Snow Effect on the ILS Glide Slope," Avionics Engineering Center, Ohio University, Athens, Ohio; February, 1979.

[74] McFarland, Richard H., "Notable Anomalous Effect of Snow on an ILS Glide Slope," Precis No. 28, Avionics Engineering Center, Ohio University, Athens, Ohio; April, 1984.

-138-

CHAPTER VIII

APPENDICES

-139-

APPENDIX A

DATA FROM PONTIAC NEAR-FIELD MONITOR

-140-

NEAR-FIELD MONITOR

DATE TIME DATE TIME RECORDER MAXIMUM ROLL II STARTED STARTED ENDED ENDED SPEED DEVIATION

1 4-20-84 21:50Z 4-25-84 19:46Z 6" 15%

2 4-26-84 16:30Z 5-4-84 16:20Z 6" 15%

3 5-4-84 16:35Z 5-11-84 17:15Z 6" 15%

4 5-11-84 17:30Z 5-17-84 15:01Z 6" 15%

5 5-17-84 14:55Z 5-d8-84 19:33Z 6" 15%

6 5-21-84 11 :45Z 5-29-84 13:50Z 6" 25%

7 5-29-84 17:00Z 6-5-84 17:00Z 6" 15%

8 6-5-84 07:53Z 6-13-84 18:23Z 6" 15%

9 6-13-84 18:45Z 6-22-84 13:40Z 6" 10%

10 6-22-84 13:50Z 6-29-84 15:51Z 6" 10%

11 6-29-84 19:56Z 7-3-84 12:30Z 6" 10%

12 7-3-84 12:50Z 7-10-84 18:12Z 6" 10%

13 7-10-84 18:16Z 7-18-84 13:56Z 6" 10%

14 7-18-84 14:05Z 7-27-84 18:50Z 6" 10%

15 7-27-84 17:38Z 8-3-84 17:42Z 6" 10%

-141-

NEAR-FIELD MONITOR

DATE TIME DATE TIME RECORDER MAXIMUM ROLL II STARTED STARTED ENDED ENDED SPEED DEVIATION

16 8-4-84 17:50Z 8-10-84 17:05Z 6" 10%

17 8-10-84 17:20Z 8-17-84 18:48Z 6" 10%

18 8-17-84 18:55Z 8-24-84 15:46Z 6" 10%

19 8-24-84 15:56Z 8-31-84 18:55Z 6" 10%

20 8-31-84 19:00Z 9-7-84 16:20Z 6" 10%

21 9-7-84 16:25Z ·9-17-84 14:12Z 6" 15%

22 9-17-84 14:15Z 9-26-84 18:07Z 6" 10%

23 9-26-84 18:15Z 10-5-84 13:45Z 6" 10%

24 10-5-84 13:50Z 10-12-84 18:55Z 6" 10%

25 10-12-84 19:00Z 10-19-84 16:55Z 6" 10%

26 10-19-84 17:00Z 10-26-84 15:00Z 6" 05%

27 10-26-84 18:10Z 11-2-84 18:30Z 6" 05% -

28 11-2-84 18:30Z 11-9-84 19:10Z 6" 05%

29 11-9-84 19:15Z 11-16-84 19:10Z 6" 05%

30 11-16-84 19: 15Z 11-23-84 18:20Z 6" 10%

-142-

NEAR-FIELD MONITOR

DATE TIME DATE TIME RECORDER MAXIMUM ROLL II STARTED STARTED ENDED ENDED SPEED DEVIATION

31 11-23-84 19:10Z 11-30-84 18:40Z 6" 15% (1)

32 11-30-84 . 18:45Z 12-3-84 16:06Z 6" 05%

33 12-3-84 16:11Z 12-5-84 17:30Z 6" 08%

34 12-5-84 17:40Z 12-7-84 19:25Z 6" 05%

35 12-7-84 19:30Z 12-13-84 15:35Z 6" 10% (2)

36 12-13-84 15:35Z 12-21-84 18:40Z 6 ·~ (3)

37 12-21-84 18:45Z 12-28-84 18:37Z 6" 15% (4)

38 12-28-84 18:45Z 1-7-85 14:20Z 6" (5)

39 1-7-85 15:04Z 1-16-85 18:10Z 6" 10%

40 1-16-85 15:05Z 1-25-85 19:35Z 6" 25%

41 1-25-85 19:40Z 2-1-85 19:50Z 6" 10% (6)

42 2-1-85 20:00Z 2-8-85 20:30Z 6" 25%

43 2-8-85 20:36 2-15-85 21:35Z 6" 25%

44 2-15-85 21:38Z 2-25-85 23:28Z 6" 25%

45 2-25-85 23:35Z: 3-4-85 19:35Z 6" 25%

-143-

NEAR-FIELD MONITOR

DATE TIME DATE TIME ROLL II STARTED STARTED ENDED ENDED

46 3-4-85 19:40Z 3-13-85 13:20Z

47 3-13-85 13:25Z 3-22-85 18: 10Z

48 3-22-85 18:10Z 3-29-85 19:10Z

49 4-5-85 14:40Z 4-10-85 19:00Z

(1) GS OTS from 17:27Z 11-24-84 until 19:80Z 11-26-84. (2) Accident at 00:33Z 12-18-84. (3) 22:27Z 12-16-85/ okay except ••• ???? (4) 90AZ.

- (5) 19:20Z 12-28-85 pen dried up. No record taken. (6) 90AC. (7) Project completed.

-144-

RECORDER MAXIMUM SPEED DEVIATION

6" 25%

6" 25%

6" 25%

6" (7)

APPENDIX B

DATA FROM PONTIAC FAR-FIELD MONITOR

LOWER ANTENNA

-145-

FAR-FIELD MONITOR - LOWER ANTENNA

DATE TIME DATE TIME RECORDER MAXIMUM ROLL II STARTED STARTED ENDED ENDED SPEED DEVIATION

1 4-20-84 20:36Z 4-27-84 15:47Z 6" 25%

2 4-27-84 16:35Z 5-4-84 16:55Z 6" 25%

3 5-4-84 17:10Z 5-11-84 16:40Z 6" 25%

4 5-11-84 16:55Z 5-14-84 10:00Z 6" 25%

5 5-21-84 21 :OOZ 5-29-84 13:50Z 6" 25%

6 5-29-84 13:55Z 6-5-84 16:36Z 6" 25%

7 6-5-84 16:45Z 6-13-84 18:21Z 6" 25%

8 6-13-84 18:27Z 6-21-84 18:07Z 6" 25%

9 6-21-84 18: 15Z 6-29-84 19:10Z 6" 25%

10 6-29-84 19:21Z 7-3-84 13:08Z 6" 25%

11 7-3-84 13:20Z 7-9-84 15:00Z -6" 25%

12 7-9-84 15:22Z 7-20-84 16:50Z 6" 25%

13 7-20-84 17:00Z 7-27-84 18:25Z 6" 25%

14 7-27-84 18:40Z 8-3-84 17:10Z 6" 25%

15 8-3-84 17:10Z 8-10-84 17 :45Z 6" 0/K 25%

-146-

FAR-FIELD MONITOR - LOWER ANTENNA

DATE TIME DATE TIME RECORDER MAXIMUM ROLL II STARTED STARTED ENDED ENDED SPEED DEVIATION

16 8-10-84 8-10-84 6" 25%

17 8-17-84 18:25Z 8-24-84 19:00Z 6" 0/K 25%

18 8-24-84 19:22Z 8-31-84 19:20Z 6" 0/K 25%

19 8-31-84 9-7-84 16:50Z 6" 0/K 25%

20 9-7-84 17:10Z 9-17-84 14:29Z 6" 0/K 25%

21 9-17-84 14:48Z 9-25-84 19:46Z 6" 0/K 25%

22 9-25-84 20:00Z 10-01-84 13:55Z 25%

23 10-05-84 14:40Z 10-12-84 18:00Z 6" 25%

24 10-12-84 18:10Z 10-19-84 17:20Z 6" 25%

25 10-19-84 17:24Z 10-26-84 18:45Z 6" 25%

26 10-26-84 18:50Z 11-02-84 18:55Z 6" 25%

27 11-02-84 19:02Z 11-09-84 18:35Z 6" 25%

28 11-09-84 18:40Z 11-16-84 18 :45Z 6" 25%

29 11-16-84- 18:50Z 11-23-84 18:43Z 6" 20%

30 11-23-84 18:50Z 11-30-84 19:00Z 6" 20%

-147-

FAR-FIELD MONITOR - LOWER ANTENNA

DATE" TIME DATE TIME RECORDER MAXIMUM ROLL II STARTED STARTED ENDED ENDED SPEED DEVIATION

31 11-30-84 19:05Z 12-07-84 20:00Z 6" 25%

32 12-07-84 20: 10Z 12-13-84 17:00Z 6" (1)

33 12-13-84 17:20Z 12-21-84 18:20Z 6" 25%

34 12-21-84 18:25Z 12-28-84 17:54Z 6" 25%

35 12-28-84 18:00Z 1-08-85 11: 50Z 6" 25%

36 1-08-85 16:54Z 1-16-85 14:27Z 6" 25%

37 ·1-16-85 14:35Z 1-24-85 14:20Z 6" 25%

38 1-24-85 14:27Z 2-01-85 19:22Z 6" 25%

39 2-01-85 19:28Z 2-05-85 13:15Z 6" 25%

40 2-06-85 15:30Z 2-11-85 14:00Z 6" 25% (2)

41 2-15-85 15:33Z 2-22-85 18:26Z 6" 25%

42 2-22-85 18:31Z 3-01-85 19:35Z 6" 25%

43 3-01-85 19:40Z 3-08-85 17:45Z 6" 25%

44 3-08-85 17:45Z 3-15-85 19:00Z 6" 25%

45 3-15-85 19:15Z 3-21-85 15:40Z 6" 25% (3)

-148-

DATE ROLL II STARTED

46 3-21-85

47 3-29-85

48 4-05-85

(1) Aircraft accident (2) Pen ran out (3) Trip to Pontiac

FAR-FIELD MONITOR - LOWER ANTENNA

TIME DATE TIME RECORDER MAXIMUM STARTED ENDED ENDED SPEED DEVIATION

15:40Z 3-29-85 18:35Z 6" 25%

18:50Z 4-05-85 18:30Z 6" 25%

18:35Z 4-10-85 18:30Z 6" 25%

-149-

APPENDIX C

DATA FROM PONTIAC FAR-FIELD MONITOR

UPPER ANTENNA

-150-

FAR-FIELD MONITOR - UPPER ANTENNA

DATE TIME DATE TIME RECORDER MAXIMUM ROLL II STARTED STARTED ENDED ENDED SPEED DEVIATION

6" per hr.

1 4-20-8 03:00Z 4-27-84 15:47Z 6" 20% I 150

2 4-27-84 16:35Z 5-07-84 10: 16Z -6" 20% I 150

3 5-07-84 14:37Z 5-14-84 19:35Z 6" 15% I 150

4 5-14-84 19:42Z 5-18-84 19:00Z 6" 35% I 150

5 5-18-84 19:15Z 5-25-84 15:50Z 6" 25% I 150

6 5-25-84 15:58Z 6-01-84 19:48Z 6" 35% I 150

7 6-01-84 19:55Z 6-08-84 19:20Z 6" 20% I 150

8 6-08-84 19:28Z 6-15-84 18:05Z 6" 35% I 90

9 6-15-84 18:90Z 6-22-84 17:55Z 6" 25% I 150

10 6-22-84 18:10Z 6-29-84 19:31Z 6" 20% I 150

11 6-29-84 19:20Z 7-07-84 22:07Z 6"

12 7-07-84 22: 15Z 7-20-84 (1) 6" 28% I 150

13 7-20-84 17:00Z 7-27-84 18:15Z 6" 25% I 150

14 7-27-84 18:15Z 8-03-84 16:50Z 6" 10% I 90

15 8-03-84 17:00Z 8-10-84 17:35Z 6" 10% I 90

-151-

FAR-FIELD MONITOR - UPPER ANTENNA

DATE TIME DATE TIME RECORDER MAXIMUM ROLL II STARTED STARTED ENDED ENDED SPEED DEVIATION

16 8-10-84 17:45Z 8-17-84 18:17Z 6" 30% I 150

17 8-17-84 8-24-84 19:00Z 6" 28% I 150

18 8-24-84 19:10Z 8-31-84 19:15Z 6" 28% I 150

19 8-31-84 19:20Z 9-07-84 16:40Z 6" 35% I 150

20 9-07-84 16:40Z 9-17-84 14:29Z 6" 35% I 90

21 9-17-84 14:44Z 9-25-84 19:24Z 6" 20% I 90

22 9-25-84 19:46Z 10-04-84 13:35Z 6" 20% I 90

23 10-04-84 13:45Z 10-12-84 17:55Z 6" 25%

24 10-12-84 18:00Z 10-18-84 17:30Z 6" 25%

25 10-19-84 17:35Z 10-26-84 18:35Z 6" 25%

26 10-26-84 18:40Z 11-02-84 19:06Z 6" 25%

27 11-02-85 19:15Z 11-09-84 18:40Z 6" 25%

28 11-09-84 18:50Z 11-16-84 18:38Z 6" 25%

29 11-16-84 18:38Z 11-23-84 18:35Z 6" 25%

30 11-23-84 18:40Z 11-30...:.84 19:13Z 6" 10% (2)

-152-

FAR-FIELD MONITOR - UPPER ANTENNA

DATE TIME DATE TIME RECORDER MAXIMUM ROLL II STARTED STARTED ENDED ENDED SPEED DEVIATION

31 11-30-84 19:15Z 12-7-84 19:46Z 6" 25% I 150

32 12-7-84 19:55Z 12-13-84 17:05Z 6" 10% (3)

33 12-13-84 17:20Z 12-21-84 18 :.OOZ 6" 25% (4)

34 12-21-84 18:20Z 12-28-84 18:05Z 6" 25% (5)

35 12-28-84 18:15Z 1-8-85 16:56Z 6" 40% I 150

36 1-8-8.5 17:01Z 1-16-85 14:47Z 6" 35% I 150

37 1-16-85 14:52Z 1-24-85 14:35Z 6" 35% 1·150

38 1-24-85 14:43Z 2-1-85 19:12Z 6" 35%

39 2-1-85 19:20Z 2-6-85 15:20Z 6" 20% I 150

40 2-6-85 15:25Z 6" 25%

41 2-15-85 20:47Z 2-22-85 18:41Z 6" 25%

42 2-22-85 18:46Z 3-1-85 19:10Z - 6" 25%

43 3-1-85 19:15Z 3-8-85 17:35Z 6" 25%

44 3-8-85 17:40Z 3-15-85 19:05Z 6" 25%

45 3-15-85 19:12Z 3-21-85 15:45Z 6" 25%

-153-

FAR-FIELD MONITOR - UPPER ANTENNA

DATE TIME DATE TIME RECORDER ROLL II STARTED STARTED ENDED ENDED SPEED

46 3-21-85 15:4SZ 3-29-85 18:30Z 6"

47 3-29-85 18:35Z 4-5-85. 18:40Z 6"

48 4-5-85 18:45Zl 4-10-85 18:30Z 6"

(1) Paper supply ran out. (2) Results collected in two days due to the system being down. (3) Aircraft accident. (4) 8m to 15m. (5) 12-22-85 and 12-24-85 no pen, no record.

-154-

MAXIMUM DEVIATION

25%

25%

25%

APPENDIX D

DATA FROM PONTIAC NEAR-FIELD MONITOR

AT TIME OF CRASH ON DECEMBER 13, 1984

-155-

_Ill

t !

I -

- L~ ~;

-156-

-[ L:f.

if- I f-·, r- -

. . ':'", <1. : ~~~

' '11

H-rt :rr++

! tiW-

=- .[ll ! - -i- ! ~

-t~~

-r-1.

t-8~ - - . -- I - --- f~ 'i

FF f+

'

-+

APPENDIX E

ANALYZED DATA

-157-

FULL AUTOLAND DATA

Date Time Carrier Flight Tail No. A/C WCH ACH*

6-2 1525 DELTA 118 N36DY L1011 28.7 53.7 6-3 0751 DELTA 124 N309DL 737 35.8 50.8 6-3 0754 DELTA 036 N314DL 737 42.2 57.2 6-3 0758 DELTA 111 N315DL 737 38.2 53.2 6-3 0802 DELTA 864 N106DL 767 36.8 61.8 6-3 0808 DELTA 133 N309DL 737 42.0 57.0 6-6 1654 US AIR 069 N327UA 737 41.4 56.4 6-7 0758 DELTA 094 N318DL 737 38.0 53.0 6-7 0811 DELTA 124 N317DL 737 36.3 51.3 6-7 0823 DELTA 175 N702DA L1011 28.2 53.2 6-9 1340 AMER 488 N223AA DC-9 38.9 53.0 6-9 1344 AMER 106 N228AA DC-9 32.5 47.5 . 6-9 1500 DELTA 122 N727DA L1011 30.0 55.0 6-10 0805 DELTA 310 N520DA 727 37.3 57.3 6-10 1026 AMER 223 N205AA DC-9 38.9 53.9 6-10 1215 DELTA 064 N313DL 737 39.1 54.1 6-10 1222 AMER 504 N203AA DC-9 38.5 53.5 6-10 1223 DELTA 337 N534DA 727 33.9 53.9 6-10 1234 DELTA 048 N301DL 737 35.0 50.0 6-10 1502 DELTA 824 N115DA 767 32.2 57.2 6-10 1626 AMER 334 N233AA DC-9 41.2 56.2 6-10 1631 AMER 228 N109AA DC-10 23.1 48.1 6-10 1641 AMER 050 N125AA DC-10 32.2 57.2 6-12 0811 DELTA 036 N317DL 737 32.0 47.0 6-12 1453 DELTA 054 N312DL 737 42.8 57.8 6-14 1513 DELTA 118 N711DA L1011 24.6 49.6 6-14 1800 DELTA 180 N728DA L10ll 23.2 48.2 6-14 1803 DELTA 176 N307DL 737 31.1 46.1 6-14 1806 DELTA 866 N104DA 767 37.1 62.1 6-15 1021 AMER 333 N312AA 767 36.1 61.1 6-15 1027 AMER 123. N218AA DC-9 42.0 57.0 6-15 1501 DELTA 189 N304DL 737 34.6 49.6 6-15 1512 DELTA 192 N306DL 737 35.1 50.1

*NOTE: ACH designates glide slope antenna crossing heights.

-158-

50 FOOT DECOUPLING DATA

Date Time Carrier Flight Tail No. A/C WCH ACH

6-2 1407 AMER 338 N6820 727 30.2 50.2 6-2 1500 DELTA 026 N311DL 737 23.7 38.7 6-3 0841 AMER 381 N622AA DC-8 29.1 44.1 6-3 1003 Al.'\fER 127 N856AA 727 19.9 39.9 6-3 1204 AMER 418 N1971 727 33.7 53. 7. 6-4 1250 BRAN 702 N456BN 727 26.6 46.6 6-4 1259 BRAN 038 N463BN 727 26.2 46.2 6-4 1310 BRAN 006 N469BN 727 38.6 58.6 6-4 1406 WEST 140 N284WA 727 33.8 53.8 6-5 0738 BRAN 101 N406BN 727 37.9 57.9 6-5 0740 BRAN 002 N461BN 727 32.9 52.9 6-7 0804 AMER 493 N885AA 727 37.2 57.2 6-7 0821 US AIR 062 N981 V3 DC-9 34.0 49.0 6-7 0946 BRAN 081 N460BN 727 29.0 49.0 6-8 0825 DELTA 447 N109DL 767 33.1 58.1 6-9 0947 BRAN 060 N466BN 727 37.3 57.3 6-9 1004 BRAN 034 N472BN 727 35.0 ss.o 6-9 1011 BRAN 302 N4SOBN 727 30.4 50.4 6-9 1012 AMER 127 N878AA 727 38.5 58.5 6-9 1014 AMER 471 N846AA 727 30.6 50.6 6-9 1145 AMER 050 N708AA 727 33.9 53.9 6-10 0817 EAST 659 N885EE 727 36.2 56.2 6-10 1155 AMER 446 N307AA 767 31.2 56.2 6-10 1210 AMER 350 N861AA 727 38.7 58.7 6-10 1526 US AIR 022 N322AR 737 38.9 53.9 6-10 1549 BRAN 039 N455BN 727 40.4 60.4 6-10 1555 BRAN 008 N461BN 727 25.4 45.4 6-11 1035 AMER 123 N228AA DC-9 45.8 60.8 6-11 1049 BRAN 082 N456BN 727 33.3 53.3 6-11 1201 AMER 446 N302AA 767 29.7 54.7 6-14 1553 BRAN 105 N472BN 727 33.0 53.0 6-14 1747 AMER 476 N2.18AA DC-9 45.9 60.9 6-15 0811 DELTA 133 N370DL 737 36.5 51.5 6-15 0929 UNITED 337 N7266U 727 30.1 50.1 6-15 0942 BRAN 001 N464BN 727 35.5 ss.s 6-15 0944 BRAN 063 N449BN 727 40.5 60.5 6-15 1004 BRAN 052 N460BN 727 40.1 60.1 6-15 1012 AMER 209 N6331 727 37.8 57.8

-159-

100 FOOT DECOUPLING DATA

Date Time Carrier Flight Tail No. A/C WCH ACH

6-3 0749 DELTA 310 N4080A 727 42.0 62.0 6-3 1007 AMER 531 N897AA 727 45.3 55.3 6-3 1011 AMER 527 N6808 727 26.8 46.8 6-3 1012 AMER 629 N6802 727 39.1 59.1 6-3 1014 BRAN 191 N457BN 727 43.0 63.0 6-3 1059 UNITED 415 N7270U 727 28.6 48.6 6-3 1200 AMER 358 N6820 727 34.1 54.1 6-4 1024 AMER 527 N6827 727 27.3 47.3 6-4 1029 AMER 531 N6807 727 37.9 57.9 6-4 1156 AMER 378 N884AA 727 42.1 62.1 6-4 1159 AMER 418 N1986 727 28.7 48.7 6-4 1244 EAST 502 N930EA 727 38.8 58.8 6-4 1253 BRAN 202 N468BN 727 31.7 51.7 6-4 1254 BRAN 715 N4 71BN 727 43.4 63.4 6-4 1301 BRAN 104 N460BN 727 44.2 64.2 6-4 1309 AIR1 202 837N 727 30.4 50.4 6-4 1400 AMER 140 N6829 727 33.7 53.7 6-4 1408 AMER 506 N6832 727 25.0 45.0

--- ~-- -----6-4 1409 AMER 472 N6815 727 34.3 54.3 6-4 1512 AMER 103 N6842 727 38.6 58.6 6-5 0749 EMER 602 N66AF DC-9 42.1 57.1 6-6 1713 JET AM 212 N7795JA 727 39.1 59.1 6-7 0752 DELTA 240 N402DA 727 31.1 51.1 6-7 0756 AMER 203 N863AA 727 24.2 44.2 6-7 0825 REPUB 221 N1798U 727 33.4 53.4 6-7 0848 AMER 323 N1995 727 36.8 56.8 6-7 0858 AMER 197 N879AA 727 46.4 66.4 6-7 0934 UNITED 337 N7456U 727 41.1 61.1 6-7 0940 BRAN 103 N472BN 727 40.5 60.5 6-7 0942 BRAN 001 N464BN 727 29.6 49.6 6-7 0954 BRAN 031 N477BN 727 33.0 53.0 6-7 1006 AMER 095 N876AA 727 42.2 62.2 6-7 1009 BRAN 143 N463BN 727 31.7 51.7 6-9' 0959 BRAN 063 N471BN 727 35.4 55.4 6-9 1002 AMER 095 N6822 727 25.1 45.1 6-9 1005 AMER 413 N6835 727 34.1 54.1 6-9 1010 BRAN 203 N408BN 727 29.3 49.3 6-9 1016 BRAN 143 N465BN 727 32.0 52.0 6-9 1020 BRAN 091 N447BN 727 49.6 69.6 6-9 1125 UNITED 415 N7467U 727 37.9 57.9 6-9 1142 AMER 102 N1981 727 33.5 53.5 6-9 1155 AMER 682 N720AA 727 16.8 36.8 6-9 1349 AMER 338 N88'0AA 727 21.2 41.2 6-10 0808 AMER 262 N856AA 727 36.4 56.4 6-10 0850 AMER 280 N6819 727 44.6 64.6 6-10 0859 JET AM 201 N770JA DC-9 48.5 63.5 6-10 0942 BRAN 711 N472BN 727 43.5 63.5

-160-

100 FOOT DECOUPING DATA (CONT.)

Date Time Carrier Flight Tail No. A/C WCH ACH

6-10 0946 BRAN 103 N462BN 727 20.1 40.1 6-10 1228 BRAN 003 N452BN 727 30.0 50.0 6-10 1245 BRAN 202 N453BN 727 37.3 57.3 6-10 1456 DELTA 216 N523DA 727 39.6 59.6 6-10 1458 DELTA 026 N307DA 737 40.6 55.6 6-10 1546 BRAN 094 N453BN 727 38.7 58.7 6-10 - 1547 BRAN 110 N463BN 727 31.2 51.2 6-11 1033 UNITED 716 N7279U 727 31.0 51.0 6-11 1229 AMER 344 N721AA 727 22.8 42.8 6-11 1234 AMER 570 N861AA 727 30.3 50.3 6-12 0847 AMER 175 N398AA 727 28.2 48.2 6-12 1544 BRAN 094 N409BN 727 31.3 51.3 6-13 1355 AMER 122 N145AA DC-10 23.2 48.2 6-14 1438 UNITED 742 -N7002U 727 41.9 61.9 6-14 1449 BRAN 044 N446BN 727 21.5 41.5 6-14 1538 BRAN 039 N451BN 727 34.7 54.7 6-14 1603 BRAN 204 N465BN 727 18.1 38.1 6.-14 1632 AMER 334 N203AA DC-9 46.7 61.7 6-15 0801 AMER 656 N719AA 727 18.8 38.8 6-15 0903 PANAM 575 N368PA 727 37.6 57.6 6-15 0950 BRAN 302 N452BN 727 40.1 60.1 6-15 0957 BRAN 091 N446BN 727 21.3 41.2 6-15 1057 NW 571 727 42.8 62.8 6-15 1117 UNITED 716 N7465U 727 29.4 49.4 6-15 1526 AMER 237 N1973 727 29.4 49.4

-161-

APPENDIX F

LANDINGS TRACKED AT DFW

-162-

Date: 6/1/84

Time Carrier Tail No. A/C D/C WCH Comments

1348 AMER 1910A 727 26.1 13SO AMER N1934 727 26.1 13S2 AMER N1908 727 29.3 13S3 AMER N6908 727 44.5 13S6 AMER N6801 727 34.9 13S9 AMER N19S7 727 37.2 1401 AMER N1168A DC-10 16.1 1402 AMER N8S28A 727 14.6 1406 AMER N8S08A 727 11.9 1408 AMER N1982 727 3S.3 1410 AMER N6812 727 39.9 1412 UNITED N7011U 727 36.8 1418 REPUB 7211C DC-9 21.S

Date: 6/2/84

Time Carrier Tail No. A/C D/C WCH Comments

1038 UNITED N7058U 727 40.6 1047 BRAN N4269 727 31.0 1100 UNITED N7278U 727 36.2 1305 BRAN N4S43BN 727 34.1 130S BRAN N4S98BN . 727 27.2 1348 WEST 727 51.0 1401 UNITED N706SU 727 20.3 1402 AMER N682S 727 30.8 1404 AMER N1934 727 33.7 1404 AMER N866AA 727 32.9 140S .AJ.\fER N6815 727 37.9 1407 AMER N6820 727 so 30.2 1447 DELTA N720A L1011 30.4 1453 .DELTA N1090L 767 27.5 1SOO DELTA N311DL 737 so 23.7 1S10 DELTA N314DL 737 3S.2 1S20 DELTA N822E DC-8 26.6 1S2S DELTA N36DY LlOll AUTO 28.7 1S20 UNITED N7019U 727 19.9 1S21 .AMER N6802 727 28.S 1S48 BRAN N460BN 727 47.S 1S49 BRAN N4S4.BN 727 31.0 1SS3 BRAN N447BN 727 17.6 1554 BRAN N406BN 727 13.0 1S58 BRAN N457BN 727 26.0 1SS9 BRAN N408BN 727 30.2 1600 BRAN N463BN 727 39.2

-163-

Date: 6/3/84

·Time Carrier Tail No. A/C D/C WCH Comments

0749 DELTA N4080A 727 100 42.0 0751 DELTA N309DL 737 AUTO 35.8 0752 BRAN N446BN 727 22.2 0753 AMER N857AA 727 26.6 0754 DELTA N314DL 737 AUTO 42.2 0756 AMER N873AA 727 25.2 0758 DELTA N315DL 737 AUTO 38.2 0802 DELTA N106DA 767 AUTO 36.8 0808 DELTA N3090L 737 AUTO 36.8 0841 AMER N622AA DC-8 50 29.1 0841 AMER N141AA DC-10 34.6 0846 AMER N850AA 727 29.9 0848 AMER N6818 727 49.4 0849 AMER N6831 727 37.1 0853 AMER 727 26.4 0901 PAN AM N361PA 727 26.4 0925 UNITED N7269U 727 39.1 0931 BRAN N453BN 727 42.9 0933 BRAN N466BN 727 23.7 0938 BRAN N462BN 727 31.4 0942 BRAN N447BN 727 21.8 0952 BRAN N460BN 727 36.2 0953 BRAN N459BN 727 23.1 0955 BRAN N409BN 727 23.7 0956 BRAN N465BN 727 31.0 1003 AMER N856AA 727 19.9 1004 BRAN N448BN 727 18.0 1005 AMER N890AA 727 35.6 1006 BRAN 727 18.0 1007 AMER N897AA 727 100 45.3 1011 AMER N6808 727 100 26.8 1012 AMER N6802 727 100 39.1 1014 BRAN N457BN 727 100 43.0 1016 AMER N871AA 727 34.8 1018 AMER N208AA DC-8 43.7 1036 UNITED N7458U 727 50.4 1059 UNITED N7270U 727 100 28.6 1101 AMER N729AA 727 32.7 1107 DELTA N308DL 737 12.8 1113 REPUB 920RW DC-9 37.1 1129 US AIR N322 737 34.8 1200 AMER N6820 727 100 34.1 1201 MEX XXMES 727 36.8 1204 AMER N1971 727 50 33.7 1206 AMER N6810 727 38.1 1208 PAN AM N4732 727 21.2 12'10 AMER N6816 727 31.0 1211 AMER N710AA 727 30.6

-164-

1212 AMER N854AA 727 27.0 1217 DELTA N504DA 727 28.4 1219 AMER N6819 727 21.5 1220 UNITED N7632U 727 22.6 1223 DELTA N31SDL 737 43.0 1224 AMER N6837 727 40.1 1226 BRAN N463BN 727 34.1 1228 AMER N889AA 727 40.3 1232 DELTA N461DA 727 33.2 1234 DELTA N401DA 727 30.5 1451 BRAN N448BN 727 54.9 1455 AMER N896AA 727 28.0 1456 DELTA N483DA 727 24.9 1458 DELTA N308DL 727 31.4 1500 DELTA N487DA 727 20.5 1503 DELTA N7 41DA LlOll 33.0 1S06 DELTA N130DL DC-8 28.9 1S08 US AIR N322AL 737 31.1 lSll DELTA N310DL 737 S1.8

Date: 6/4/84

Time Carrier Tail No. A/C D/C WCH Comments- ---------

1024 AMER N6827 727 100 27.3 1029 AMER N6807 727 100 37.9 1030 US AIR N324AL 737 20.9 1049 NW N271US 727 43.9 1050 REPUB R957N 727 23.3 lOSS BRAN N452BN 727 13.0 llSO AMER N6816 727 28.3 11S6 AMER N884AA 727 100 42.1 11S9 AMER N1986 727 100 28.7 1203 AMER N6817 727 18.4 1206 AMER N6813 727 Sl.O 1207 PAN AM N4741 727 22.6 1209 DELTA N530DA 727 24.9 1210 AMER N6824 727 21.0 1213 AMER N876AA 727 27.4 1214 AMER N6809 727 19.9 121S DELTA N490DA 727 28.3 1218 DELTA N313DL 737 31.0 1220 AMER N867AA 727 34.1 1222 AMER N849AA 727 26.0 1224 DELTA N406DA 727 41.2 1227 MEX 727 13.8 1230 AMER N30SAA DC-9 37.7 1232 AMER N223AA DC-9 26.0 1244 EAST N930EA 727 100 38.8 12SO BRAN N456BN 727 so 26.6 1251 BRAN N454BN 727 27.6

-165-

1253 BRAN N468BN 727 100 31.7 1254 BRAN N471BN 727 100 43.4 1259 BRAN N463BN 727 so 26.2 1301 BRAN N460BN 727 100 44.2 1303 BRAN N467BN 727 26.4 1304 BRAN N451BN 727 33.3 1309 AIR1 837N 727 100 30.4 1310 BRAN N469BN 727 so 38.6 1350 AMER N879AA 727 20.7 1352 AMER N720AA 727 19.9 1356 AMER N227AA DC-9 32.4 1357 AMER N889AA 727 31.0 1400 AMER N6829 727 100 33.7 1402 AMER N863AA 727 32.2 1405 NW N749US 727 36.2 1406 WEST N284WA 727 so 33.8 1408 AMER N6832 727 100 25.0 1409 AMER N6815 727 100 34.3 1414 AMER N1972 727 19.3 1422 FRON N737SF 737 200 36.8 1441 BRAN N462BN 727 24.5 1443 BRAN N406BN 727 26.0 1445 DELTA N474DA 727 31.4 1448 DELTA N824 DC-8. 34.3 1452 DELTA N301DL 737 23.7 1453 DELTA N312DL 737 25.1 1454 DELTA N10SDA 767 30.3 1457 DELTA N413DA 727 23.7 1503 AMER N104AA DC-10 23.2 1505 DELTA N318DL 737 32.4 1508 DELTA N427DA 727 37.0 1S10 AMER N891AA 72t 29.5 1512 AMER N6842 727 100 38.6 1S13 UNITED N7008U 727 17.8 1S16 DELTA N73SDA L1011 1S.5 1S17 AMER N6818 727 37.2 1519 DELTA N727 L1011 16.3

-166-

Date: 6/5/83

Time Carrier Tail No. A/C D/C WCH Comments

0738 BRAN N406BN 727 50 37.9 0740 BRAN N46l.BN 727 50 32.9 0749 EMERALD N66AF DC-9 100 42.1 0937 AIR! N407BN 727 29.7 0938 BRAN N409BN 727 25.4 0941 BRAN N458BN 727 29.7 0950 UNITED N7274U 727 40.8 0952 BRAN N468BN 727 26.0 0955 BRAN N450BN 727 28.7 0956 BRAN N456BN 727 20.6 1000 BRAN N47IBN 727 32.7 1002 BRAN N472BN 727 38.9 1003 BRAN N448BN 727 37.7 1004 BRAN N466BN 727 17.8 1007 AMER N6815 727 29.3 1007 AMER Nll7AA DC-10 27.0 1011 AMER N870AA 727 30.0 1013 BRAN N455BN 727 16.2 1017 ~R N708AA 727 30.8 1021 AMER N302AA 767 27.3 1023 AMER N851AA 727 40.4 1029 AMER N227AA DC-9 400 40.0 1029 AMER N881AA 727 24.5 1035 AMER N6821 727 29.7 1042 REPUB N895 DC-9 400 41.2 1043 BRAN N469BN 727 14.1

Date: 6/6/84

Time Carrier Tail No. A/C D/C WCH Comments

1654 US AIR N327UA 737 AUTO 41.4 1713 JET Al-i N7795JA 727 100 39.1 1728 REPUB N963N DC-9 25.4 1730 PAN AM N361PA 727 40.9 1744 AMER N6823 727 200 36.1 1745 AMER_ N223AA DC-9 30.7 1746 BRAN N448BN 727 12.9 1748 AMER N205AA DC-9 42.1 1749 AMER N6816 727 39.0 1751 DELTA Nll2DL 767 39.8 1752 AMER N207AA DC-9 38.4 1755 AMER N889AA 727 30.4 1759 DELTA N310DA 737 22.9 1801 DELTA N469DA 727 30.7 1803 BRAN N446BN 727 24.0 1803 DELTA N505DA 727 18.5

-167-

Date: 6/7/84

Time Carrier Tail No. A/C D/C WCH Comments

0739 BRAN N446BN 727 28.6 0744 BRAN N408BN 727 26.S 0746 BRAN N457BN 727 36.9 0748 BRAN N463BN 727 2S.8 0749 BRAN N469BN 727 21.2 07S2 DELTA N402DA 727 100 31.1 0756 AMER N863AA 727 100 24.2 07S8 DELTA N318DL 737 AUTO 38.0 0800 AMER N1908 · 727 31.7 0801 AMER 727 16.9 0803 AMER 727 26.9 0804 AMER N88SAA 727 so 37.2 0808 DELTA N110DL 767 21.9 'OK' 0811 DELTA N317DL 737 AUTO 36.3 0813 AMER N847AA 727 26.9 0816 EMERALD N66AF DC-9 20.8 0818 DELTA N313DL 737 400 38.8 0821 US AIR N981V3 DC-9 so 34.0 0823 DELTA N702DA L1011 AUTO 28.2 082S REPUB N1798U 727 100 33.4 0828 AMER N9676 747 37.S 'OK' 0844 AMER N21SAA DC-9 43.8 0847 AMER N1982 727 32.3 0848 AMER N199S 727 100 36.8 0850 EAST N88722 727 24.2 08SS AMER N126AA DC-10 200 34.4 'Marginally acceptable' 08S8 AMER N879AA 727 100 46.4 0859 AMER N6822 727 1SO 40.7 0900 AMER N227AA DC-9 1SO 37.3 0934 UNITED N7456U 727 80 41.1 0936 AIR1 N407BN 727 200 34.0 0940 BRAN N472BN 727 100 40.5 0942 BRAN N464BN 727 100 29.6 0946 BRAN N460BN 727 so 29.0 09S2 BRAN N471BN 727 200 18.S 09S4 BRAN N477BN 727 100 33.0 09S9 BRAN N4S4BN 727 90 2S.4 1001 BRAN N406BN 727 23.S 'Broken autopilot' 1003 BRAN N4S6BN 727 2S.8 1006 AMER N876AA 727 100 42.2 1009 BRAN N463BN 727 100 31.7 1011 BRAN N458BN 727 30.6 1013 AMER N6839 727 300 3S.9 'Looked good' 1016 AMER N854AA 727 24.6 1019 BRAN N409BN 727 24.4 . 1303 BRAN N464BN 727 26.1

-168- .

Date: 6/8/84

Time Carrier Tail No. A/C D/C WCH Comments

0748 AMER N6817 727 43.7 0752 BRAN N456BN 727 14.9 075S DELTA N305DL 737 27.0 07S6 BRAN N462BN 727 27.0 0801 DELTA N498DA 727 27.4 0808 AMER N843AA 727 38.S 0812 AMER N844AA 727 25.8 0819 DELTA N312DL 737 33.S 0825 DELTA N109DL 767 so 33.1 0832 US AIR N995UJ DC-9 38.5 0834 AMER Nl39AA DC-10 26.2

Date: 6/9/84

Time Carrier Tail No. A/C D/C WCH Comments

0843 AMER N706AA 727 23.6 084S AIR CAN CAA6 727 38.9 0847 AMER N215AA DC-9 40.5 0848 AMER N887AA 727 38.3 08SO AMER N6842 727 22.0 08S1 AMER N128AA DC-10 30.1 0855 AMER N225AA DC-9 40.8 08S6 AMER N6812 727 3S.1 0900 AMER N967S 747 37.4 0903 AMER N109AA DC-10 33.S 0906 JET AM N778JA DC-9 39.3 0932 UNITED N7449U 727 26.4 0947 BRAN N466BN 727 40 37.3 0948 BRAN N463BN 727 13.8 09S4 BRAN N448BN 727 25.1 09S9 BRAN N471BN 727 100 3S.4 1002 AMER N6822 727 100 2S.1 1004 BRAN N472BN 727 so 3S.O 100S AMER N683S 727 90 34.1 1010 BRAN N408BN 727 100 29.3 'Fly left at 200 feet' 1011 BRAN N4SOBN 727 50 30.4 1012 AMER N878AA 727 so 38.5 1014 AMER N846AA 727 so 30.6 1016 BRAN N465BN 727 90 32.0 1020 BRAN N447BN 727 100 49.6 1022 AMER N707AA 727 200 33.2 1024 AMER Nll2AA DC-10 33.2 1026 AMER N216AA DC-9 32.8 1028 REPUB N915RW DC-9 30.3 1030 AMER N6808 727 17.8 1036 TRAAM N742TV 747 36.4 .

-169-

1045 BRAN N451BN 727 26.4 1057 UNITED N7284U 727. 39.3 1107 DELTA N736DL L1011 35.1 'Decoupled at 4 nm DME' 1112 DELTA N305DL 737 1300 31.2 'Erratic G/S' 1125 UNITED N7467U 727 80 37.9 1135 DELTA N128DL DC-9 200 24.0 1142 AMER N1981 727 100 33.5 1145 AMER N708AA 727 50 33.9 1153 AMER N6824 727 20.9 1155 AMER N720AA 727 100 16.8 1158 DELTA N505DA 727 18.8 'OK' 1333 AMER N1903 727 33.5 1337 AMER N869AA 727 25.5 1340 AMER N223AA DC-9 AUTO 38.9 1342 AMER N105AA DC-10 22.0 1344 AMER N228AA DC-9 AUTO 32.5 1345 AMER N6804 727 34.7 'Erratic G/S' 1346 AMER N1995 727 35.5 'Erratic G/S' 1349 AMER N880AA 727 100 21.2 1351 AMER N137AA DC-10 39.3 1354 AMER N1997 727 34.7 1356 AMER N848AA 727 25.9 1358 AMER N833AA 727 23.6 1411 UNITED N7006U 727 21.7 1412 REPUB N956N DC-9 37.0 1415 NW G55US 727 33.5 1443 BRAN N466BN 727 44.7 1444 MEX XA1EW 727 22.0 1446 DELTA N825E DC-8 150 28.6 1448 DELTA N495DA 727 36.0 1449 DELTA N308DL 737 38.2 1451 DELTA N320DL 737 25.5 1452 AMER N896AA 727 14.0 1454 DELTA N721DA L1011 23.6 1457 DELTA N104DL 767 33.5 1458 DELTA N544DA 727 36.0 1500 DELTA N727DA L1011 AUTO 30.0 1506 AMER N6815 727 32.8 1509 DELTA N471DA 727 33.5 1511 DELTA N315DL 737 33.5 1513 UNITED N7021U 727 30.5 1515 AMER N6820 727 30.1 1517 US AIR N317AU 737 21.3

Date: 6/10/84

Time Carrier Tail No. A/C D/C WCH Comments

0752 AMER N720AA 727 30.3 'Good' 0754 DELTA N302DL 737 35.1 0755 DELTA N497DA 727 33.2

-170-

07S8 AMER N867AA 727 40.5 'Satisfactory' 0801 DELTA N305DL 737 42.8 080S DELTA N520DA 727 AUTO 37.3 0806 TRANS AM N742TV 747 46.6 0808 AMER N8S6AA 727 60 36.4 0811 AMER N716AA 727 200 44.3 'Oscillations' * 0813 AMER N199S 727 1SO 47.8 'Oscillations' * 0817 EAST N88SEE 727 so 36.2 0823 REPUB N766RC DC-9 30.9 0837 AMER N232AA DC-9 44.1 0839 AMER N1977 727 33.2 0843 AMER N234AA DC-9 40.6 0847 AMER N707AA 727 "33.2 0849 AMER N104AA DC-10 4S.1 08SO AMER N6819 727 44.6 08S2 AMER N6828 727 33.2 08S4 AMER N125AA DC-10 3S.S 08S9 JET AM N779JA DC-9 100 48.S 0911 PAN AM N361PA 727 40.3 0925 UNITED N7253U 727 26.6 0927 BRAN N458BN 727 27.4 0931 BRAN N451BN 727 33.2 0942 BRAN N472BN 727 100 43.5 0946 BRAN N462BN 727 100 20.1 'Dip at 100 feet and

oscillations at outer marker 09S4 BRAN N464BN 727 41.6 1000 BRAN N459BN 727 so 36.6 1002 BRAN N468BN 727 so 40.4 1004 BRAN N471BN 727 75 37.0 100S BRAN N4SSBN 727 50 42.7 1007 AMER N701AA 727 100 37.4 1010 Al"iER N6820 727 32.2 'Autopilot out-of-order' 1012 BRAN N469BN 727 so 37.7 1014 AMER N6809 727 100 31.8 101S BRAN N449BN 727 100 38.3 1017 AMER N6808 727 S40 32.8 1019 AMER N6824 727 200 40.8 1020 EAST N8986E 727 150 S2.8 1022 AMER N872AA 727 150 48.9 1026 AMER N205AA DC-9 AUTO 38.9 1029 AMER N6834 727 1000 42.8 1030 US AIR N323AL 737 70 3S.7 1053 UNITED N7260U 727 200 42.4 1103 UNITED N7448U 727 80 34.3 1110 DELTA N307DL 737 22.8 1124 US AIR N322AU 737 25.5 1127 REPUB N489S DC-9 21.7 1143 DELTA N127DL DC-9 37.0

*NOTE: Aircraft parked in glide slope critical areas

-171-

1151 AMER N788AA 727 33.5 1155 AMER N307AA 767 50 31.2 1201 AMER N886AA 727 150 34.3 1202 AMER N869AA 727 17.8 1206 AMER N866AA 727 300 30.5 1208 Ai.'1ER N845AA 727 80 36.4 1210 AMER N861AA 727 50 38.7 1212 AMER N6825 727 300 28.2 1214 AMER N6814 727 200 26.6 1215 DELTA N313DL 737 AUTO 39.1 1217 AMER N870AA 727 400 29.9 'Oscillations' 1220 MEX XAMEJ 727 150 26.3 1222 AMER N203AA DC-9 AUTO 38.5 1223 DELTA N534DA 727 AUTO 33.9 1228 BRAN N452BN 727 100 30.0 1230 DELTA N468DA 727 15.5 1231 AMER N6838 727 200 46.6 'Erratic' 1234 DELTA N301DL 737 AUTO 35.0 1236 UNITED N7461U 727 23.6 1240 BRAN N448BN 727 27.0 1244 BRAN N467BN 727 27.8 1245 BRAN N453BN 7.27 100 37.3 1430 MEX XAMEL 727 36.2 1440 BRAN N460BN 727 22.8 1447 BRAN N469BN 727 28.4 1450 DELTA N712DA L1011 23.6 1454 DELTA N722DA L1011 13.0 1456 DELTA N523DA 727 100 39.6 1458 DELTA N307DA 737 100 40.6 1501 AMER N867AA 727 300 34.7 'Dip at 300 feet' 1502 DELTA N115DA 767 AUTO 32.2 1504 AMER N725AA 727 37.2 1506 DELTA N130DL DC-8 56.2 1509 DELTA N315DL 737 41.8 'Unacceptable' 1510 DELTA N320DL 737 51.6 1511 DELTA N535DA 727 27.8 1513 DELTA N538DA 727 35.5 1515 AMER N843AA 727 45.6 1517 DELTA N462DA 727 28.4 1526 US AIR N322AR 737 50 38.9 1529 AMER N719AA 727 36.4 1535 AMrRANS N7573A 707 36.2 1540 BRAN N470BN 727 27.0 1543 BRAN N446BN 727 200 31.2 1546 BRAN N453BN 727 100 38.7 1547 BRAN N463BN 727 100 31.2 1549 BRAN N45SBN 727 so 40.8 1551 NW N291US 727 700 26.4 'Oscillations' 1552 BRAN N448BN 727 20.3 1554 BRAN N447BN 727 1500 32.4 1555 BRAN N461BN 727 50 25.4 1557 BRAN N467BN 727 19.7

-172-

1608 JET AM N778JA DC-9 45.6 1610 MEX XAHON 727 35.7 1613 AMER N350AA 727 44.7 1614 AMER N6812 727 350 28.6 1616 AMER N713AA 727 36.2 1618 AMER N708AA 727 27.4 1620 AMER N6835 727 43.7 1622 AMER N1982 727 250 30.9 1624 BRAN N472BN 727 29.7 1626 AMER N233AA DC-9 AUTO 41.2 1631 AMER N109AA DC-10 AUTO 23.1 1635 JET AM ~779JA DC-9 39.7 1641 AMER N125AA DC-10 AUTO 32.2

Date: 6/11/84

Time Carrier Tail No. A/C D/C WCH Comments

0846 AMER N703AA 727 38.3 'Nice G/S' 0847 AMER N111AA DC-10 25.9 0850 AMER N130AA DC-10 33.6 0853 AMER N6802 727 35.9 'Fly down at 150 feet' 0858 AMER N207AA DC-9 21.7 0859 AMER N6841 727 100 24.7 0901 AMER N725AA 727 34.9 0905 PAN AM N363PA 727 43.5 'Fine coupled approach' 0932 UNITED N7257U 727 100 32.8 0933 AIR1 N836N 727 34.0 0935 BRAN N470BN 727 36.6 0937 BRAN N463BN 727 30.1 'Good' 0940 BRAN N452BN 727 42.0 'Good G/S' 0947 BRAN N455BN 727 40.3 0950 BRAN N443BN 727 34.0 0951 .BRAN N450BN 727 28.2 0955 BRAN N454BN 727 32.2 0958 BRAN N459BN 727 29.3 1003 BRAN N469BN 727 22.4 1005 AMER N874AA 727 32.2 'Good G/S' 1007 BRAN N453BN 727 13.2 1012 AMER N870AA 727 200 34.3 'Bumps at 300 feet' 1015 BRAN N447BN 727 39.5 'Fine' 1016 BRAN N458BN 727 16.7 1023 AMER N6834 727 36.3 1026 AMER N720AA 727 34.3 1028 AMER N223AA DC-9 31.3 1033 UNITED N7279U 727 100 31.0 1035 AMER N228AA DC-9 50 45.8 1038 AMER N859AA 727 34.3 1044 FRONT N7363F 737 20.7 1049 BRAN N456BN 727 50 33.3 1052 METRO N929ML DC-9 51.4

-173-

1105 UNITED N7453U 727 30.3 1120 REPUB N768RC DC-9 38.0 1125 DELTA N304DL 737 200 53.2 1126 AMER N6804 727 2000 24.4 1136 DELTA N723DA L1011 30.3 1201 AMER N302AA 767 50 29.7 1203 PAN AM N4741 727 33.0 1204 AMER N845AA 727 34.3 1206 AMER N1994 727 29.7 1209 AMER N6813 727 37.4 1211 AMER N855AA 727 25.1 'Autopilot problems' 1214 AMER N883AA 727 200 42.0 1217 DELTA N317DL 737 29.3 1220 MEX N552NA 727 31.6 1229 AMER N721AA 727 100 22.8 1230 UNITED N7272U 727 34.5 1232 DELTA N530DA 727 25 36.1 1234 AMER N861AA 727 100 30.3 1236 DELTA N735DA L1011 27.4 1239 BRAN N460BN 727 35.5

Date: 6/12/84

Time Carrier Tail No. A/C D/C WCH Comments

0745 BRAN N463BN 727 33.6 0750 DELTA N318DL 737 29.8 0754 AMER N1901 727 43.2 0756 DELTA N404DA 727 40.3 0757 DELTA N309DL 737 32.8 0759 BRAN N452BN 727 27.5 0801 AMER N6836 727 29.2 0802 AMER N6821 727 28.6 0803 AMER N884AA 727 31.3 0805 AMER N860AA 7.27 41.5 0811 DELTA N317DL 737 AUTO 32.0 0812 DELTA N498DA 727 32.3 0814 DELTA N472DA 727 14.2 0815 DELTA N101DA 767 36.7 0826 US AIR N935UJ DC-9 36.7 0828 REPUB Nl961N DC-9 23.2 0830 AMER N9675 747 56.1 0834 AMER N705AA 727 26.5 0841 AMER N723AA 727 40.7 0847 AMER N398AA 727 100 28.2 0849 AMER N870AA· 727 37.8 0850 AMER N1910 727 36.7 0852 AMER N113AA DC-10 42.6 0854 AMER N887AA 727 200 26.3 'Dive at 250 feet" 0856 AMER N241AA DC-9 27.1 0857 AMER N849AA 727 31.3

-174-

•. -

0858 AMER N210AA DC-9 30.7 0900 AMER N719AA 727 17.9 0906 PAN AM N361PA 727 33.8 0910 METRO N1069T DC-9 48.6 0930 TRANS AM N742TV 747 25.0 'Poor G/S' 0934 BRAN N468BN 727 36.5 0936 UNITED N7288U 727 32.3 0944 BRAN N464BN 727 27.1 0947 BRAN N469BN 727 45.1 1448 DELTA N316DL 737 500 39.4 1451 BRAN N408BN 727 29.4 'A little low' 1453 DELTA N312DL 737 AUTO 42.8 1456 BRAN N460BN 727 15.6 1457 AMER N6821 727 31.3 'Erratic' 1458 DELTA N1292L DC-9 42.8 1501 DELTA N320DL 737 21.3 1502 AMER N1992 727 39.2 'Were not coupled' 1503 DELTA N477DA 727 12.5 1506 DELTA N475DA 727 200 40.9 'It worked real well' 1508 AMER N101AA DC-10 39.0 'Hand flown' 1512 US AIR N317 AU 737 20.6 1513 DELTA N741DA L1011 34.4 'Spacing move. To 350

feet it was right on' 1524 DELTA --- Nffi 3Dk- 707 31.7 1526 DELTA N718DA L1011 22.9 1538 BRAN N455BN 727 31.7 1539 BRIT C-BEBM DC-10 33.6 1543 NW N290US 727 30.2 1544 BRAN N409BN 727 100 31.3 1545 BRAN N472BN 727 25.5 1546 JET AM N781JA DC-9 42.2 1549 BRAN N443BN 727 23.6 1550 BRAN N463BN 727 25.0 1555 AMER N705AA 727 35.9 1556 BRAN N451BN 727- 26.7 1559 BRAN N458BN 727 300 14.8 1606 BRAN N464BN 727 26.3 1608 AIR1 N836N 727 36.7 1614 AMER N867AA 727 30.9 1618 BRAN N446BN 727 37.1 1619 AMER N882AA 727 33.2 1620 AMER N873AA 727 800 35.0 1623 JET AM N1004L DC-9 39.9 1627 AMER N1980 727 28.6 1630 REPUB N774RC DC-9 38.4

Date: 6/13/84

Time Carrier Tail No. A/C D/C WCH Comments

0750 DELTA N308DL 737 34.4 0752 DELTA N421DA 727 24.4

-175-

0754 BRAN N451BN 727 27.9 0755 DELTA N311DL 737 25.6 1234 AMER N866AA 727 39.2 1237 DELTA N723DA L1011 34.6 1240 BRAN N460BN 727 21.3 1250 EMERALD N38641 DC~9 26.3 1350 AMER N6804 727 21.7 1352 AMER N1901 727 29.4 1355 AMER N145AA DC-10 100 23.2 1357 AMER N71.5AA 727 31.3 1448 MEX XAMEF 727 35.2 1452 DELTA N304DL 737 32.9 1454 DELTA N103DA 767 29.8 1456 DELTA N312DA 737 23.4 1501 DELTA N301DL 737 29.4 1503 AMER N6831 727 38.2 1504 AMER N896AA 727 21.7 1506 FRONT N739F 737 37.1 1507 DELTA N748DA 727 32.9 1509 DELTA N508DA 727 29.4 1510 AMER N1971 727 40.9 1511 DELTA N1303L DC-8 23.3 1513 DELTA N1734 L1011 21.0 1516 US AIR N312AU 737 37.5 1518 AMER N6839 727 47.8 1519 AMER N778AA 727 40.5 1521 BRAN N477BN 727 29.6 1523 DELTA N712DA L1011 31.9 1529 UNITED N7017U 727 31.7 1532 NW N263US 727 31.1

Date: 6/14/84

Time Carrier Tail No. A/C D/C WCH Comments

1432 MEX KAMEK 727 32.1 1434 BRAN N466BN 727 400 49.4 1438 UNITED N7002U 727 100 41.9 1447 DELTA N108DL 767 21.6 1449 BRAN N446BN 727 100 21.5 1451 DELTA N464DA 727 40.0 1453 DELTA N477DA 727 16.2 1455 DELTA N303DL 737 26.2 1457 DELTA N316DL 737 36.5 1458 DELTA N530DA 727 16.8 1459 DELTA N822E DC-8 17.7 1501 DELTA N304DL 737 31.9 1503 AMER N858AA 727 36.0 1505 US AIR N325AU 737 30.6 1506 DELTA N306DL 737 37.7 1508 AMER N6836 727 33.5

-176-

1511 UNITED N7081U 727 200 21.6 1513 DELTA N711DA L1011 AUTO 24.6 'Off C/L left' 1533 NW N251US 727 32.3 1538 BRAN N451BN 727 100 34.7 1545 BRAN N409BN 727 14.5 1550 BRAN N454BN 727 26.4 1553 BRAN N472BN 727 50 33.0 'Nice ILS' 1554 BRAN N465BN 727 24.1 1555 BRAN N406BN 727 35.4 1556 BRAN N471BN 727 27.9 1559 JET AM N778JA DC-9 28.7 1603 BRAN N465BN 727 100 18.1 1610 BRAN N453BN 727 16.4 1611 AMER N885AA 727 34.8 1613 AIR1 N407BN 727 35.2 1614 BRAN N458BN 727 28.1 1616 JET AM N779JA DC-9 55.9 1618 AMER N219AA DC-9 27.5 1620 AMER N6807 727 30.2 1622 AMER N1986 727 37.7 'Unsatisfactory' 1629 AMER N1908 727 43.4 'Unsatisfactory' 1632 AMER N203AA DC-9 100 46.7 1634 AMER N104AA DC-10 25.6 1644 US AIR N316AU 737 34.0 1658 REPUB N9330 DC-9 30.8 1705 DELTA N109DL 767 29.0 1740 AMER N210AA DC-9 38.6 1744 AMER N845AA 727 40.6 1745 AMER N208AA DC-9 150 32.1 'Good to 150 feet' 1747 AMER N218AA DC-9 50 45.9 1749 AMER N846AA 727 27.1 1750 AMER N878AA 727 17.2 1757 DELTA N469DA 727 40.9 1800 DELTA N728AA 11011 AUTO 23.2 'Satisfactory' 1803 DELTA N307DL 737 AUTO 31.1 'It worked real well' 1806 DELTA N104DA 767 AUTO 37.1 'Good ILS all the way' 1810 DELTA N531DL DC-9 32.3 1811 PAN AM N363PA 727 43.4 1812 DELTA N472DA 727 27.9 'Autopilot would not lock up' 1814 DELTA N478DA 727 35.4 1820 DELTA N318DL 737 32.3 1823 UNITED N7266U 727 21.6 1827 AMER N709AA 727 24.1

Date: 6/15/84

Time Carrier Tail No. A/C D/C WCH Comments

0734 BRAN N459BN 727 28.7 0749 DELTA N315DL 737 28.7 0751 EMERALD N930EA DC-9 31.5

-177-

0800 DELTA N317DL 737 22.7 0801 AMER N719AA 727 100 18.8 0804 AMER N881AA 727 30.6 0806 AMER N6813 727 41.9 'ILS worked fine" 0808 AMER N1978 727 33.6 0811 DELTA N370DL 737 50 36.5 'Had to disconnect auto land

due to oscillations' 0817 REPUB DC-9 36.5 0829 AMER N9676 747 39.8 0836 AMER N225AA DC-9 50.7 0843 AMER N221AA DC-9 35.0 0843 JET AM N781JA DC-9 48.6 0844 AMER N1972 727 28.9 0849 AMER N113AA 727 36.5 0851 AMER N882AA DC-10 28.7 0854 AMER N849AA 727 36.3 'Coupled was real good' 0903 PAN AM N368PA 727 100 37.6 'CAT II was fine' 0929 UNITED N7266U 727 50 30.1 0933 AIR1 N407BN 727 200 51.3 'A good, smooth approach

all the way down' 0942 BRAN N464BN 727 50 35.5 0944 BRAN N449BN 727 50 40.5 'Sure glad it was VFR' 0947 BRAN N451BN 727 150 42.1 'We got a dip at 150 feet' 0950 BRAN N452BN 727 100 40.1 0952 OZARK N9832 DC-9 26.2 0957 BRAN N446BN 727 100 21.3 0959 BRAN N463BN 727 150 40.7 'Oscillations and porpoising

at 150 feet' 1000 BRAN N447BN 727 150 35.2 1004 BRAN N460BN 727 50 40.1 1007 BRAN N462BN 727 26.5 1009 BRAN N448BN 727 200 21.8 'Fly left at 200 feet' 1011 AMER N853AA 727 200 49.0 'Fly left at 200 feet' 1012 AMER N6331 727 50 37.8 1014 AMER N6807 727 48.6 1015 BRAN N470BN 727 42.9 1016 AMER N6841 727 44.4 1021 AMER N312AA 767 AUTO 36.1 'Good all the way to the runwE 1027 AMER N218AA DC-9 AUTO 42.0 'It was good. Disconnected at

the second high speed.' ;

1030 BRAN N456BN 727 26.0 1031 REPUB N9344 DC-9 40.9 1048 BRAN N463BN 727 30.8 1057 NW N229US 727 100 42.8 'Pretty good' 1105 UNITED N7 453U 727 50.5 1115 DELTA N315DL 737 150 32.5 1117 UNITED N7465U 727 80 29.4 'Better than most' 1126 REPUB DC-9 19.5 1459 AMER N850AA 727 26.4 1501 DELTA N304DL 737 AUTO 34.6 1506 DELTA N317DL 737 27.5 1509 DELTA N464DA 727 19.1 'It took a dive close in'

-178-

1512 DELTA N306DL 737 AUTO 35.1 1514 AMER N343AA 727 41.3 1516 BRAN N449BN 727 50.5 'Flattens out at 100 feet' 1520 US AIR 737 27.5 1526 AMER N1973 727 100 29.4 'Satisfactory' 1528 UNITED N7002U 727 46.7 'Problem with autopilot' 1531 BRAN N465BN 727 32.9 1539 BRAN N463BN 727 250 23.7 'Porpoising' 1542 JET AM N779JA DC-9 44.8 1545 BRAN N454BN 727 24.4 1549 NW N263US 727 29.8 1550 BRAN N467BN 727 30.6 1552 BRAN N469BN 727 34.2 1553 BRAN N457BN 727 27.5 1555 WE STAIR N2215Y DC-8 29.4 1557 BRAN N471BN 727 24.8 1600 BRAN N450BN 727 35.8

-179-

APPENDIX G

MEASUREMENT SENSITIVITY ANALYSIS

-180-

This appendix presents a sensitivity analysis of the method used to measure the wheel crossing heights. It shows that the elevation angle is the more sensitive measurement.

The wheel crossing height can be written as

WCH = d taneEL + a

where a = h - 1.

The last term in equation (1) is a constant which is measured only once each time the theodolite is positioned. Thus, the only error induced by this term is that of the measurement of eyepiece height, which should be accurate to at least 0.01 feet.

The first term in equation (1) is dependent on two variables, namely d and

eEL. Choosing the Boeing 727 as an example, we can expect a wheel crossing height of 35 feet. This figure is based on an antenna crossing height of 55 feet and a glidepath-to-wheel height of 20 feet. With no loss in generality we can assume that the theodolite eyepiece is positioned level with the threshold, so that a = 0. Thus, the expected angle of elevation with respect to the theodolite is:

To determine the sensitivity of the measurements, assume a small change in the measured quantity. For the distance d, assuming a change of 1.0 foot,

WC~ = (1098.54 + 1.00)tan(1.825) = 35.03'

WCH_ = (1098.54 - 1.00)tan(1.825) = 34.97'

For the angular measurement assume an increment of 0.05 degrees. This leads to:

we~= (1098.54)tan(1.825 + 0.050) = 35.96'

WCH_ = (1098.54)tan(1.825- 0.050) 34.04'

These calculations show that the more critical measurement in the procedure is. that of the elevation angle.

-181-

APPENDIX H

EFFECTS OF ANTENNA LOCATION ON REFERENCE DATUM HEIGHTS

FOR AN IDEAL GLIDE SLOPE

-182-

The equation of the zero difference in depth of modulation (DDM) path for an ideal image glide slope operating above an ideal ground plane (refer to figure H-1) is

tan ep (1)

where

d = Offset of antenna mast from runway centerline

Sp = Path angle

Equation (1) is used to calculate the path angles with respect to the ori­gin for the distances to the samples required for the regression analyses. From these samples the reference datum heights are calculated based on the procedure in FAA Order 8240.47. The results of these calculations are summarized in figures H-2 through H-19 for path angles from 2.6 to 3.2 degrees. The graphs are plotted for antenna offset distances from cen­terline of 250', 450' and 650'.

-183-

I ..... 00 ~ I

· z = Jd2+--?-- 1 tanQP

z

--------

Figure H-1. Illustration of the zero DDM path formed in space by an ideal glide slope site.

y

I ...... 00 U1 I

0 0 0 . 0 CD I I I I I

., l . I I I I I 1

--,

I 0 0 0 .

- :...__ - f- T -I-I

I-- - -I

-1

0 r-

0 ,.....0 t-o lL . ..... a

co

J:O Do a:~

0 tn

0 0 0 . 0 ~

0

-

~ ! -I- -

I I

I

--t

I

I I

--1

L

I~

. I ~

~r?-t

650'

~ -~

I

~~~ ~ ;:;.::;..___

h f

'-- I T

I~ ~ ~

I

I I~ ~

r

_/

I

~~

I

.........:: I J 0 0

J ~~~ I I I I I I -l . 0

"\) ~

0 I STANCE AU~NG CENTERLINE (fl) lX 10

Figure H-2. RDH versus distance of antenna mast from threshold along centerline for varying distances of the mast off center­line for a path angle of 2.5 degrees.

110.000

0 0 0

0

0 CD

0 0 0

0

0

"

,_,.0 ._g LLo ....,o

co

I IO

...... oo

00 a:~ 0\ 0 I IJ)

0 0 0

0

0 ::I'

0 0 0

0

0

'1>

~

~ ~ ..........:

~~~ ~ ~ ~ 250- ~~

~ ~ ~

~ ~ ~ ~

v

~ /

~ ~ .,......-::

Wo 70o000 0 I STANCE ALllNG CENTERLINE (fTl (X 101 J

Figure H-3. RDH versus distance of antenna mast from threshold along centerline for varying distances of the mast off center­line for a path angle of 2.6 degrees.

~

OoOOO

0 0 0 . 0 co

0 0 0 . 0 r-

,.....o t-8 LL . .._,a

ID

I J:O t-' 00

oo -...1 a::~ I 0

lf)

0 0 0 . 0 ;j'

0 0 0 . 0

'1>

~ ~ ~

v

~ /

650-"\ ~ 450'- ~ 250-~ ~/

~ v /

v v ~

v ~ ~

~ ~. ~ !Y.Ooo 70.000 80.000 90.000 100.000 110.000 120.000

DISTANCE ALtjNG CENTERLINE (fTl (XlOll

Figure H-4. RDH versus distance of antenna mast from threshold along centerline for varying distances of the mast off center­line for a path angle of 2.7 degrees.

~ !;::/"

0.000

I ...... CXl CXl I

0 0 0 . 0 IX)

0 0 0

0 ,...

0 """"o 1-o LL· ...._.a

to

ID oo a:~

0 Ul

0 0 0 . 0 ';jA

0 0 0 . 0 ~

/

~ ~ /

~ ~ ./

~

•

~ ~ ~

~/"

~ ~

v

~ 650/

~~ ~ v

450 ... ~ 250-

~ v

~

'1io.ooo 70.000 80.000 90.000 100.000 110.000 120.0001 130.000 140.000 0 I STANCE All~NG CENTERLINE (fTl (X 10 l

Figure H-5. RDH versus distance of antenna mast from threshold along centerline for varying distances of the mast off center­line for a path angle of 2.8 degrees.

·,

I ....... 00 1.0 I

D D D . D 10

D D D . D ......

4 ~

7

~ ,.......o .-g LL . ,_,o

10

~ ~

~ 65a ~

v 45 1-b\ f\ ~ 25d-

ID oo a:~

D lll

~ ~ P'

~ P'

~ D D D . D ;:J'

~ I?" ~

~ ~ ~ D D D ~ ~

-- --

. D

'1io.ooo 70.000 80.000 90.000 100.000 110.000 120.0001 130.000 1~0.000 DISTANCE ALCJNG CENTERLINE (fll (XlO I

Figure H-6. RDH versus distance of antenna mast from threshold along centerline for varying distances of the mast off center­line for a path angle of 2.9 degrees.

I 1-' \0 0 I

0 0 0 . 0 ~-~

(I)

0 0 0 . 0 r-

~ I L-~ i ~

~ i

~ ~0

r-g LL . .....,o

10

l

~ ~ ~ 650, \ ~ ~ 450-h\ ~ 2.50- _l

IO oo a:~

0 Ul

~ ~ .L_

~ ~ ~ 0 0 0 . 0 ~ v

/ ::j<

~ v /

0 0 0 ~ . 0 -

"to.ooo 70.000 80.000 90.000 100.000 110.000 120.0001 130.000 11!0.000 0 I STANCE ALONG CENTEAL I NE (fTl (X 10 l

Figure H-7. RDH versus distance of antenna mast from threshold along centerline for varying distances of the mast off center­line for a path angle of 3.0 degrees.

I ...... \0 ...... I

0 0 0 . 0 CD

0 0 0 0 . 0 ,.....

? ~· /~ I

~ ~ ~ I

,.......o ._g lL . ._.a

ID

~ ~ ~ 650, ~ ~ v I

450, h\ 250 1\/ ID Do a:~

0 In

~ ~ I

/

~ v

~ 0 0 0 . 0

~ v

~ ::2'

~ ~ ~ 0 0 ~ 0 . 0 --

'Go.ooo 70.000 80.000 90.000 100.000 110.000 120.0001 130.000 0 I STANCE ALONG CENTERLINE (fTl (X 10 I

1ij0.000

Figure H-8. RDH versus distance of antenna mast from threshold along centerline for varying distances of the mast off center­line for a path angle of 3.1 degrees.

I ..... 10 N I

a a a . a ([)

--- - [:0 ~~

a a a . b r--

~ ~ ~

~ v /

,.....a 1-g lL• ._.a

ID

l0 ~ 650-

/

~ v

450-h\ ~ 250'-:r:a oa a:~

a t? ~ ~

1/)

~ ~ ~ a a a . a ~ v

/ :::1'

l0 ~ 0 v a a . a ----

"lio.ooo 70.000 80.000 90.000 100.000 110.000 120.0001 130.000 1110.000 0 I STANCE ALCJNG CENTEAL I NE (fTl (X 10 J

Figure H-9. RDH versus distance of antenna mast from threshold along centerline for varying distances of the mast off center­line for a path angle of 3.2 degrees.

I ...... \0 w I

0 0 0 . 0 (I)

0 0 0

0 I'

0 ~0

1-o I..L· '-Jo

10

IO oo a:~

0 Ill

0 0 0 . 0 :j'

0 0 0 . 0

~ ~

7

~

~ ~

///

~ ~ ~

~ ~ /

~ ~

~ 7

~ 650'-450,. I~ ~ ~ 2.50' \/

~ ~ /

~

•

'l>o. ooo 70.000 80.000 90.000 100.000 110.000 120.0001 130.000 1110.000 DISTANCE ALONG CENTERLINE (fTJ lXlO l

Figure H-10. RDH versus distance of antenna mast from threshold along centerline for varying distances of the mast off center­line for a path angle of 3.3 degrees.

0 0 0 . 0 CD

0 0 0 . 0 r--

0

J=g LL . ...... a

ID

I ..... :r:o \0 oo ~ 0 I a:.

a:o Ul

0 0 0 . 0 :::1'

0 0 0 . 0

1>

~

~ 0 ~ ./

650-

~ ~ ~:::: /

45o' ./ v 250

~ b/ v /

~ ~ v /

~ ~ v _/ ~

~ ~ [7

v ---:?' /

DISTANCE ALONG CENTERLINE lFTl lXlO

Figure H-11. ARDH versus distance of antenna mast from threshold along centerline for varying distances of the mast off center­line for a path angle of 2.5 degrees.

I

I

' ' '

~

1!0.000

0 0 0 . 0 CD

0 0 0 . 0 r--

,......o 1-g LL . ......,o

(0

I :co .... \0

Do U1 a:~ I a:o

Ul

0 0 0 . 0 :::1'

0 0 0 . 0

'1i

/ ~ ~ ~ ~ ~ /

650, L:: ~ ~ 450-h\' 25d- ~ v ~ ~

v v

~ ~ v

v _/

~ k vp / /

~ ~ v I

/ I

DISTANCE AL~NG CENTERLINE .(fTl (Xl0 1l l \

Figure H-12. ARDH versus distance of antenna mast f!rom threshold along centerline for varying distances of thle mast off center­line for a path angle of 2.6 degrees.

v ~

0.000

I ...... \0 0\ I

0 0 0 . 0 IX)

0 0 0 . 0 ,.....

~ ,__a 1-g LL• ....,a

CD ~ v ~ /

~ ~ ~ / ~ IO oo a:~ a: a

Ill

65q '-\ ~ ~ '/ 450-h\ 25d- i\/ ~

~ ~ / / 0 0 0 . 0 :J'

~ ;:::; / /

~ ;:::; /

./ 0 0 0 . 0

'1i ~ v /

O.ooo '70.000 80.000 90.000 100.000 110.000 12 DISTANCE AU~NG CENTERLINE (fTJ lXlOlJ

Figure H-13. ARDH versus distance of antenna mast from threshold along centerline for varying distances of the mast off center­line for a path angle of 2.7 degrees.

~ f-'

0.000

I t-' \0 -...! I

0 0 0 . 0 Q)

0 0 0

0 I'

0 ~a

1-o LL • ...... a

ID

:co oo a:~ a:o

Ill

0 0 0 . 0 ::1'

0 0 0 . 0

I I I I r -

. /

V (/1f;/ _, .---V/

/vv 650 / /~ 450' , v/ // 25d-~ \/ /~/ /~~v

v[/;V ///

/~~/ /~~v vv /1

'1io. ooo --- ---1.....-~~~~+~_LJ_l_j

70.000 80.000 90.000 100.000 110.000 120.000 1 130.000 1ij0.000 DISTANCE AL(jNG CENTERLINE (fll (XlO I

Figure H-14. ARDH versus distance of antenna mast from threshold along centerline for varying distances of the mast off center­line for a path angle of 2.8 degrees.

I ...... 1.0 00 I

0 0 0 . 0 -CD

0 0 0 . 0 ,......

....d v v / ·r-

k: yl/ v ,_.0 1-0

0 LL • ......,a

(D

v t;; ~/ 650/ k ~v v 450-h\ ~ 25&

IO Do a:C: a:o

l/)

A ~ ~ v

I I I v ~ rY"'

/ 0 0 0 . 0 ::1'

~ v 1/

/ ./

~ y l7

/ 0 0 0 ~

v 0 - -- ----·--

'1io. ooo 70.000 80.000 90.000 100.000 110.000 120.0001 130.000 11!0.000 0 I STANCE ALCJNG CENTERLINE (fTl (X 10 J

Figure H-15. ARDH versus distance of antenna mast from threshold along centerline for varying distances of the mast off center­line for a path angle of 2.9 degrees.

I ...... \0 \0 I

0 0 0 . 0 a>

~ 0 0 0 . 0 r-

~ I~ /

~ /v I

v ,......o 1-g LL• "--~0

(0

~ v / I~ 650,

~ v v 450-h\ v 250'-:J:O oo a:~ a:o

Ul ~ ~

v ~

~ :;::;- / L

0 0 0 . 0 ::I'

~ :::: / /

~ ~ v 0 0 0 v

1 - - --.

0

1io.ooo 70.000 80.000 90.000 100.000 110.000 120.000 1 130.000 lijO.OOO DISTANCE AL~NG CENTERLINE (fTJ (XlO I

Figure H-16. ARDH versus distance of antenna mast from threshold along centerline for varying distances of the mast off center­line for a path angle of 3.0 degrees.

I N 0 0 I

a a a . a CD v ~ a a a . a I'"

~ v v

v v v v

' v / a

~a

1-a LL· .....,a

ID

v ~ v / /

650 v ~ v 450' h\ ~ 250' / v J:a oa a:~ a: a

Ul K ~ v

/ V' v v ~ / /

a a a . a

v v t:/v / ~ /

=I' v vv ~

a a a VI

---

. a '1io.ooo 70.000 80.000 90.000 100.000 110.000 120.0001 130.000 11!0.000

0 I STANCE RL~NG CENTEAL I NE lFTl lX 10 J

Figure H-17. ARDH versus distance of antenna mast from threshold along centerline for varying distances of the mast off center­line for a path angle of 3.1 degrees.

I N 0 1-' I

0 0 0 . 0 ([)

0 0 0 . 0 r--

,......a 1-g LL• .....,a

ID

:co Do a:~ a: a

Ul

0 0 0 . 0 ~

0 0 0 . 0

V vv I~

Vl0V _/ __,~1/

vv:-v 650 '---- / /V 450' 1

, 1/1 __,...-v 25d h\ 1\/ /'/

/~~v v/v_/v~

/ //

__,.-v !0v _,/ //'V

~~-~v I~V1~ri-!-IT-~-+~+-~-L~

'1>o.ooo 70.000 80.000 90.000 100.000 ·110.000 120.0001 130.000 1110.000 0 I STANCE Rll'JNG CENTERLINE lFTl lX 10 l

Figure H-18. ARDH versus distance of antenna mast from threshold along centerline for varying distances of the mast off center­line for a path angle of 3.2 degrees.

I N 0 N I

0 0 0 . 0 m

0 0 0 . 0 ,.....

0 ,.....0 1-o LL • ....... o

CD

IO oo a:~ a:o

L/)

0 0 0 . 0 ::t'

0 0 0 . 0

v ~

~50, 50 250'

v / / v ~ v

/ / / v I~ ~/

-- ---

L

,/ v / v /~ v v. / / ,../

~ ~ v L/

v / v b\ 1\/ ~ /~ ~ v / v. lY'

"l>o.ooo 70.000 80.000 90.000 100.000 110.000 120.0001 130.000 D I ST RNCE RL(jNG CENTEAL I NE (fTl (X 10 J

Figure H-19. ARDH versus distance of antenna mast from threshold along centerline for varying distances of the mast off center­line for a path angle of 3.3 degrees.

L_

11!0.000

APPENDIX I

DETERMINATION OF THE EXACT EQUATION FOR THE GLIDEPATH

FOR UNIFORM, LATERAL TERRAIN SLOPES AND EFFECTS ON THE

REFERENCE DATUM HEIGHTS

-203-

To determine the effects of uniform, lateral terrain slopes on the reference datum heights, the exact equation for the glidepath in space is derived. The analysis is similar in form to previous work, but the simplifying assumption that a is small is abandoned so that the exact effects of lateral gradients on reference datum heights could be determined -and the accuracy of the approximate equation can be examined.

Referring to figure 2-251, the equation for the cone in its natural coor­dinate system (designated by primed coordinates) can be written as

(2)

where d is the distance the antennas are of~set from the runway centerline and ep is the glidepath angle. Using the rotation of axes transformation equations

x'= x cos a - z sin a

z~= z cos a + x sin a

the equation for the cone in the gravitational reference system of coor­dinates becomes

(3)

(4)

2 2 2 2 (z cos a+ x sin a) = {[(x cos a- z sin a)-d] + y } tan ep (5)

The equation of the glidepath is determined by making x = 0, which defines the intersection of the tilted conic with the localizer plane as

2 2 2 2-(z cos a ) = [ (-z sin a - d) + y ] tan ep (6)

Completing the square yields

z =

2 2 2 r (d + y ) tan ep

L 2 2 2 cos a - sin atan 6p

+ -2

d sin a tan ep

2 - 2 2 COS a - S1n atan 6p

2 d sin a tan 6p

+ -----------------------2 2 2 cos a - sin a tan ep (7)

-204-

which is the exact equation for the glidepath. Letting a go to zero gives

z =

which verifies the exact equation, since equation (8) is the ideal path equation for a flat earth site.

(8)

Figure I-1 shows a comparison of the exact equation with the approximate equation developed in 1969 by McFarland and Redlich. Figure I-2 shows the effects of uniform, laterally downsloping terrain on the reference datum heights for gradients of up to 5%. Clearly, the reference datum heights are very insensitive to small, lateral terrain·slopes.

-205-

-I-LL.

0 0 CD

0 0 II) .

-o 0

UJ:S: u z UJ a: UJc LL.C LL."! 1-4

0

c c N .

0 0 ...

300.000

Figure I-1. Difference between exact and approximate equations for zero DDM path formed by a null reference glide slope in the presence of uniform, laterally downsloping terrain.

-206-

0 0 0 •

II) II)

0 0 II)

•

"

·A"RDH

11)~------------------------~----------

-o t-o LL.o -. " II)

(f)

t­:I:o (!)o ..... II)

w· :r::~

0

RDH

gL-----------------------~---A~----------•

'" II)

0 0 II)

• N+--------+--------r-------~------~------~ ~.oo t.ooo 2.ooo 3.ooo q.ooo s.ooo

PERCENT SLfJPE

Figure I-2. Response of RDH and ARDH versus lateral terrain gradient.

-207-

APPENDIX J

BIAS ELIMINATION

-208-

The purpose of this appendix is to demonstrate the validity of the tech­nique used in eliminating a constant-bias occurring in the raw CDI trace or modeling trace.

The technique is simply to subtract the bias from the trace. The assump­tion is that the scattering fields are not changed drastically by making minor adjustments to the antenna positions. Hence, the resulting trace should exhibit effects almost identical to those created by simply subtracting a constant bias. To verify the assumption by measurement is not an easy task. As one knows, the airplane cannot exactly repeat the same path every time. The problem is overcome by using the computer model.

One site has been chosen for the verification. With reference to figure 2-280, the modeling trace has a -13 ~A bias. The antenna heights are 14.13, 28.26, and 42.39 feet for lower, middle, and upper antennas, respec­tively. Physically moving all antennas down to 13.85, 27.70, and 41.55 feet results in moving the calculated trace down about the 0 ~A indication, therefore, giving the trace zero bias. Figure J-1 shows the new calculated trace. One can note the similar characteristics of the traces shown in figures 2-280 and J-1. By using the -13 ~A bias to compensate for the off­set of the trace shown in figure 2-280, zero bias trace is obtained.

-209-

I N ...... 0 I

N J:

~

........ <C( ..3-

8 -I Date: 1 Oct 80 Run No.: 1-16A Facility: TMB CE

00 u

N J: 0 II')

8 ~--------~--------~---------.---------.-------·--.-------~ + 0 10,000 20,000 30,000 Distance from Antenna Most (Feet)

Figure J-1. Calculated CDI after adjusting antenna heights and shifted bias of figure 2-280 traces.

APPENDIX K

TERRAIN PLATE SELECTING PROCESS

-211-

The method presented in this appendix requires less work in selecting the data points and the compromise in accuracy is considered minimal for most users. An effective approach, based on the Physical Optics theory, to a first approximation is to take the terrain vertical variations perpen­dicular to the runway centerline. The PO theory suggests that the dominant contribution to the total reflected signal comes from the first Fresnel zone. The dimensions of the first Fresnel zone can be up to 4000 feet in length in the direction parallel to the runway with the width less than 150 feet. Therefore, the most important ground to be considered is the 75 feet on either side of a vertical plane including the antenna and the aircraft. The lateral terrain variation can be made by making a piecewise linear approximat~on to the ground profile along the vertical plane and setting up plates in the 2-D OUGTD model to have the same profile. As the aircraft moves, the vertical plane will move; therefore, the ground profile is changed.

If the terrain has a bowl shape such that strong reflections from areas other than along the vertical plane including the antenna and the aircraft occur, the accuracy of this approach is greatly diminished. An example of how to select points from a topographical map is demonstrated. Figure K-1 is a sketch of a topography of the terrain survey data. Before the selecting process starts, one must decide how many flat plates and how many elevation steps are sufficient to represent the ground roughness. In the example, four elevation points and eight plates are chosen. One may recognize that these numbers (4 and 8) are in matrix form. It should be pointed out that eight 2-D connecting plates can be formed only by nine edges which are parallel to the X-axis. There is no unique guideline on how to select these numbers; however, if fewer numbers of plates or eleva­tion points are used, the plate model is obviously not a good represen­tation of the terrain considered. Figure K-1 shows the x marks that designate the chosen points.

Perhaps the most appropriate method of selecting a sufficient number of reflecting plates, while ensuring that an appropriate number are available to.model the site, is to use the Rayleigh Criteria. This criteria may be used to determine whether or not a terrain variation will effect the signal.

RC = 4wrsina A

r = Standard Deviation of Roughness

a = Grazing Angle

A ~ 3' (For glide slope)

A good rule of thumb in the reflecting plate selection is to only include a terrain variation as a new plate if the Rayleigh Criteria value is> 0.1; however, the minimum plate length should never be less than A• This cri­teria should be utilized only about the center of the first Fresnel zone or approximately 1,500'.

-212-

I N ..... w I

Scale 1 em= 100 feet

· Runway 11.5)1--------

___... )f= -Jo--------5 )(-

Profile Une from Antenna to Projected Aircraft

---1 7 }lr-

II- -·-' "( Base of Antenna

Mast (Reference J Point)

6 ' 500

X Coordinate Input Data

r-5 ' 1000

Distance in Feet from the Reference Point

-9 -9 -3 -2

I! \ ~

1~00

Figure K-1. An example of site topography (2 sheets).

AI c

---1--f.!J - 2 :;-£

..... ~

I 2000

I N ...... .j:>. I

II c: :J ..&; u

l

I 2000

')(

I

2500

Projection of an

10 11 Aircraft on the

...._ - ::.-- Runway Center-

~ Uno Extended

. - f-·-

I

3000

Distance in Feef from the Reference Point

Profi le U ne from Antenna to Projected Aircraft

I 3500

Aircraft

~ I I I I

9.6

1

4000

The data file corresponding to the x marks is shown in figure K-2. The first line in the file is the control line ~ith format 215. The first and second values are the number of step elevations and the number of plates plus 1, respectively. The following lines are the X, Y, and Z values of the marked points with 3Fl0.3 format. From figure K-2, it should be noted that from the second through the fifth lines is the elevation information in the X direction including the antenna. From the sixth through the ninth lines is the elevation information in the X direction at 500 feet down the runway centerline.

Once the data file is constructed as described above, the profile as shown in figure K-3 is calculated when given two different points (for example, the antenna and aircraft locations). The calculated profile is automati­cally determined by the computer.

-215-

4 9 -10000. o.o 11.5 o.o o.o 11.5 165. o.o 10. 325. o.o 1. 470. o.o o.o -10000. 500. 10. 110. 500. 10. 205. 500. 5. 340. 500. o.o 640. 500. -3. -10000. 1000. 5. o.o 1000. 5. 190. 1000. 1. 315. 1000. -3. 450. 1000. -5. -10000. 1400. o.o -140. 1400. o.o 130. 1400. o.o 230. 1400. -3. 360. 1400. -5. -10000. 1530. o.o -30. 1530. o.o 155. 1530. -9. - ---- -----~

245. 1530. -9. 560. 1530. -9. -10000. 1700. 2. -70. 1700. 2. 170. 1700. -3. 290. 1700. -2. 550. 1700. -2 -10000. 2100. 4. o.o 2100. 4. 190. 2100. 2. 280. 2100. 4. 570. 2100. 4.7 -10000. 2900. 10. -90. 2900. 10. 140. 2900. 10. 330. 2900. 10. 550. 2900. 6. -10000. 4000. 9.6 o.o 4000. 9.6 130. 4000. 9.6 440. 4000. 9.6

Figure K-2. Calculated profile for the 3-D case.

-216-

I N ...... -...)

I

tf=IJ, Z=O)

z

t Base of the Antenna Mast

}

(2900,9 .8) ~

(1575,-9)

Figure K-3. Terrain profile along a line from the antenna mast to the point below an aircraft located as shown in figure K-1 at Springfield, Ohio Municipal Airport.

(4000,0) ... y

APPENDIX L

BIAS ELIMINATION

-218-

The purpose of this appendix is to demonstrate the validity of the tech­nique used in eliminating a constant bias occurring in the measured CDI trace or modeling trace. The technique is simply to subtract the bias from the trace. The assumption is that the scattering fields are not changed drastically by making minor adjustments to the antenna positions. Hence, the resulting trace should exhibit effects almost identical to those created by simply subtracting a constant bias. To verify the assumption by measurement is not an easy task. As one knows, the airplane cannot exactly repeat the same path every time. The problem is overcome by using the com­puter model.

One site has been choosen for the verification. With reference to figure 2-309, the modeling trace has a -13 ~A bias. The antenna heights are 14.13, 28.26, and 42.39 feet for lower, middle, and upper antennas, respec­tively. Physically moving all antennas dow9 to 13.85, 27.70, and 41.55 feet results in moving the calculated trace down about the 0 ~A indication, therefore, giving the trace zero bias. Figure L-1 shows the new calculated trace. One can note the similar characteristics of the traces shown in figures 2-309 and L-1. By using the -13 ~A bias to compensate for the off­set of the trace shown in figure 2-309, zero bias trace is obtained.

-219-

I N N 0 I

N :I:

~

-<( ..3-

0 0 -I

00 u

N :I:

~ .-

8 '+0

Date: Site: Run 1: System:

10,000 20,000 Distance from Antenna Mast (feet)

Figure L-1. Calculated CDI after adjusting antenna heights.

1 Oct 80 TMB 9l 1-16A CEGS

30,000

APPENDIX M

OUTLINE FOR ENDFIRE GLIDE SLOPE SEMINAR

-221-

INTRODUCTION AND HISTORY ----- McFarland TDC Ohio State University Work Watts 1972 Tamiami 1973-1977 Staunton 1974 NAFEC Rock Springs 1977-1980 Butte, Montana 1980-1982

ANTENNA THEORY OF OPERATION Array ----- McFarland Antenna ----- Wat.ts

MONITORING History ---- McFarland Theory -----Watts Measurements

Ground ----- Watts Air ----- McFarland

TERRAIN EFFECTS Use of model Measurements

.,.222-

McFarland McFarland

APPENDIX N

STRIP CHART DATA

-223-

PILOT: 1-A

MEASURED FLIGHT DIRECTOR (FD) RAW DATA (RD) VALUES

CASE 1 CASE 2 CASE 1 CASE 2 73500 3500T 73500 3500T 73500 3500T 73500 3500T

MAX -above 48 60 84 30 24 60 72 112

GLIDE SLOPE

J.JA +be loY< 30 24 30 66 30 NA 57 42

LOCALIZER DEVIATION 60 10 125 20 185 60 110 40 MAX ± FT.

DISTANCE FROM T-D OF MAX EXCURSim ... o 10500 ... o 1000

ON GLIDE SLOPE

DISTANCE FROM T-D OF MAX EXCURSim 6000 12500 9000 14000 ON LOCALIZER

FEET

MAX HEAD DEV 1.6 degrees 2.4 degrees 4.0 degrees 2.5 degrees

MAX PITCH 4.0 degrees 4.0 degrees 5.3 degrees 7.0 degrees

MAX BANK 3.5 degrees 2.0 degrees 3.0 degrees 4.2 degrees

TOUCH DOWN (WHEELS) FT. 1000 1000 1500 1400

-224-

PILOT: 1-B

MEASURED FLIGHT DIRECTOR (FD) RAW DATA ( RD) VALUES

CASE 1 CASE 2 CASE 1 CASE 2 73500 3500T 73500 3500T 73500 3500T 73500 3500T

MAX -above 66 74 78 156 45 162

GLIDE SLOPE

JJA +belo'll 36 18 54 12 72 0

LOCALIZER DEVIATION 50 20 70 30 72 75 MAX ± FT.

DISTANCE FROM T-D OF MAX EXCURSim 2000 500 1500

ON GLIDE SLOPE

DISTANCE FROM T-D OF MAX EXCURSim 13500 1800 1000 ON LOCALIZER

FEET

MAX HEAD DEV. 0.8 degree 1.2 degrees 4.4 degrees

MAX PITCH 3.0 degrees 2.5 degrees 6.5 degrees

MAX BANK 2.5 degrees 3.5 degrees 3.5 degrees

TOUCH DOWN (WHEELS) FT. 1000 1000 1000

Due to running out of paper, no records were taken for Case 2, raw data.

-225-

PILOT: 1-C

MEASURED FLIGHT DIRECTOR (FD) RAW DATA (RD) VALUES CASE 1 CASE 2 CASE 3 CASE 1 CASE 2 CASE 3

ALL CASE S RECORDED AT 73500 and 3500T

MAX -above 66 97 37 51 24. c 25 42 72 90 50 48 0

GLIDE SLOPE ~A +belo'.i 36 42 55 42 19 24 53 25 none 50 42 54

LOCALIZER DEVIATION 60 30 80 25 60 40 110 75 150 60 70 30 MAX ± FT.

DISTANCE FROM T-D OF MAX EX.CURSIO~ 2500 16500 19000 1000 10500 500

ON GLIDE --- --- ----- r--- -

SLOPE FEET

DISTANCE FROM T-D OF MAX EXCURSim 17500 9000 16000 9000 7000 21000 ON LOCALIZER

FEET

-MAX HEAD DEV. 1.6 1.2 1.2 4 3.6 2.0

degrees degrees degrees degrees degrees degrees

MAX PITCH 5.0 3.0 3.5 2.5 3.0 2.5 degrees degrees degrees degrees degrees degrees

MAX BANK 6.0 4.5 3.5 11.0 9.5 3.5 degrees degrees degrees degrees degrees degrees

TOUCH DOWN (WHEELS) FT. 2000 1000 1500 2000 1500 250

-226-

PILOT: 1-D

MEASURED FLIGHT DIRECTOR (FD) RAW DATA (RD) VALUES

CASE 1 CASE 2 CASE 1 CASE 2 73500 3500T 73500 3500T 73500 3500T 73500 3500T

MAX -above 46 96 60 54 33 66 42 36

GLIDE SLOPE

JJA +be loy; 30 .. o 54 12 66 84 78 66

LOCALIZER DEVIATION 65 .. o 54 20 100 30 110 40 MAX ± FT.

DISTANCE FROM T-D OF MAX EXCURSim 2500 6500 2500 4000

ON GLIDE SLOPE

DISTANCE FROM T-D OF MAX EXCURSim 15500 9000 25500 16500 ON LOCALIZER

FEET

MAX HEAD DEV 0.8 degree 0.8 degree 3.2 degrees 2.4 degrees

MAX PITCH 4.0 degrees 2.0 degrees 7.5 degrees 3.5 degrees

MAX BANK 2.5 degrees 4.5 degrees 4.0 degrees 3.5 degrees

TOUCH DOWN (WHEELS) FT. 1500 1200 1000 1000

.14 = 30

-227-

PILOT: 1-E

FLIGHT MEASURED DIRECTOR (FD)

VALUES CASE 3

73500 3500T MAX

-abovE 30 42 GLIDE SLOPE

JJA +belo'lo 12 72

LOCALIZER DEVIATION 100 40 MAX ± FT.

DISTANCE FROM T-D OF MAX EXCURSim ... o

ON GLIDE SlOPE

DISTANCE FROM T-D OF MAX EXCURSim 17500 ON LOCALIZER

FEET

MAX HEAD DEV. 0.8 degree

MAX PITCH 5.0 degrees

MAX BANK 2.5 ·degrees

TOUCH DOWN (WHEELS) FT. 1000

-228-

PILOT: 1-F

MEASURED FLIGHT DIRECtOR (FD) RAW DATA (RD) VALUES

CASE 1 CASE 2 CASE 1* CASE 2 73500 3500T 73500 3500T 73500 3500T 73500 3500T

MAX -above 24 36 63 78 132 0

GLIDE SLOPE

JlA +belo\i 36 54 66 48 132 162

LOCALIZER DEVIATION 120 25 60 30 280 170 MAX ± FT.

DISTANCE FROM T-D OF MAX EX.CURSim ... o ... o 3000

ON GLIDE SLOPE

DISTANCE FROM T-D OF MAX EXCURSim 13500 15000 19000 ON LOCALIZER .

FEET

MAX HEAD DEV 2.4 degrees 2.0 degrees 10.0 degrees

MAX PITCH 3.0 degrees 4.5 degrees 8.5 degrees

MAX BANK 3.8 degrees 5.0 degrees 12.5 degrees

TOUCH DOWN (WHEELS) FT. 1000 1500 1000

* moderate turbulance

-229-

PILOT: 1-G

MEASURED FLIGHT DIRECTOR (FD) RAW DATA (RD) VALUES

CASE 1 CASE 2 CASE 1* CASE 2 73500 3SOOT 73500 3500T 73500 3500T 73500 3500T

MAX -above so 24 90 24 90 168 60 90

GLIDE SLOPE

J.lA +bela~ 30 60 30 66 114 0 54 30

LOCALIZER DEVIATION 180 so 65 30 260 190 170 110 MAX ± FT.

DISTANCE FROM T-D OF MAX EXCURSim .. o 10500 2000 2000

ON GLIDE -SLOPE

DISTANCE FROM T-D OF MAX EXCURSim 25000 16000 22500 12500 ON LOCALIZER

FEET

MAX HEAD DEV 1.6 degrees 2.4 degrees 12.0 degrees 6.0 degrees

MAX PITCH 1.0 degrees 5.0 degrees 6.5 degrees 5.0 degrees

MAX BANK 3.0 degrees 3.0 degrees 12.5 degrees 9.5 degrees

TOUCH DOWN (WHEELS) FT. 1100 1000 2000 1000

* moderate turbulance

-230-

PILOT: 1-H

MEASURED FLIGHT DIRECTOR (FD) RAW DATA ( RD) VALUE;S

CASE 1 CASE 2 CASE 1 CASE 2 CASE 3* ALL CASES RECORDED AT 73500 and 3500T

MAX -abovE 126 168 42 168 96 120 66 150 168 150

GLIDE SLOPE

lJA +belm. 72 90 36 66 36 30 60 6 42 6

LOCALIZER DEVIATION 140 so 130 95 160 95 100 140 . MAX ± FT.

DISTANCE FROM T-D OF MAX EXCURSIOl' 2000 2000 1000 2000 4500

ON GLIDE - - ---- ----- --- -SLOPE

DISTANCE FROM T-D OF MAX EXCURSIOl' 12000 17000 15500 21000 18000 ON LOCALIZER

FEET

MAX HEAD DEV. 2.4 2.8 5.2 4.5 4.8 degrees degrees degrees degrees degrees

MAX PITCH 5.5 s.o 5.0 7.5 6.0 degrees degrees degrees degrees degrees

MAX BANK 4.0 5.50 8.0 6.0 9.5 degrees degrees degrees degrees degrees

TOUCH DOWN .. o (WHEELS) FT. 1000 2000 1500 1500

* moderate turbulance

-231-

PILOT: 1-I

MEASURED FLIGHT DIRECTOR (FD) RAW DATA (RD) VALUES

CASE 1 CASE 2 CASE 1 CASE 2 73500 3SOOT 73500 3500T 73500 3500T 73500 3500T

MAX -above 42 60 54 66 66 63 54 150

GLIDE SLOPE . J.IA +belo~ 36 90 60 57 48 42 60 36

LOCALIZER DEVIATION so 20 85 30 185 110 280 70 MAX ± FT.

DISTANCE FROM T-D OF MAX EXCURSim 500 ... o 10500 500

ON GLIDE SLOPE

DISTANCE FROM T-D OF MAX EXCURSIO~ 7500 15500 11000 19000 ON LOCALIZER

FEET

MAX HEAD DEV 1.6 degrees 1.7 degrees 6.0 degrees 4.4 degrees

MAX PITCH 4.5 degrees 4.0 degrees 4.5 degrees 3.5 degrees

MAX BANK 5.0 degrees 6.5 degrees 13.5 degrees 7.0 degrees

TOUCH DOWN (WHEELS) FT. 1200 1500 1700 2000

100% Tolerance

-232-

PILOT: 1-J

MEASURED FLIGHT DIRECTOR (FD) RAW DATA (RD) VALUES

CASE 1 CASE 2 CASE 1 CASE 2 73500 3500T 73500 3500T 73500 3500T 73500 3500T

MAX -above 150 162 30 NA 30 48

GLIDE SLOPE

lJA +belo'll 42 0 72 150 72 54

LOCALIZER DEVIATION 115 60 60 45 130 50 MAX± FT.

DISTANCE FROM T-D OF MAX EXCURSim 2500 1000 12500

ON GLIDE SLOPE

DISTANCE FROM T-D OF MAX EXCURSim 17500 7000 20700 ON LOCALIZER

FEET

MAX HEAD DEV 1.6 degrees 1.8 degrees 2.8 degrees

MAX PITCH 4.5 degrees 3.5 degrees 5.0 degrees

MAX BANK 3.0 degrees 2.0 degrees 3.0 degrees

.

TOUCH DOWN (WHEELS) FT. 1500 1000 1500

-233-

PILOT: 1-K

MEASURED FLIGHT DIRECTOR ( FD) VALUES

CASE 1 CASE 2 CASE 3 73500 3500T 73500 3500T 73500 3500T

MAX -above 93 42 48 6 12 0

GLIDE SLOPE

llA +belo'lti 30 15 66 54 18 54

LOCALIZER DEVIATION so so so 20 40 10 MAX ± FT.

DISTANCE FROM T-D OF MAX EXCURSim 5500 16500 300

ON GLIDE SLOPE

DISTANCE FROM T-D OF MAX EXCURSim 4500 18500 11500 ON LOCALIZER

FEET

MAX HEAD DEV 0.6 degree 1.6 degrees 1.0 degree

MAX PITCH 5.0 degrees 3.0 degrees 3.5 degrees

MAX BANK 3.5 degrees 3.0 degrees 1.9 degrees

TOUCH DOWN (WHEELS) FT. 1500 1000 500

-234-

PILOT: 1-M

MEASURED FLIGHT DIRECTOR (FD) RAW DATA (RD) VALUES

CASE 1 CASE 2 CASE 3 CASE 1 CASE 2 ALL CASES RECORDED AT 73500 and 3500T

MAX -abovE 42 48 60 25 12 6 69 32 26 66

GLIDE SLOPE

llA +belo";o 30 12 33 48 11 54 114 66 45 0

LOCALIZER DEVIATION 30 9 40 30 45 12 160 15 121 30 MAX ± FT.

DISTANCE FROM T-D OF MAX EXCURSim 0 10500 200 9400 12000

ON GLIDE SLOPE

DISTANCE FROM T-D OF MAX EXCURSim 8500 8000 15500 21500 13600 ON LOCALIZER

FEET

MAX HEAD DEV 0.8 1.2 0.6 2.8 1.6 degree degrees degree degrees degrees

MAX PITCH 4.5 4.0 3.0 6.0 3.3 degrees degrees degrees degrees degrees

MAX BANK 2.5 3.5 1.7 4.5 2.5 degrees degrees degrees degrees degrees

TOUCH DOWN (WHEELS) FT. 1000 1500 1000 1500 1000

-235-

PILOT: 1-N

MEASURED VALUES

CASE 1 73500 3500T

MAX -above

GLIDE SLOPE

llA +belo¥

LOCALIZER DEVIATION MAX ± FT.

DISTA.l.~CE

FROM T-D OF MAX EXCURSim

ON GLIDE SLOPE

DISTANCE FROM T-D OF MAX EXCURSIOli ON LOCALIZER

FEET

MAX HEAD DEV

MAX PITCH

MAX BANK

TOUCH DOWN (WHEELS) FT.

+ 125% Tolerance Ill (FD) missed

42 42

54 48

80 20

12500

18000

0.4 degree

4.5 degrees

1.5 degrees

2000

#2 (RD) two times #3 (PD) two times

FLIGHT DIRECTOR (FD) RAW DATA (RD)

CASE 2 CASE 3 CASE 1 73500 3500T 73500 3500T 73500 3500T

60 54 18 24 70 66

42 6 12 54 27 69

106 45 60 30 120 55

10500 ... o 5500

12500 13500 10000

1.6 degrees 0.4 degrees 1.6 degrees

4.5 degrees 3.0 degrees 3.0 degrees

2.5 degrees 1.0 degree 4.2 degrees

1500 1500 1500

-236-

APPENDIX 0

SAMPLE OF AN ANALOG STRIP CHART RECORDING

-237-

-238-

APPENDIX P

LEGEND PAGE FOR PROVIDING IDENTIFICATION AND

QUANTITATIVE ASSESSMENT OF THE ANALOG STRIP CHARTS

-239-

PITCH ( ) c::>-

c::r DEGREES ( '6 ~

!\lENT •• • ll.lftKER ( ) c:::r.

IL ~ ( ) c:.::>-1

t ( ,< ->-~ BANK ( -j_~- ~ DEGREES

TEST NO. c ? < r ~ C::Jc ~

3 <:::::; .

DISPlACEMENT ( ) f=;. FROM ~.:.._: LOCALIZER < ) < >i F EfT c J < r

c ;< >-: 4 c 'c-=r

DISTANCE c 'c--:::r FROM < 'cy T/D POINT ~ FEET X 100 ( ) ( >-

c::Y. IS c::::>-=-'

( ) c r: ALTITUDE

< ,LY, c:::>

FEET ( ::J( ~ ~ c=:r ~

• ~ ~ c:::>-

AIRSPEED ~ KNOTS c :::J( >-

~____; ~ ~ 7 ( )( ).!

Rlli''WAY ~

~ HEADING c:::>· DEVIATION ~

EVENT ... • NARKtR DEGREES c::::>-(=): I L W£"/t;Hr ~N WN£ELS~ c::::>-=-I FLY UP ~

GUDE SLOPE ~ ..--:) ( . )J DEVIATION ... c:::::r

RECORDING SPEED MICROAMPS ~ ____ MMPS FLY DOWN

( } c . ~ c::J--==

-240-

APPENDIX Q

FORTRAN PROGRAM LISTINGS FOR THE COMPUTATION OF

GLIDE SLOPE REFERENCE DATUM HEIGHTS

-241-

Introduction. This report describes two computer programs for calculating the reference datum height (RDH) and achieved reference datum height (ARDH) as specified by FAA Order 8240.47, "Determination of Instrument Landing System (ILS) Glidepath Angle, Reference Datum Heights, and Ground Point of Intercept," dated May 5, 1983.

The programs described in this document are designed to make determination of the reference datum heights as expedient as possible. The programs are written in standard FORTRAN IV language, and may easily be adapted for use with personal computers and programmable pocket calculators, provided their memory capacity is sufficient. Two programs are available, one using stan­dard input and output data files and the other using an interactive com­puter terminal for input and output.

The interactive program RDHINT is basically intended for use in the field. The user can access the IBM-4341 mainframe at Ohio University with a por-table computer terminal and modem, allowing the reference datum height calculations to be completed within minutes from any location having availability to a telephone line.

The standard program RDH is designed for use on the output data generated by mathematical models of glide slope systems. In addition, the program may be used to calculate the reference datum heights from measured data by simply creating an input data file with the differential amplifier samples in the appropriate format. This feature is particularly useful when the input and/or output data files must be saved.

Analytical Method. The programs described in this paper are based on a linear regression analysis. The discussion here will be qualitative only, and readers wishing to examine the analysis in more detail should refer to FAA Order 8240.47.

The best-fit straight line (BFSL) is determined by fitting a straight line to a set of samples of the differential amplifier trace over a segment of the glidepath. The method used is a least squares linear regression analy­sis, which determines the slope and height of the BFSL relative to the reference point. The regression analysis is performed on a set of evenly spaced data points over a selected sampling interval. The RDH is computed in ILS Zone 2 and the ARDH is computed over the glidepath segment starting 6000 feet from the threshold to ILS Point C. The height of the BFSL above the runway threshold is the designated reference datum height.

The computations are straightfoward in that only basic algebraic expressions are used, but the quantity of required calculations makes the use of some type of computational aid desirable for preventing human errors. The method used to compute the reference datum heights is shown in figure Q-1.

Usage. The following procedure is intended as a guide for use of the programs. Before running the program the site data and differential amplifier data must be collected. The site data required are:

-242-

( START ) ----,r----

READ SITE INPUT DATA

INITIALIZE VARIABLES

READ DIFF. AMP DATA

COMPUTE RECTANGULAR COORD! NATES FROM DIFF AMP DATA

COMPUTE BFSL EQUATION FROM COORDINATE DATA

WRITE OUTPUT PATH ( RD H OR ARD H ) .

STOP

-243-

WRITE SITE INPUT DATA

OFFSET

BACK

c

EL

REFANG

WIDTH

K

Perpendicular distance from runway centerline to reference point in feet.

Distance along runway centerline from threshold to centerline abeam reference point in feet.

Distance along extended runway centerline from threshold to ILS Point C in feet.

Threshold elevation minus reference point elevation in feet.

Reference path angle in degrees.

Reference path width (angle between 75 microamp points) in degrees.

Number of differential amplifier samples. Determines which reference datum height (RDH or ARDH) is computed.

K = 21 K = 11

RDH is computed ARDH is computed

Note that the variable 'C' is used only in ARDH calculations and must be omitted from the site data input file when the RDH is being computed.

The differential amplifier samples are determined according to the proce­dure defined in Order 8240.47. The differential samples are entered in units of the small divisions on the chart recording for the interactive program so that the user does not have to perform any calculations by hand. This assumes that the flight recording was made with the standard ± 100 microamps full scale setting on the chart recorder, and that the chart paper is divided into 100 unit divisions (2 microamps per unit division). For the standard program RDH, the input differential samples are entered to an input file in units of microamps, making RDH compatible with the outputs of math models in use at Ohio University. The program can be used on measured differential amp data by entering the samples (in microamps) to a data file with the format (17X,F10.5). Note also that a site data file must be created containing the inputs listed above in the format 7(F10.5,/,I2).

The interactive program RDHINT provides user insructions to the terminal. To use this program with the output directed back to tlie terminal, the user simply enters the command 'RDHINT' and the appropriate file definitions will be automatically set. If the output data are to be sent to another logical device (i.e. disk, tape, printer, etc.), then the following file definitions must be set before executing the program:

Unit 5 - User Terminal Unit 6 - User Terminal Unit 10 - Desired Output Device

-244-

To terminate execution of the program the user must enter a null return.

To run the program RDH the following must be done:

1. Define the following unit devices for the I/O statements;

Unit 5 -Unit 6 -Unit 13-

Data file containing differential amp samples Output Data file containing site input data

2. Create an input data file containing the site data listed above. This file should be called 'SITE DATA A1' and must be in the format (7(F10.5,/),I2).

3. Create an input data file containing the differential amplifier samples (in microamps) in the format (17X,F10.5). The math model OUGS3D output 'FILES' has this format.

4. Begin execution of the program by entering the command

RDH 'FN' 'FT' 'FM'

where the file identifiers are for the input file as defined in step 3.

5. The output file will be placed on the users 'C' disk and will be named 'FN RDH C1' where FN is defined as in step 4.

Every attempt has been made to insure accuracy and ease of use with these programs. Please report any problems to the author.

-245-

Sample Interactive Sessions.

rdhint EXECUTION BEGINS •••

RDHINT - REFERENCE DATUM HEIGHT PROGRAM ENTER TITLE FOR THIS RUN (UP TO 72 CHARACTERS)

ardh sample interactive programming session ENTER NUMBER OF DIFF AMP SAMPLES - RDH=21 , ARDH=11

11 ENTER REFERENCE OFFSET FROM CENTERLINE (FT)

400 ENTER REFERENCE SETBACK FROM THRESHOLD (FT)

1000 ENTER REFERENCE PATH ANGLE (DEGREES)

3.0 ENTER REFERENCE PATH WIDTH (DEGREES)

0.7 ENTER THRESHOLD MINUS REFERENCE ELEVATION (FT)

2.0 ENTER DISTANCE FROM THRESHOLD TO ILS POINT C (FT)

908 ARDH SAMPLE INTERACTIVE PROGRAMMING SESSION

REFERENCE DATA :

OFFSET FROM CENTERLINE SETBACK FROM THRESHOLD THRESHOLD-REFERENCE ELEVATION REFERENCE PATH ANGLE REFERENCE PATH WIDTH DISTANCE TO ILS PT. C

400.0000 1000.0000

2.0000 3.0000 0.7000

908.0000

IS INPUT DATA CORRECT??? (YES=1, N0=2) 1

ENTER 11 DIFF AMP SAMPLES (UNIT DIVISIONS) ENTER POINT 1 -5 ENTER POINT 2 -4 ENTER POINT 3 -3 ENTER POINT 4 -2 ENTER POINT 5 -1 ENTER POINT 6

-246-

0 ENTER POINT 7 1 ENTER POINT 8 2 ENTER POINT 9 3 ENTER POINT 10 4 ENTER POINT 11 5

SAMPLE DCDI(UNIT DIVISIONS) 1 -5.0 2 -4.0 3 -3.0 4 -2.0 5 -1.0 6 o.o 7 1.0 8 2.0 9 3.0

10 4.0 11 5.0

ARE THE DIFF AMP SAMPLES ALL CORRECT ??? (YES=1, N0=2) 1

ENTER RETURN TO RECEIVE DATA ARDH SAMPLE INTERACTIVE PROGRAMMING SESSION

REFERENCE DATA :

OFFSET FROM CENTERLINE 400.0000 SETBACK FROM THRESHOLD 1000.0000 THRESHOLD-REFERENCE ELEVATION 2.0000 REFERENCE PATH ANGLE 3.0000 REFERENCE PATH WIDTH 0.7000 DISTANCE TO ILS PT. C 908.0000

SAMPLE THETA X xo y

1 2.953 7000. 7011. 362. 2 2.963 6491. 6503. 337. 3 2.972 5982. 5995. 311. 4 2.981 5472. 5487. 286. 5 2.991 4963. 4979. 260. 6 3.000 4454. 4472. 234. 7 3.009 3945. 3965. 208. 8 3.019 3436. 3459. 182. 9 3.028 2926. 2954. 156.

10 3.037 2417. 2450. 130. 11 3.047 1908. 1949. 104.

-247-

XLIT

2546. 2037. 1528. 1018.

509-. o.

-509. -1018. -1528. -2037. -2546.

R· ,

ANGLE = 2.902 DEG.

REFERENCE HORIZONTAL ADJUSTMENT 155.82031 FT. REFERENCE VERTICAL ADJUSTMENT 7.89951 FT.

ARDH = 56.6 FT.

-248-

Fortran Program Listing.

C*********************************************************************** c c c c c c c c c c c c c

THIS PROGRAM CALCULATES THE RDH (REFERENCE DATUM HEIGHT) OR ARDH (ACHIEVED REFERENCE DATUM HEIGHT) ACCORDING TO THE REGRESSION Al\fALYSIS OF FAA ORDER 8240.4 7.

WRITTEN BY: LARRY BRADY 1-11-85 AVIONICS ENGINEERING CENTER DEPARTMENT OF ELECTRICAL ENGINEERING . OHIO UNIVERSITY, ATHENS, OHIO

C*********************************************************************** c c C INPUT VARIABLES c c c C OFFSET- DISTANCE FROM REFERENCE TO RUNWAY C CENTERLINE IN FEET c C BACK- DISTANCE ALONG CENTERLINE FROM THRESHOLD C TO CENTERLINE ABEAM REFERENCE IN FEET c C C- DISTANCE ALONG EXTENDED CENTERLINE FROM C THRESHOLD TO ILS POINT C IN FEET c C EL- THRESHOLD ELEVATION MINUS REFERENCE ELEVATION C IN FEET c C REFANG- REFERENCE PATH ANGLE IN DEGREES c C WIDTH- REFERENCE PATH WIDTH BETWEEN 75 MICROAMP POINTS C IN 90 AND 150 HZ. REGIONS IN DEGREES c C K- INTEGER WHICH DETERMINES NUMBER OF SAMPLES AND C REFERENCE DATUM HEIGHT ( RDH OR ARDH) COMPUTED c C K=21, RDH IS COMPUTED C K=ll, ARDH IS COMPUTED c C DCDI(I)- DIFFERENTIAL AMPLIFIER SAMPLES IN MICROAMPS c c C***********************************************************************

DIMENSION DCDI(21),THETA(21),X(21),Y(21),X0(21),XLIT(21) READ(13,49)0FFSET,BACK,C,EL,REFANG,WIDTH,K

-249-

49 FORMAT(6(F10.5,/),I2) c C PRINT INPUT DATA c

WRITE(6,26)0FFSET,BACK,EL,REFANG,WIDTH 26 FORMAT(1X, 1 REFERENCE DATA : 1 ,//,5X, 1 0FFSET FROM CENTERLINE

(1 ,T40,F9.4,/,5X, 1 SETBACK FROM THRESHOLD 1 ,T40,F9.4,/,5X, 1 THRESHO

(LD-REFERENCE ELEVATION 1 ,T40,F9.4,/,5X, 1 REFERENCE PATH ANGLE 1

(,T40,F9.4,/,5X, 1 REFERENCE PATH WIDTH I ,T40,F9.4) IF(K.EQ.ll)WRITE(6,66)C

66 FORMAT(5X, 1 DISTANCE TOILS PT. C = 1 ,T40,F9.4) WRITE(6 ,67)

67 FORMAT(/) c c C INITIALIZE VARIABLES AND CONSTANTS c

c

XA=24304.4+BACK XB=3500.0+BACK XC=BACK+C X6=6000.0+BACK IF(K.EQ.21)X1=XA IF(K.EQ.21)X2=XB IF(K.EQ.11)X1=X6 IF(K.EQ.l1)X2=XC XINC=(X1-X2)/FLOAT(K-1) XCURR=X1+XINC XTOTAL=O.O YTOTAL=O.O SUMXY=O.O SUMXSQ=O.O

C READ IN DIFF AMP DATA c

DO 1000 I=1 ,K READ(5,55)DCDI(I)

55 FORMAT(l7X,F10.5) c C THIS SECTION CALULATES THE PATH ANGLES c

THETA(I)=REFANG+(WIDTH/150.0)*DCDI(I) . 1000 CONTINUE

c C CALCULATE DISTANCE INCREMENTS AND LOAD ARRAY c

DO 2000 I=1 ,K X(I)=XCURR-XINC XO(I)=SQRT(X(I)**2+0FFSET**2) Y(I)=TAN(THETA(I)/57.29578)*XO(I) XTOTAL=XTOTAL+X(I) YTOTAL=YTOTAL+Y(I) XCURR=X(I)

-250-

2000 CONTINUE c C PERFORM STATISTICAL ANALYSIS c

XAVG=XTOTAL/FLOAT(K) YAVG=YTOTAL/FLOAT(K) SUMXY=O.O SUMXSQ=O.O DO 3000 I=1 ,K XLIT(I)=X(I)-XAVG SUMXY=SUMXY+XLIT(I)*Y(I) SUMXSQ=SUMXSQ+XLIT(I)**2

3000 CONTINUE c C COMPUTE OUTPUT DATA c

c

RATIO=SUMXY/SUMXSQ ANGLE=ATAN(RATIO)*S7.29578 DISPL=YAVG/RATIQ-XAVG HEIGHT=YAVG-RATIO*XAVG IF(K.EQ.Zl)RDH=(BACK+DISPL)*TAN(ANGLE/57.29578)-EL IF(K.EQ.11)ARDH=(BACK+DISPL)*TAN(ANGLE/57.29578)-EL

C PRINT OUTPUT DATA c

WRIT£(6,60) 60 FORMAT(lX, 'SAMPLE' ,4X, 'THETA' ,SX,' X ',SX,' XO ',5X,

* I y ' , 6X, ' XLIT ' , /) WRITE(6,61)(I,THETA(I),X(I),XO(I),Y(I),XLIT(I),I=1,K)

61 FORMAT(3X,I3,5X,F5.3,5X,F6.0,5X,F6.0,5X,F6.0,5X,F7.0) WRITE(6,63)DISPL,HEIGHT

63 FORMAT(1X,/,' REFERENCE HORIZONTAL ADJUSTMENT ',T35,F10.5,' FT. <' ,/,' REFERENCE VERTICAL ADJUSTMENT ',T35,F10.5,' FT.',/) WRITE(6,62)ANGLE

62 FORMAT(5X,'ANGLE = I ,F5.3,' DEG.'/) IF(K.EQ.21)WRITE(6,64)RDH IF(K.EQ.11)WRITE(6,65)ARDH

64 FORMAT(lX,5X,' RDH = 1 ,F5.1,' FT.') 65 FORMAT(1X,5X,'ARDH = I ,F5.1,' FT.')

STOP END

-251-

C*********************************************************************** c c c c c c c c

·C. c c c

THIS PROGRAM CALCULATES THE RDH (REFERENCE DATUM HEIGHT) OR ARDH (ACHIEVED REFERENCE DATUM HEIGHT) ACCORDING TO THE REGRESSION ANALYSIS OF FAA ORDER 8240.47.

WRITTEN BY: LARRY BRADY 1-11-85 AVIONICS ENGINEERING CENTER DEPARTMENT OF ELECTRICAL ENGINEERING OHIO UNIVERSITY, ATHENS, OHIO

c C*********************************************************************** c c C INPUT VARIABLES c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c

OFFSET-

BACK-

c-

EL-

REFANG-

WIDTH-

K-

DCDI(I)-

DISTANCE FROM REFERENCE TO RUNWAY CENTERLINE IN FEET

DISTANCE ALONG CENTERLINE FROM THRESHOLD TO CENTERLINE ABEAM REFERENCE IN FEET

DISTANCE ALONG EXTENDED CENTERLINE FROM THRESHOLD TO ILS POINT C IN FEET

THRESHOLD ELEVATION MINUS REFERENCE ELEVATION IN FEET

REFERENCE PATH ANGLE IN FEET

REFERENCE PATH WIDTH BETWEEN 75 MICROAMP POINTS IN 90 AND 150 HZ. REGIONS IN DEGREES

INTEGER WHICH DETERMINES NUMBER OF SM1PLES AND REFERENCE DATUM HEIGHT (RDH OR ARDH) COMPUTED

K=21, RDH IS COMPUTED K= 11 , ARDH IS COMPUTED

DIFFERENTIAL AMPLIFIER SAMPLES IN UNIT DIVISIONS (2 MICROAMPS = 1 UNIT DIVISION)

C*********************************************************************** DIMENSION DCDI(21),THETA(21),X(21),Y(21),X0(21),XLIT(21),INBUF(18) INTEGER TITLE(18) PRINT 1

-252-

1 FORMAT(/////' RDHINT- REFERENCE DATUM HEIGHT PROGRAM ') c C ENTER TITLE OF THIS RUN c

PRINT 2 2 FORMAT(' ENTER TITLE FOR THIS RUN (UP TO 72 CHARACTERS) ')

READ(S,3,END=99)TITLE c C ENTER TYPE OF CALCULATION DESIRED (RDH OR ARDH) c 19 PRINT 33 33 FORMAT(' ENTER NUMBER OF DIFF AMP SAMPLES - RDlJ.=21 ARDH=ll 1

)

READ(5,3,END=99)INBUF CALL NCVT(INBUF,K,ROUT,DOUT,IDEC,IERR) IF(IERR.NE.O)GO TO 19 IF(K.NE.21.AND.K.NE.11)GO TO 19

C IF(K.NE.21.0R.K.NE.11)GO TO 19 c C ENTER REFERENCE POSITION DATA c 20 PRINT 4 4 FORMAT(' ENTER REFERENCE OFFSET FROM CENTERLINE (FT) ')

READ(5,3,END=99)INBUF CALL NCVT(INBUF,IOUT,OFFSET,DOUT,IDEC,IERR) IF(IERR.NE.O)GO TO 20

c 21 PRINT 5 5 FORMAT(' ENTER REFERENCE SETBACK FROM THRESHOLD (FT) ')

READ(5,3,END=99)INBUF

c

CALL NCVT(INBUF,IOUT,BACK,DOUT,IDEC,IERR) IF(IERR.NE.O)GO TO 21

C ENTER REFERENCE PATH DATA c 22 PRINT 6 6 FORMAT(' ENTER REFERENCE PATH ANGLE (DEGREES) ')

READ(S,3,END=99)INBUF

c

CALL NCVT(INBUF,IOUT,REFANG,DOUT,IDEC,IERR) IF(IERR.NE.O)GO TO 22

C ENTER REFERENCE PATH WIDTH. THIS IS THE ANGLE BETWEEN THE 75 C MICROAMP POINTS IN THE 90 AND 150 HZ. REGIONS DETERMINED FROM C THE PATTERN 'B' (LEVEL RUNS) MEASUREMENTS. c 23 PRINT 7 7 FORMAT( 1 ENTER REFERENCE PATH WIDTH (DEGREES) 1

)

READ(5,3,END=99)INBUF

c c

CALL NCVT(INBUF,IOUT,WIDTH,DOUT,IDEC,IERR) IF(IERR.NE.O)GO TO 23

C ENTER SITE DATA

-253-

c 25 PRINT 9 9 FORMAT(' ENTER THRESHOLD MINUS REFERENCE ELEVATION (FT) ')

READ(5,3,END=99)INBUF

c

CALL NCVT(INBUF,IOUT,EL,DOUT,IDEC,IERR) IF(IERR.NE.O)GO TO 25

C IF ARDH IS DESIRED, DISTANCE C FROM THRESHOLD TO ILS PT. 'C' C MUST BE ENTERED c

IF(K.EQ.21)GO TO 222 26 PRINT 10 10 FORMAT(' ENTER DISTANCE FROM THRESHOLD TOILS POINT C (FT)')

READ(5,3,END=99)INBUF CALL NCVT(INBUF~IOUT,C,DOUT,IDEC,IERR) IF(IERR.NE.O)GO TO 26

222 CONTINUE 3 FORMAT(18A4) c c C INITIALIZE VARIABLES AND CONSTANTS c

c

X.A=24304.4+BACK XB=3500.0+BACK XC=BACK+C X6=6000.0+BACK IF(K.EQ.21)X1=XA IF(K.EQ.21)X2=XB IF(K.EQ.11)X1=X6 IF(K.EQ.11)X2=XC XINC=(X1-X2)/FLOAT(K-1) XCURR=X1 +XINC XTOTAL=O.O YTOTAL=O.O SUMXY=O.O SUMXSQ=O.O

C PRINT INPUT DATA c 1001 WRITE(6,747)TITLE 747 FORMAT(1X,18A4/)

WRITE(6,91)0FFSET,BACK,EL,REFANG,WIDTH 91 FORMAT\1X,'REFERENCE DATA :',//,5X,'OFFSET FROM CENTERLINE

<',T40,F9.4,/,5X,'SETBACK FROM THRESHOLD 1 ,T40,F9.4,/,5X,'THRESHO (LD-REFERENCE ELEVATION ',T40,F9.4,/,5X,'REFERENCE PATH ANGLE ' (,T40,F9.4,/,5X,'REFERENCE PATH WIDTH 1 ,T40,F9.4)

IF(K.EQ.11)WRITE(6,66)C 66 FORMAT(5X,'DISTANCE TOILS PT. C ',T40,F9.4)

WRITE(lO ,67) 67 FORMAT(/) c C CHECK TO SEE INPUT DATA IS CORRECT

-254-

c 726 PRINT 727 727 FORMAT(' IS INPUT DATA CORRECT??? (YES=1, N0=2)')

READ(5,3,END=99)INBUF CALL NCVT(INBUF,J,ROUT,DOUT,IDEC,IERR) IF(IERR.NE.O)GO TO 726 IF(J.EQ.2)GO TO 19 IF(J.NE.1)GO TO 726

c C READ IN DIFF AMP DATA c

PRINT 88,K 88 FORMAT(' ENTER' ,I2,' DIFF AMP SAMPLES (UNIT DIVISIONS) ') 29 DO 1000 I=1,K 89 PRINT 77,I 77 FORMAT(' ENTER POINT ',I2)

READ(5,3,END=99)INBUF CALL NCVT(INBUF,IOUT,DCDI(I),DOUT,IDEC,IERR) IF(IERR.NE.O)GO TO 89

c C THIS SECTION CONVERTS SMALL UNIT DIVISIONS ON STANDARD CHART C RECORDER PAPER INTO MICROAMPS FOR THE PATH ANGLE CALCULATION c

DCDIUA~2.0*DCDI(I) c C THIS SECTION CALCULATES THE PATH ANGLES c

THETA(I)=REFANG+(WIDTH/150.0)*DCDIUA 1000 CONTINUE c C CHECK TO SEE IF DIFF AMP DATA· IS CORRECT c 299 PRINT 331 331 FORMAT(//1X,'SAMPLE' ,T10,'DCDI(UNIT DIVISIONS)')

DO 4000 M=1,K WRITE(6,92)M,DCDI(M)

92 FORMAT(5X,I3,T10,F6.1) 4000 CONTINUE 330 PRINT 332 332 FORMAT(' ARE THE DIFF AMP SAMPLES ALL CORRECT??? (YES=1, N0=2)')

READ(5,3,END=99)INBUF CALL NCVT(INBUF,L,ROUT,DOUT,IDEC,IERR) IF(IERR.NE.O)GO TO 330 IF(L.EQ.1)GO TO 333 IF(L.NE.2)GO TO 330

334 PRINT 335 335 FORMAT(' ENTER NUMBER OF INCORRECT DIFF AMP SAMPLE')

READ(5,3,END=99)INBUF CALL NCVT(INBUF,N,ROUT,DOUT,IDEC,IERR) IF(IERR.NE.O)GO TO 334 IF(N.GT.K)GO TO 334

336 PRINT 337,N

-255-

337 FORMAT(' ENTER DIFF AMP SAMPLE NUMBER ',I2,/) READ(5,3,END=99)INBUF CALL NCVT(INBUF,IOUT,DCDIR,DOUT,IDEC,IERR) IF(IERR.NE.O)GO TO 336 DCDI(N)=DCDIR DCDIUA=DCDIR*2.0 THETA(N)=REFANG+(WIDTH/150.0)*DCDIUA GO TO 299

333 CONTINUE c C CALCULATE DISTANCE INCREMENTS AND LOAD ARRAY c

DO 2000 I=1,K X(I)=XCURR-XINC XO(I)=SQRT(X(I)**2+0FFSET**2) Y(I)=TAN(THETA(I)/57.29578)*XO(I) XTOTAL=XTOTAL+X(I) YTOTAL=YTOTAL+Y(I) XCURR=X(I)

2000 CONTINUE c C PERFORM STATISTICAL ANALYSIS c

XAVG=XTOTAL/FLOAT(K) YAVG=YTOTAL/FLOAT(K) SUMXY=O.O SUMXSQ=O.O DO 3000 I=1 ,K XLIT(I)=X(I)-XAVG SUMXY=SUMXY+XLIT(I)*Y(I) SUMXSQ=SUMXSQ+XLIT(I)**2

3000 CONTINUE c C COMPUTE OUTPUT DATA c

c

RATIO=SUMXY/SUMXSQ ANGLE=ATAN(RATI0)*57.29578 DISPL=YAVG/RATIO-XAVG HEIGHT=YAVG-RATIO*XAVG IF(K.EQ.21)RDH=(BACK+DISPL)*TAN(ANGLE/57.29578)-EL IF(K.EQ.11)ARDH=(BACK+DISPL)*TAN(ANGLE/57.29578)-EL

C PRINT OUTPUT DATA c 888 PRINT 889 889 FORMAT(' ENTER RETURN TO RECEIVE DATA')

READ(5,3,END=1234) 1234 CONTINUE

WRITE(10,747)TITLE WRITE(10,91)0FFSET,BACK,EL,REFANG,WIDTH IF(K.EQ.11)WRITE(10,66)C WRITE(10,60)

-256-

60 FORMAT(lX,/,'SAMPLE' ,4X,'THETA' ,SX,' X I ,SX,' xo I ,SX, *' y I ,6X,' XLIT I,/)

63 FORMAT(lX,/,' REFERENCE HORIZONTAL ADJUSTMENT ',T35,Fl0.5,' FT. (

1 ,/,' REFERENCE VERTICAL ADJUSTMENT ',T35,Fl0.5,' FT.',/)

WRITE(l0,6l)(I,THETA(I),X(I),XO(I),Y(I),XLIT(I),I=l,K) 61 FORMAT(3X,I3,5X,F5.3,5X,F6.0,5X,F6.0,5X,F6.0,5X,F7.0)

WRITE(l0,62)ANGLE 62 FORMAT(lX,/,SX,'ANGLE = 1 ,F5.3,' DEG.'/)

WRITE(l0,63)DISPL,HEIGHT IF(K.EQ.2l)WRITE(l0,64)RDH IF(K.EQ.ll)WRITE(l0,65)ARDH

64 FORMAT(SX,' RDH = I ,FS.l,' FT.') 65 FORMAT(SX,' ARDH = I ,FS.l,' FT.')

STOP c C NULL RETURN ENTERED ENDS PROGRAM c 99 PRINT 98 98 FORMAT(' NULL RETURN ENTERED. END OF PROGRAM')

END

-257-

APPENDIX R

MAINTENANCE AND REFURBISHMENT OF TAMIAMI TEST FACILITY

~258-

One of the key items leading to a successful conclusion of this contract has been the test site maintained at Tamiami airport in Miami, Florida. Through the cooperation of the Dade County Airport Authority, Ohio University maintains a test station consisting of numerous ILS antenna arrays and ancillary transmitting equipment.- This allows rapid set-up and testing of most·types of ILS facilities. The station is located on an essentially ideal ground plane, allowing easy isolation of system pertur­bations.

After several years of operation, the test site has been in need of refur­bishment. A special contract was issued for this purpose, and the resulting improvements should allow even better results to be obtained in the future. The _following list reflects the repairs and improvements made to the test site.

1. Infrequently used equipment was moved to the transmitter site for International Flight Service (IFST) for storage.

2. The 40-foot house trailer was scrapped.

3. The remaining four (4) buildings and the equipment were repaired, refurbished and/or restored in accordance with the following list:

A. LOCALIZER HUT

(1) Seal and repair roof (2) Replace power drop, local electrician (3) Repair door frame, rotted (4) Rework floor damaged by rot (5) Replace all screens on siding (6) Paint siding (7) Install floor under seal (8) Floor joist replacement (9) Place gravel under hut

(10) Remove & relocate storage boxes for "0" Ring antennas (11) Apply weed killer (12) Acquisition of materials (13) Repair RF feeder cable (14) Cut & remove light lane tower (15) Remove fungus, polish & wipe all electronics (16) Repair power distribution box at array (17) Paint V-Ring antennas & obstruction- lights

B. "FIFTH WHEEL" TRAILER

(1) Seal & repair roof (2) Prime & paint structure and siding (3) Remove &·relocate 8 distribution boxes,

tube-type monitoring equipment (4) Re-level, install gravel (5) Repair lighting equipment

-259-

(6) Apply weed killer (7) Repair, replace floor sections (8) Install seals for floor (9) Exterminate

C. GLIDE SLOPE HUT

(1) Clear out, clean, inspect power drop and feed (2) Replace communication antennas & masts (3) Remove bees (hive) (4) Replace siding screws (5) Replace & seal roof (6) Install gravel under hut (7) Apply weed killer (8) Repair rot on door frame & baseboard (9) Replace floor

(10) Remove fungus, polish & wipe electronics (11) Paint structure (12) Paint equipment rack (13) Remove cables (14) Replace dehumidifier (15) Repair air conditioner (16) Inspect & repair power cables and touch-up paint on

GS tower

D. DOBBINS HUT

(1) Remove equipment and store at IFST (2) Seal & repair roof (3) Prime and paint structure on all sides (4) Repair floor rot (5) Install gravel base (6) Paint and repair furniture items (7) Repair lights (8) Refurbish telescoping tower

E. HOUSE TRAILER

(1) Remove tower sections to IFST (2) Sort, wipe, seal stored equipment (3) Relocate equipment to IFST (4) Remove and terminate electrical supply (5) Salvage useful items from trailer (6) Prepare trailer for removal (7) Haul trailer to salvage, disposal

Initial contacts were made with FAA personnel responsible for the Boeing B-727-200 advanced flight simulator at the Mike Monroney.Aeronautical Center, Oklahoma City, Oklahoma. The following data relevant to simulator operation and capabilities were obtained:

1. As many as 20 variables can be recorded on magnetic tape during simulations. This allows foe data reduction to be mostly auto­mated.

-260-

APPENDIX S

INITIAL ASSESSMENT OF LOCALIZER STANDARDS AND TOLERANCES

-261-

Initial contacts were .made with FAA personnel responsible for the Boeing B-727-200 advanced flight simulator at the Mike Monroney Aeronautical Center, Oklahoma City, Oklahoma. The following data relevant to simulator operation and capabilities were obtained:

1. As many as 20 variables can be recorded on magnetic tape during simulations. This allows for data reduction to be mostly auto­mated.

2. The sample rate of the computer-generated CDI (Course Deviation Indicator) is 5 Hz.

3. The simulator is intended to serve primarily as a training system for pilots, and secondarily as a research tool.

4. The coordinate system used in the simulation is centered at the touchdown point, 1000 feet ·from threshold.

5. The simulation uses a rectangular coordinate system that provides a resolution of 10 feet and uses latitude/longitude data.

6. The simulator is programmed for approaches to 9 airports at pre­sent, with the most advanced being OKC (Oklahoma City). Only long runway_siwtlations (~eater than 9000 feet) exist at present.

7. Some sample code has already been written and implemented on the simulator to generate derogated localizer courses.

8. The best time for conducting simulator experiments is during shortened work weeks (less than 5 days), since training sessions are not held at these times.

9. The flight director used in the simulator is a Collins 329B-9A, and is part of a FD-109Z system.

The simulator can be a very important tool in the evaluation of aircraft and pilot responses to non-ideal ILS signals. In addition, performing this type of work could allow the realism of the simulation to be improved by introducing non-ideal localizer and glide slope courses which are known to exist.

-262-