multi surf aircrafts

8/2/2019 Multi Surf Aircrafts

1/16

*Corresponding author. Tel.:#44-141-330-5042; fax:#44-141-330-5560.

E-mail address: [email protected] (L. Smrcek)

Aircraft Design 2 (1999) 191}206

Aerodynamic characteristics of multi-surfaceaircraft con"gurations

Z. PaH tek, L. Smrcek*

VZLU, Aeronautical Research and Test Institute, 19905 Prague, Czech Republic

Department of Aerospace Engineering, University of Glasgow, G12 8QQ, Scotland, UK

Abstract

This paper describes the wind tunnel testing of a specially designed aircraft model allowing systematic

variation of geometric parameters related to overall aircraft con"gurations. The experiment work was

carried out in the 1.8 m low-speed wind tunnel at VZLU, Aeronautical Research and Test Institute in Prague.

The resultant data created an aerodynamic database for numerical modelling and veri"cation. In addition,

numerical validation of CFD package FLUENT was performed in the computerised #uid dynamic laborat-

ory at the Department of Aerospace Engineering, University of Glasgow as a part of the ongoing research

collaboration between both institutions. The wind-tunnel test program had two aims

E to provide basic aerodynamic data e!ects of multi-surface aircraft con"gurations with a view to assessing

the degree to which speci"c design features such as a combination of canard, wing and tail plane are

bene"cial to aircraft aerodynamic performance

E to provide an aerodynamic database for numerical validation.

The aerodynamic data for di!erent geometrical aircraft con"gurations with the same wing, fuselage,

horizontal tail and canard surfaces were compared. The results show the magnitude of lift, drag and pitching

moment depending on the angle of attack and various positions of canard, wing and horizontal tail with

respect to one another. They have been compared with studied references. The experimental results are in

accordance with theoretical predictions. 2000 Elsevier Science Ltd. All rights reserved.

1. Introduction

Aircraft development in many ways means to improve its aerodynamic characteristics all the

time. This aim can be achieved by modifying existing classical aircraft con"gurations or by

adopting the more radical approach of multi-surface aircraft geometrical con"gurations. The most

common way of dealing with such a problem is to delete the longitudinal stabiliser aft of the wing

1369-8869/99/$ - see front matter 2000 Elsevier Science Ltd. All rights reserved.

PII: S 1 3 6 9 - 8 8 6 9 ( 9 9 ) 0 0 0 1 5 - 4


2/16

Nomenclature

C*

lift coe$cient

angle of attack

C"

drag coe$cient

C

zero lift pitching moment coe$cient (positive nose up)C

pitching moment coe$cient (positive nose up)

c wing chord

h

stabiliser horizontal position (see Fig. 1)

stabiliser vertical position (see Fig. 1)

h

canard horizontal position (see Fig. 1)

canard vertical position (see Fig. 1)

and substitute it by something else. Such an aircraft con"guration is being used in both civilian andmilitary applications. The purpose of this research was to "nd the con"guration using canard, wing

or tail which will improve aircraft performance in comparison to the conventional aircraft surfacearrangement.

Multi-surface aircraft geometry has been known since the beginning of aviation, but it was only

during the 1960s that some of these aircrafts got past the prototype stage into serial production.

The "rst of these were military aircraft and in the late 1970s the "rst civil canard application wasused for sport/recreation aircraft design. In 1985 the prototype of a new canard commuter aircraft

was built with the view of bringing the project into serial production. The trial was successful,

however, serial production was later suspended due to the high cost and technical problems of theaircraft involved. The results of analytical analysis of some unconventional aircraft con"gurations

were published mainly by Keith [1], Rokhsaz and Selberg [2}4]. Some experimental work by

Feistel [5], Ostowar [6] had been studied prior to the research to identify the wind tunnel set-upand model geometry.

The research which is presented in this paper examines the geometric parameters of the aircraft

model and the e!ect on the aerodynamic characteristics such as, lift curve gradient and maximum

lift coe$cient, drag polars and pitching moment coe$cient. The experimental tests results for all

selected model variants are presented in this paper.

2. Wind tunnel model

A specially designed modular aircraft wind tunnel model was made which allowed changes to the

aircraft con"guration. The metal aircraft model was constructed from modules such as canard,

fuselage, wing and horizontal tail. The basic model dimensions are shown in Fig. 1. Modularconstruction was used to enable the creation of three basic aircraft con"gurations, i.e. conventional

wing}horizontal tail aircraft con"guration, unconventional with foreplane canard-wing and

canard-wing}horizontal tail aircraft con"gurations. It was also possible to change the relative

position of canard, wing and horizontal tail as shown in Fig. 1.

192 Z. Pa&tek, L. Smrcek/ Aircraft Design 2 (1999) 191}206


3/16

Fig. 1. Wind tunnel model [dimensions in mm].

The wing was rectangular with a pro"le NACA 0015 along the span, without dihedral or twist.

The modular aircraft model allowed positioning the wing in low or high con"guration. Canard and

horizontal stabilisers were also of rectangular shape with an area 30% of the wing, NACA 0012pro"le and without dihedral or twist. It was possible to place the canard in nine positions on the

forward fuselage or to take it away. Also, the horizontal tail could be "xed in four positions on the

"n or taken away.

3. Wind tunnel test set-up

All tests were conducted in the 1.8 m low-speed wind tunnel at VZLU, Aeronautical Research

and Test Institute in Prague in the Czech Republic [7]. This institute has provided very goodquality wind tunnel facilities for the Czech aircraft industry for many years. The particular windtunnel used in this study was an atmospheric open-section, closed return, with a maximum velocity

of 55 m/s. Forces and moments were measured on a six component overhead gravitational balance.

The model was mounted in the wind tunnel by a system of connecting wires. The model hanging

in the wind tunnel can be seen in Fig. 2. The Reynolds Number based on the wing chord was3.35;10 which is at least 10 times lower than for real-size aircraft but still not below critical

values.

For calculations of non-dimensional force coe$cients reference used was the wing area including

part of the wing shadowed by the fuselage.

Z. Pa&tek, L. Smrcek/ Aircraft Design 2 (1999) 191}206 193


4/16

Fig. 2. Model in the wind tunnel.

The basic characteristic length for pitching moment coe$cient calculation was the wing (mean

aerodynamic) chord. Pitching moment is related to the quarter chord.

All measurements were without sideslip.

4. Experimental data analysis

The main features of the results from the wind-tunnel test programme are presented and

analysed in this section. Selected aerodynamic data are represented by lift, drag and pitching

moment coe$cients and are shown and analysed for three basic aircraft con"gurations.

1. Conventional aircraft wing}horizontal tail con"guration.

2. Canard-wing con"guration.

3. Canard-wing}horizontal tail con"guration.

The various combination of vertical and longitudinal position with respect to one another

of canard, wing and horizontal tail is listed in the tables for each aircraft con"guration measured.



5/16

Table 1

Con"guration wing}tail h

High wing 0

4c 0.5c

1.5c

1c

Low wing 4c 1.5c

2c

All measured data relate to the untrimmed wind tunnel model which means that chords of canard,

wing and horizontal tail are always parallel to the longitudinal axis of the fuselage.

5. Wing}horizontal tail con5guration

Measured in the wind-tunnel were six combinations of wing and tail relative positions. Threecombinations were for low-wing con"guration and three for high wing con"guration. The combi-

nations and geometry of wing and tail that were tested are listed in Table 1. For the high wing

con"guration the vertical position of the stabiliser was in the region c442c, where the

horizontal position was kept constant at h"4c. The horizontal tail incidence with respect to the

wing was zero.

Figs. 3a}c present experimental data for a wing}horizontal tail con"guration.

5.1. Lift curve, C*

Fig. 3a presents the measured lift curve for each of the six con"gurations examined in the study.From the "gure it is possible to observe that the lift curve slope is a!ected by the vertical position of

the tail relative to the high or low wing con"guration. Maximum recorded increase in C*

deriva-tive was in the region of 0.15 rad\. This was due to the reduction of wing and fuselage e!ect on the

tail #ow (downwash).

5.2. Maximum lift coezcient, C*

Increase in vertical distance,

, between the wing and tail for both low and high wing

con"gurations was the main reason for increment inC

* as shown in Fig. 3a. This e!ect is againexplained by the reduction of wing and fuselage e!ect on the tail #ow.

5.3. Minimum drag coezcient, C"

The experimental results are illustrated in Fig. 3b. In the case of high wing con"guration the

minimum drag coe$cient was decreasing with increased vertical distance,

. But for higher values

of

this decrease in C"

was not signi"cant. Maximum C"

reduction occurred when the

distance

was between zero and 1.5c and was C" "0.0012.



6/16

Fig. 3. Aerodynamic coe$cients of the conventional con"guration. (a) Lift curve. (b) Drag polar. (c) Pitching moment.

In the case of low wing aircraft con"guration, minimum drag coe$cient was not a!ected with

change in vertical position of tail on the "n.

5.4. Pitching moment coezcient, C

Fig. 3c shows the pitching moment coe$cient curves. From this "gure it can be observed that

increased vertical distance between the wing and tail makes the curve slope more negative, thus the

neutral point was shifting backwards. Also, by increasing the distance between the wing and tail, the

downwash on the tail was reduced and this resulted in a decrease in zero lift pitching moment, C

.



7/16

Fig. 4. Aerodynamic coe$cients of the canard-high wing con"guration. (a) Lift curve. (b) Drag polar. (c) Pitching

moment.

6. Canard-wing con5guration

The results from the experiments which were carried out to "nd the e!ect of various canard-wing

positions, with respect to one another, on aerodynamic forces and pitching moment are presentedin Figs. 4 and 5. Table 2 shows that 18 canard-wing combinations which were measured in the wind

tunnel.



8/16

Fig. 5. Aerodynamic coe$cient of the canard-low wing con"guration. (a) Lift curve. (b) Drag polar. (c) Pitching

moment.

Nine variations of measurement were a combination of canard and high wing con"gurations. Inthis case horizontal distances between the canard and the wing varied in the region 2c4h

44c

and the vertical distances in the region of, !c440. The remaining 9 con"gurations were

canard and the low wing. The horizontal distance between canard and wing changed in the region

2c4h44c and the vertical distance in the region of 04

4c.



9/16

Table 2

Con"guration Canard-wing h

High wing !1c

2c !0.5c

0

!1c

3c !0.5c

0

!1c

4c !0.5c

0

0

Low wing 2c 0.5c

1c

03c 0.5c

1c

0

4c 0.5c

1c

6.1. Lift curve, C*

As shown in Figs. 4a and 5a the variation of lift with angle of attack was almost linear in the!6((7 degree range. A slight increase in lift curve slope due to the increased h

distance is

noticeable in the linear range ofC*

for both low and high wing con"gurations. Changes in verticaldistance,

, did not in#uence the lift curve slope.


As can be seen in Figs. 4a and 5a the horizontal distance between canard and wing, h

, has nota!ected the maximum lift coe$cients but the canard vertical position with respect to the wing

a!ected C* noticeably. The lowest C* was recorded for high wing con"guration when thevertical distance between canard and wing was

"!0.5c. In this con"guration the canard

unfavourably a!ected the #ow over the wing, particularly in the region of angle of attackcorresponding to C

* .


Shortening of the horizontal distance between canard and wing in#uenced the minimum drag

coe$cient only in the region ofh

between 3c and 2c when the reduction ofC"

was only 0.001

as can be Figs. 4b and 5b.



10/16

Table 3

Con"guration

Canard-wing}tail

h

h

High wing 2c !1c!0.5c

0

3c !1c

!0.5c 4c 1.5c

0

4c !1c

!0.5c

0

4c !0.5c 0

!0.5c 4c 0.5c

Low wing 2c 00.5c

1c

0

3c 0.5c 4c 1.5c

1c

0

4c 0.5c

1c

4c 1c 1c

1c 4c 2c

Increased vertical distance between the canard and the wing contributed to the reduction of

minimum drag coe$cient in both cases when the canard was shifted above or below the wing level.

The highest C"

was recorded when "0.


As indicated in Figs. 4c and 5c the pitching moment curve was a!ected by the horizontal

distance, h

, between the canard and the wing in the case of high and low wing con"gurations. For

increased h

the neutral point moved forward resulting in a reduction of the static longitudinal

stability. It should be noted that all con"gurations are unstable.Change in vertical distance between canard and wing

was followed by a slight change in the

zero lift pitching moment coe$cient. This was due to the fact that the change in vertical canard

position with respect to the centre of gravity also changed canard drag moment which contributeto the overall pitching moment. The distance,

, did not a!ect the curve gradient.

7. Canard-wing}horizontal tail con5guration

The experimental programme involved tests of twenty two di!erent canard-wing}horizontal tail

aircraft con"gurations with relative positions to each other as stated in Table 3.



11/16

Half of the model con"gurations were with high wing. In 9 cases the canard position was

adjusted so that the horizontal and vertical distance from the wing were in the region

2c4h44c,!c4

40 while the tail was in the "xed position h

"2c and

"1.5c. In

two cases when canard position was at h"4c and

"!0.5c the tail position was changed to

h"4c and

"0 and

"0.5, and the e!ect on the aerodynamic characteristics were

observed.The other half of the model con"gurations were with low wing. Also in nine cases the canard

position was adjusted so that the horizontal and vertical distances from the wing were in the region

2c4h44c, 04

41c while the tail was in the "xed position h

"4c and

"1.5c. In two

cases the canard was placed at h"4c and

"1c the horizontal tail position was adjusted to

h"4c,

"1c and

"2c and the e!ect on the aerodynamic characteristics were observed.

The results are presented in Figs. 6}8.

7.1. Lift curve, C*

If the horizontal canard distance from the wing h

is increased from 2c to 4c, accordingto Figs. 6a}8a, the increase in the lift curve gradients is about 0.1 rad\ in the linear range

ofC*

.

The vertical distance between canard and wing in#uenced the lift curve slope by 0.4 rad in some

model con"gurations. This e!ect was not observed in the canard-wing con"guration nor did the

vertical position of the horizontal tail a!ect C*

in the case of the conventional horizontal wing}tailcon"guration. This leads to the explanation that the canard a!ected the #ow over the tail directly

and also at the same time a!ected the wing #ow and hence the downwash, which resulted in

di!erent tail #ow characteristics compared with conventional wing}tail con"gurations.

When the vertical distance between the tail and canard (!

) was less than 1.5c, the tail was

a!ected most by the canard and this resulted in the reduction of C*

.

The e!ect of the vertical distance between the wing and the tail on C*

can be expected to be thesame as in the case of conventional model con"guration. This means increased vertical distance,

, will result in a slight increase in C*

.


In high wing con"guration the horizontal distance between canard and wing, h , did not a!ectC*

, but for the low wing con"guration an increase in h

resulted in a slight positive incrementin C

* .

As Figs. 6 and 7 show the vertical distance between canard and wing,

, a!ected C*

in the

same way as in the case of canard-wing con"guration. In the case of"!0.5c the critical angle

of attack was reduced and a decrease in C*

was recorded. This was due to the unfavourablecanard e!ect on the wing #ow.

From the similarity between results observed for canard-wing and canard-wing}horizontal tail

con"gurations it is possible to conclude that relative position of canard and tail does not a!ect

C*

signi"cantly.



12/16

Fig. 6. Aerodynamic coe$cients of the canard-high wing}horizontal tail con"guration. (a) Lift curve. (b) Drag polar.

(c) Pitching moment.


According to the experimental results presented in Figs. 6b}8b changes in position of both

canard and tail surface did not a!ect minimum drag coe$cient.


Pitching moment curves as seen in Figs. 6c and 7c were signi"cantly a!ected by two geometrical

parameters of the model. Firstly, increasing horizontal distance between the canard and the wing,



13/16

Fig. 7. Aerodynamic coe$cients of the canard-low wing}horizontal tail con"guration. (a) Lift curve. (b) Drag polar.

(c) Pitching moment.

h

, shifts the neutral point forward and the pitching moment curve slope is increased. It should benoted that all con"gurations are unstable.

Secondly, the geometrical parameter which a!ected the pitching moment was the canard vertical

position,

, and the tail vertical position,

. This could be explained by analysing and

comparing experimental results for di!erent model geometrical con"gurations.Even this e!ect was not observed in the case of canard wing and conventional wing}horizontal

tail con"gurations. It was concluded that, for example, the di!erences in the pitching moment

curves, between conventional and canard-wing}horizontal tail con"gurations, were the result of

the canard #ow e!ect on the tail. This was caused directly by canard #ow wake and also by canard



14/16

Fig. 8. Canard-wing}horizontal tail con"guration. (a) Lift curve. (b) Drag polar. (c) Pitching moment.

e!ect on the wing #ow and its downwash, which also a!ects and changes horizontal tail #owcharacteristics.

The experiments show that a signi"cant change in pitching moment coe$cient occurs in the

practical range of angles of attack in relation to the direct e!ect of the canard on the horizontal tail.This e!ect was signi"cantly reduced by increasing the vertical distance,

between canard and

horizontal tail. However, according to the experimental data for the modular model, only a change of

vertical distance was needed to eliminate the negative e!ect of canard #ow wake on the tail. When

the horizontal distance between canard and tail is, (h#h

)"6c or 7c the vertical distance



15/16

between canard and tail must be !

"2c. If the horizontal distance was (h

#h

)"8c

the vertical distance between canard and tail needed to eliminate canard e!ect on tail was

(!

)"2.5c.

When the angle of attack was increased beyond the critical value, it was observed that the "rst

#ow separation occurred on the canard, then the pitching moment slope gradient was reversed and

the neutral point moved backwards.The canard vertical position also a!ected the zero lift pitching moment coe$cient, C

. The

reason and e!ect was the same as in the case of canard-wing and conventional wing}horizontal tail

con"gurations.

8. Discussion

All measurements relate to an untrimmed model. In fact, all models cannot be trimmed in this

condition because the zero-lift pitching moment is almost zero. The only deviation is due to thefuselage drag acting at its centerline and therefore contributing a nose-up pitching moment

(positive, but not enough). This is obviously not a problem, but this makes the conclusions withregard to C

rather super#uous since all di!erences in the aircraft design process will be negated

by an appropriate choice of the horizontal tail or canard angle of incidence. This will mean that,

for the conventional aircraft model the horizontal tail lift will decrease and for the canard it will

increase for a given angle of attack. This is true for the case of positive static margin, however, for

negative static margin the observations could be di!erent.

The conventional con"guration (Fig. 3c) is the only one with a stable pitching moment slope.There should be a small e!ect due to the high wing position, which brings the wing aerodynamic

centre backwards relative to the moment reference point with increasing incidence, introducing

some non-linearity. But more signi"cantly, this con"guration has a fairly linear behaviour up toand including the stall. This implies that the hotizontal tail is always stabilising and (probably)

remains unstalled. Therefore the measured maximum lift coe$cients are useful values, because the

aircraft will be trimable up to and including the stall. That is not the case with the canard andthree-surface con"guration. All of these show a strong pitch down behaviour at about 83 angle of

attack. This is most probably caused by canard stall, because the further forward the canard is

located, the larger is the pitch-down moment. For the case of the canard-wing}tail con"guration

(Figs. 7 and 8) the slope becomes clearly negative at high angles of attack, indicating that above 83angle of attack the model behaves like a conventional aft-tail (stable) con"guration. However, with

the centre of gravity at the appropriate location forward it would be much too stable to be #yable

at angles of attack more than 83.For a real design the canard angle of attack will be increased in order to have a useful positive

zero-lift pitching moment and a much more forward moment reference point to achieve positive

stability. As a consequence, the critical angle of attack will be reduced even more, making the

maximum lift coe$cient even smaller.

The canard con"gurations will have an inferior trimmed maximum lift performance. Obviouslythe situation would become totally di!erent for controllable or free-#oating canards.

Small di!erences in the zero-lift drag were observed. These can be explained by the increased

wetted area in the case of the three surface model. These di!erences are obviously small inproportion to the zero-lift drag of the fuselage.



16/16

Of more interest is, however, the maximum lift-to-drag ratio. The only noticeable e!ect found

from the experiments was that the conventional con"guration seems to have the highest value,

probably due to the lowest tailplane lift.

9. Conclusions

Wind tunnel tests and their results are presented for di!erent canard-wing}horizontal tail

aircraft con"gurations in low-speed applications. The data indicates that the interference between

canard, wing and tail will a!ect basic aerodynamic characteristics. This e!ect depends on therelative position of all three lifting surfaces. A specially designed modular aircraft wind tunnel

model was made. This model proved to be practical in creating a su$cient number of aircraft

con"gurations. Obtained experimental data represents the aeronautical database for numerical

validation which has been performed in the Department of Aerospace Engineering at the Univer-sity of Glasgow using the CFD code FLUENT [8].

Acknowledgements

This project has been supported by the VZLU, Aeronautical Research and Test Institute in

Prague and the Department of Aerospace Engineering at the University of Glasgow. The authors

acknowledge the help and support given by the leadership of both institutions.

References

[1] Keith MW, Selberg BP. Aerodynamic canard/wing parametric analysis for general-aviation application. AIAA

Journal of Aircraft 1985.

[2] Selberg BP, Rokhsaz K. Aerodynamic tradeo!study of conventional canard and trisurface aircraft systems. AIAA

Journal of Aircraft 1986.

[3] Rokhsaz K, Selberg BP. Analytical study of three-surface lifting systems. Paper 850866, Society of Automotive

Engineering Inc., 1986.

[4] Rokhsaz K, Selberg BP. Three-surface aircraft: optimum vs. Typical. AIAA Journal of Aircraft 1989.

[5] Feistel TW, Corsiglia VR, Levin DB. Wind-tunnel measurements of wing-canard interference and a comparison with

various theories. Paper 810575, Society of Automotive Engineerings Inc., 1982

[6] Ostowar C, Naik D. An experimental study of a three lifting surface con"guration. IICAS-86-1.94.

[7] Patek Z. Basic aerodynamic characteristics of some unconventional aircraft con"gurations. CVUT Theses, 1991.[8] Zverina O, Smrcek L. Computation of basic aerodynamic characteristics of multi-surface aircraft con"gurations.

University of Glasgow, Department of Aerospace Engineering, Report No. 9921.


multi surf aircrafts

Documents