ad#: d - apps.dtic.mil · component part notice this paper is a component part of the following...

COMPONENT PART NOTICE

THIS PAPER IS A COMPONENT PART OF THE FOLLOWING COMPILATION REPORT:

TITLE: Proceedings of the U.S. Army Sympsium on Gun Dynamics (7th)

Held in Newport, Rhode Island on 11-13 May 1993.

To ORDER THE COMPLETE COMPILATION REPORT, USE AD-A278 075

THE COMPONENT PART IS PROVIDED HERE TO ALLOW USERS ACCESS TO INDIVIDUALLY

AUTHORED SECTIONS OF PROCEEDING, ANNALS, SYMPOSIA, ETC. HOWEVER, THE COMPONENTSHOULD BE CONSIDERED WITHIN THE CONTEXT OF THE OVERALL COMPILATION REPORT AND

NOT AS A STAND-ALONE TECHNICAL REPORT.

THE FOLLOWING COMPONENT PART NUMBERS COMPRISE THE COMPILATION REPORT:

AD#: AD-P009 060 AD#:

AD#: thru ADN:

AD#: AD-P009 091 AD#:

Accesion For

NTIS CRA&IDTIC TAB T ICU nannounced n_ I LECTE

n .............................. M AY 1 7 1994

By~ D......B y .................. . . . . . . .Dil-t; ibtition } "

Availability Codes -

Avail and I orDist Special

DTIC FORM 463 OPI: DTIC-TIDMAR 85

\ WILKERSON, BERMAN, AND LI

*TITLE: A MODAL SURVEY OF THE MIAI MAIN WEAPON SYSTEM

Dr. Stephen Wilkerson, Mr. Morris Bcrman, and Mr. Ting Li A D-P009 084U.S. Army Research LaboraloryATTN: AMSRL-WT-PDAberdeen Proving Ground, MD 21(X)5-5066(410) 278-6131

*ABSTRACT:

When a tank fires its main weapon system, a complex chain of dynamic events begins. The projectile is accelerated downan imperfect gun tube (gun tubes are never perfectly straight) being forced by the burning propellant gases. During this time.considerable forces and interactions between the projectile and gun tube are possible. In sonic cases, the response of the twosystems (ie., the projectile and gun tube) are not fully understood. One method of examining the dynamics of thesecomplicated chain of events is to develop straightforward numerical models. As a first step in assuring the accuracy of thesemaodels, verification of the assumptions, such as geometry and boundary conditions, must be examined.

This paper discusses an experimental modal survey of the MIAI main w,'apon system. Both horizontal and verticalcomponents are examined to find the actual frequencies and mode shapes of the system. A simple numerical model isdeveloped using the finite element method and subsequent!y compared to the experimental results of the modal survey. Adiscussion of the system's attributes, as well as the techniques and assumptions used to develop the finite element model arediscussed at length. Possible shortcomings in the numerical approximation are outlined as well.

*BIOGRAPHY:

*PRESENT ASSIGNMENT: October 1989 to Present: Rcsearch Engineer, The U.S. Army Research Laboratory, AberdeenProving Ground, MD 21005-5066.

*PAST EXPERIENCE:

July 1983 to October 1989: Mechanical Engineer, The Naval Surface Weapons Center, 10901 New Hampshire Ave., SilverSpring, MD 20903-5000

1982 to 1983: Aerospace Engineer, Naval Air Systems Command, Washington, DC.

*DEGREES HELD:

The Johns Hopkins University, Baltimore, MD. Ph.D., Department of Mechanics, April 6, 1990. Dissertation title:"A Boundary Integral Approach to Three-Dimensional Underwater Explosion Bubble Dynamics."

George Washington University, Washington, DC. Masters of Science, Structural Engineering, May 1985. Thesis title: "Shipand Submarine Response to an Underwater Explosion." GPA - 3.50.

The Johns Hopkins University, Baltimore, MD, Bachelor of Science, Mechanical Engineering, May 1982, GPA - 3.00.

94-14479401 \\lll\\\l\l\lll \\ll i\\\lI \\\ll

A MODAL SURVEY OF THE MIAI MAIN WEAPON SYSTEM

Dr. Stephen Wilkerson, Mr. Morris Berman, arid Mr. Ting LiU.S. Army Research Laboratory

A'TN: AMSRL-WT-PDAberdeen Pioving Ground, MD 21(X)5-5066

(410) 278-6131

OBJECTIVE

The primary purpose of this modal test was to provide experimental verification for a finite element (FE) model of theMIAI tank M256 main gun. The FE model is being developed for the gun accuracy improvement program. Major verticaland horiz'.mtal rigid body and bending modes of the main gun will be obtained from experimental modal analysis (EMA).The realism of the FE boundary conditions (between the main gun and its supporting structu'es) may be, enhanced bycomparison to the EMA model. The secondary objectives of this experiment included measuring the nonlinearity of the gunmodal responses and determining, if any, the effects of the hydraulic gun elevation mechanism on the gun dynamics.

TEST SETUP

All of the heat shields on the gun were removed and the accelerometers were attached directly to the gun tube. Theaccelerometer locations were situated in two lines, 90' apart. This positioning permits one set of vertical data and one setof horizontal measurements to be obtained. In order to resolve the fifth bending knode, 15 locations, 10 inches apart, werechosen from the muzzle to the king nut. Only four locations, due to inaccessibility, were measured within the turret. Table 1and Figure 1 detail the placement of the measurement locations. The chosen excitation locations were at the muzzle (location200) aud in the king-nut area (location i00). Both vertical and horizontal excitation was utilized. The excitation source wasa 50-ib electrodynamic shaker. To ensure uniform energy distribution over the frequency range of interest, a controlled truerandom signal was utilized as the excitation signal.

SIGNAL PROCESSING

A 16-channel Genrad 2515 spectrum analyzer was utilized to acquire the measurements. Since more than 16 measurementchannels were utilized, 2 runs were required for each configuration. Frequency response functions (FRFs) and coherencefunctions were retained in the final data set. In addition, several auto spectra were also retained to evaluate the excitationsignal. The FRFs were collected with the following parameters:

Maximum Frequency 640 HzFrame Size 2560 FrequencyFrequency 0.25 HzNumber of 50Window Type Hanning

FRFs collected to assess nonlinearity of the structure utilized a higher frequency resolution in order to detect smallfrequency shifts.

FINITE ELEMENT MODEL DESCRIPTION

The M256 120-mm gun system consists of a number of important parts which contribute to the system's dynamiccharacteristics. Figure 2 is a cut-a-way, three-dimensional view showing some of the parts in the cradle assembly area. Thisfigure, as well as the model, do not consider the mantelet or trunnion mounts. This is discussed in the conclusion sectionof this paper. The objective of this fiist FE model, namely, a beam element representation of the system, was to make asimple, easily modified, numerical model of the M256 gun-recoil system. Then this simple model will be used as a Iearningtool for more sophisticated FE models in the future. The complicated set of boundary conditions in the system is firstexamined with the simplistic model. That model is then compared with experiments and improved based on observeddiscrepancies. Model attributes are incrementally changed until a satisfactory numerical representation of the M256's dynamiccharacteristics is made. The characteristics (realistic geometric and boundary conditions) of the simple model are thenincorporated into more sophisticated and robust thiee-dimensional FE models which are not so quickly modified or analyzed.

402

WILKERSON, BERMAN, AND LI

ExcitationLocation 100

18 20/21

19 9 61 5 14 13 12 11 10 9 8 7 6 5 4 3 21

ExcitationLocation

200

Figure 1. Measurement Location Diagram.

BREECH

•CRADLE

BEARING

THRUST NUT

KING NUT ,

S• PISTON

ADAPTOR

Figure 2. Cut-a-Way View of the M256 Assembly.

403


Table 1. Measurement Location Dimension

Location N:. Distance From Muzzle Description of Gauge location

1 .05 @ Muzzle

2 10. On Gun Tube

3 20. On Gun Tube4 30. On Gun Tube

5 40. On Gun Tube

6 50. On Gun Tube

7 60. On Gun Tube

8 70. On Gun Tube

9 80. On Gun Tube

10 90. On Gun Tube

11 100. On Gun Tube

12 110. On Gun Tube

13 120. On Gun Tube

14 130. On Gun Tube

15 135. On Gun Tube

16 180. On Cradle

17 208.7 On RF of Gun Tube

18 218. On Breech

19 218. On Breech

20 -191. On Elev. Mechanism

21-1. On Elev. Mechanism

After examining the iiechanical drawings of the system and then examining its associated parts, as well as its assemblyand disassembly, a simplified mode! of the M256 consisting of what is believed to be critical components was developed.The critical components were combined into nine individual parts. These nine essent.al parts are highlighted in Figure 2.Additionally, Figure 3 shows a computer aided design/computer aided manufacture (CAD/CAM) drawing of the simplifiedparts which are going to be included in the FE model. For the beam element model, each of the parts is represented usingconcentric cylindrical beam elements with associated properties to the pieces shown in Figure 3. However, the adaptedbearing, king-nut, and thrust nut were included in the beam element model of the piston. The piston was assumed 'o herigidly attached, as was the breech, to the gun tube at the contact points. Similarly, the cradle, which supports the structure,was modeled, at first, with rigid contact points where the piston rcsttd on the cradle's surfaces (Wilkerson ei a. 1993).Initially, it was understood that this would be insufficicnt for modeling the recoil system's motion. Nonetheless, it wasassumed that this would be sufficient for finding the firs, five vertical flexural frequencies and associated mode shapescorrectly. Due to clearances between the piston and cradle in the real system, that was not thc case In particular, the contact

40,4


/m

Piston and Spring

-K-

Adapter Ki g-u

Cradle,- us u

Figure 3. CAD/CAM Simplification of M256 Critical Components.

points, where the piston slides inside the cradle, required more flexibility than the rigid connections allowed. Subsequently,the rigid connections were replaced by gap elemenu;. These gap ctements gave the system more flexibility and betterapproximated the actual systems dynamic characteristics. (Note: gap elements are simulated with regard to the vertical modalanalysis as spring-dashpots, but act as gap elements allowing the system to recoil.) The final model consisted of the gun tube,breech, piston, and cradle assembly. The cradle was simply supported at the same location as the trunnions. The elevatingmecha iism was approximated in the model as a .pring-dashpot which attached between the cradle assembly and a rigidmount.

In order to check the FE model's geometric prorenies, it was assembled incrementally. First, the model of just the guntube was compared with some experimental results of the free-free vibrational frequencies of that part (Rowelamp 1987).This was a good initial check of the model's most important part, namely, the gun tube. Alterwards, the breech, pist.n, andcradle assemhies were included in the model. Results comparirg experimental frequencies to the numerical model'spredictions are given in Table 2 and Figure 4 for the gun tube alone. A comparison of the experimental and numerical modelfor the assemb'cd M256's frequencies and mode s;hapes are given in the resulLs section of this paper.

405


Table 2. (,ion Tu))c Only - Free-Free

- Beam Model 341) Model

Mode Shape Experimental ~7 r . I Ero

1 37.5 36.4 3.0 35.6 5.3

2 106.3 106.1 02104.5 1.75

3 211.3 214.4 1.5 210.6 0.34

4 337.0 394.3 3.6 336.5 0.15

5 482.5 510.3 5.8 4861.3

ROHN114 IN SC3CI4IEPUNK7 C I ESI'.15AN ERIEG. AN IOWI4NUCNO.1201.3 337.5 482 5

OVERALL INS5 .3 1 3 § E -0 3. JA

4Z-APR-A7 r -,=. -

FR 0 (Nv2), (LIN),00se

Figure 4. Frequency Response Function for Gun-Tube in Free-Fk~cc Mode.

RESULTS

The natural frequencies found in the experimental analysis for v -itlcal 'irid horizontal flex ural components are. summarizedin Tables 3 and 4. Table 3 also summarizes the numerical predictioi.; fro ii~te model in the vertical directiion, No numericalpredictions have been made for the horizontal flexural components. The modc indicator funý'.ions for frequency ranges0-200 Hiz and 200-640 Hz are given in Figures 5 and 6 for the vertical and Figures 7 and 8 for thc horizontal, respectively.

406

WRKERSON, BERMAN, AND Ll

Table 3. Vertical Modal Parameters

Magnitude Freqitncy 7Mode Flexural Damping % ir, Driving J ErroiLabel Mode of Critical Point, FRF Exp. Calc. (%)

S--- 2.724 11.24 i 1.6 10.5 112 1.096 20.02 30.27 31.4 3.8

3 2 4.074 4.78 79.5 85.7 7.8

4 3 1.127 4.37 182. 178.9 1.7

5 4 1.143 4.66 276, 287.9 4.0

6 4 2.355 11.9 376. 366. 2.8

7 5 2.392 1.37? 495. 455. 8.8

Table 4. Horizontal Modal Parameters

Mode Damping % Magnitude in Frequency

Label of Critical Driving Point FRF (HZ)

1 1.33 25.65 16.57

2 2.17 2.36 40.85

3 4.74 4.96 73.8

4 2.33 1.88 134.6

5 2.49 2.56 170.5

6 4.69 1.70 240.0

7 3.91 .925 332.4

8 .512 1.24 342.0

9 2.53 3.24 479.4

A small nonlinearity study was performed on the M256 cannon. This study was performed in only the venicalconfiguration with the excitation forc at the king nut. Data was collkctexl it six excitation levels (0 lb, 4 lb, 6.2 lb, 8 lb,10 lb, 12.5 Ib) from 0 Hz to 320 Hz at seven locations. The first two modes were quite linear. Modes 3 (SO Hz) and 4(182 Hz) manifested some interesting effects (Figure 9). Both of these modes showed two lower wnplitude peaks precedingthe frequency from which the modal parameters were extracted. In the nanlinearity study, these initial peaks shewed a largevariation in amplitude and frequency compared to the tihird (primary) pe &k. This observation is a good indicator that thesystem can exhibit nonlinear behavior, particularly in certain frequern, y ranges. No attempt was made to simulate thisbehavior with the numerical model. Further, it is very likely that the tw( peaks observed are primarily the result of boundaryconditions (interfaces between tube and piston, or piston and cradle) and not the dynamic properties of the M256 tibe itself.

407

WILKERSON, BERMAN, AN[) LI

-- dl I I I I I f

00~ a,4.0o 0 0n~ o

0 1 0.22

l• - Io.02o

if 6-, Y0 0.00

0,61000 .. 4...... 0"

' 023K- I.2.22

3.0 - 0.0 32

- -7 0 4 20.

I4 . I14.20

ay 0. 6 66 - 0. 3 a

o0.4000 . . . . .. [ _

2!

0 I

0 4 1

0.00 50.00 100.00 1500.0n 200.00L Ibn F, q, e*0oy 2022

I :200X. - 00 0 2

0 7 - A 0- 93 - I :1 0 : I

Figure 5. Mode indicator function for vertical frequencies out to 200 Hz.

Sod. Indcator dI t ct Ion

1Y , x 0' 1 $0_ _ _ _ _ _ _ _ _ _ -- . T-r - - - - -1 a

a 2 4 0 . 1

5 x 4 1• .7

4 32 4 3 2 1

33 0.0060

o 44 3 000

20 - 0 0

0.6002 - -L- -- --

i 0 -, 4 4 1. 6y

1. - 0.4322

0 40000-

0. . . . . .. . ... .. . . . .. ... . .. . . ... .. .. .. il. 1 1?0oo.00 000.02 400 00 0o 2o 422 02

Li nwar ri .,io.,L y 2qsl

Figure 6. Mode indicator function for vertical frcqucncics from 2(X) to fi411 it'.

408


M d. I n ic Laor fu n t LIon

I I .- 0 5 1 17

- 40 1 1 24

_ 30Y 0. .1 .

4 13 34i 4 0 - 0 1 4 6

S 34 0 0 00

-0 0 0 - -.-

1 0 0 0442

,0 9 - A p R - 9 3 1i

1 I-I16 1.6

4 _ 4 4 .

24X20 0 .3Z3 I

I 4

0.00 o o0. 1 00.00 .00 20 0 .0 0-- L ,o orl t F rI q ,,v ~c y ( liii

1~~~ 1

: o4? o

10 6 -A PO - 6 3 3 I ] 1 4: ]

_. Figure 7. Modc indicator function for horizontal frequencics out to 200) Hz.

I 0 . 4 I I

t 1 ( - 321 .43

30 0 2. 440 ~ - 426

a

40

0

0.6

30 4 2 90

30 - 0.04)0

...... IFi~~~~~ure~~~~ 8.. Mmc n iao uc nfrh rzo tll- un sfo ()t 640 ' H/. o

49

0 .6000 -- -

..... .... . .. ..

LI,.r ,' 0 li

I 0000. duo?, I

Figure 8. Mod~c indicattor function ior horizonta l rcquciicics from 2(X) to 64() IHI

4 I)9


Frequncy raapon.. a unct I n

h I _ ,o.o 0 0 E

TJ

S. 0 a t- 0 4

101.3 160.00 40 00 101.04

LX nc r 1 c0qu n 0 X 4 u 3 )I I |OOK. 140. I 0 ' 00 0*40. . . . .. . ... 3X •, LOOjX~ 4 0 ,* 3 4 l OOK.• 14.---4*

LOCATION 1I ,roa0Ro L, oC LOCATI 0N 14 /r3ACE L O LOCAT1ON 4 / 0RCE LOC LOCAI10N 1 4 rItCr E LOC

oK ar v co Or r. r loin r r

Figure 9. Nonlinear study, mode indicator function for 0 to 320 Hz.

Figures 10-16 show the numerical and experimental mode shapes for each of the first seven major frequencies obtained.As can be seen in the figures, the early mode shapes (i.e., rigid body mode and the first two flexural frequencies) compaiewith the numerical predictions quite well. However, due to insufficient accessibility in the cradle region of the M256, theshapes of flexural modes 3-5 are not completely resolved in the experiment. Consequently, the actual assumed shape ispenciled into the experimental figures based on the estimations made by the numerical model.

CONCLUSIONS

The confidence in the accuracy of the modal parameters for the first five modes of the vertical configuration and the fusIfour modes of the horizontal configuration is high. There is significantly less confidence in the accuracy of the damping andfrequency values for the higher order modes. In the vertical configuration, several modes appear in the mode indicatorfunction and frequency response functions which were not extracted in the modal analysis. These modes have mode shapeswhich are extremely similar to mode shapes which were extracted. It is believed that these extraneous modes result fromcomplex boundary conditions and are not present in the dynamics of the tube itself, but only the system as a whole (tube,plus its supporting stnrcture).

The current FE tricdel shows reasonable agreement with the experimental results. However, the experimental mode shapesfor flexural modes 4 and 5 must be resolved to determine if the numerical predictions are correct and to verify that there aremultiple flexural mode 4 shapes. This can only be done by increasing the number of sensors on the gun tube inside of the.cradle. This could have been accomplished by attaching the sensors to the inside of the gun tube; however, it was not thoughtof at the tune of the test.

There was also a fair amount of difficulty in numerically predicting both the first rigid body mode and the first flexuralmode. Both modes are dependent on the Tpring constant used to simulate the elevating mechanism. A stiffet sprirng constantmakes it easier to duplicate the rigid body frequency while making the first flexural mode too stiff. Therefore, it was

410


Z x

EXPERIMENTAL RESULTS

ANSYS 4.4A1MAR 18 199310: 45 :0 4

NUMERICAL RESULTS PLOT NO. 2POSTI DISPL.

STEP:1ITR- 2FREQ-10.457DHX -1.201

DSCA-10.183'V 1iDIST-122.287XF -43.53

Figure 10. Compnrisonl of rumrerical and cxierimcmal vertical minoc shapes (rigid Ihdy imoxe).

411

WILKERSON, BERMAN, AND lI


-I.LL L

ANSYS 4.4A1MAR 18 1993

NUMERICAL RESULTS 10:45:18PLOT NO. 3POST1 DISPL.STEPI1!TERP3FREQ-31.411DMX -1.291

DSCA'9.476ZV -1DIST-122.287xF -91.53

Figuie i 1. Comparison of numcrical and experimcntl vertical mfodC shapes (first flexural mode).

412

WILKERSON, BERMAN, AND HI


ANSYS 4.4A1

MAR 18 199310:45:25NUMERICAL RESULTS PLOT NO. 4POSTI DISPL.STEP-IITER-4FREQ 85.74DMX -1.472

DSCA-8.305ZV -1DIST:122.287ILF -93.53

Figure 12. Comparison of numerical and cxperinnenLal vertical mode shapes (second flexural mode).

413

WILKERSON, BERMAN, AND LH



NUMERICAL RESULTS 10:45:36PLOT NO. 5POST1 DISPL.STEP-1ITER-5FREQ-178 .939DHX -1.355

DSCA-9.023zv -DIST-122 .287XT -93.53

Figure 13. Comparison of numerical and expcrimental vertical mode shapes (third flexural mode).

414

WILKERSON, BERMAN, AND L1

x


ANSYS 4.4AIHAR 18 1993

NUMERICAL RESULTS 10:%52:59PLOT NO. 7POSTI DISPL.STEP61.ITER-6FREQ-287 .869DHX -1.371

DSCA-8.92zv =1DIST:122.287Xr .93.53

Figure 14. Comparison of nuLmerical and experimental vertical mode shapes (fourth ilexural mode).

415


x


=L

&NSYS 4.4A1MAR 18 1993

NIUIERICAL RESULTS 10: 53 : 07PLOT NO. 8POSTi DISPL.STEP-IITER-7FREQ-365.945DHX -0.551369

DSCA-22.179ZV -1DIST-122.287XF -93.53

Figure 15. Comparison of numerical and experimental vertical mode shapes (Iburth flexural mode ?).

416

WILKERSON. BERMAN. AND LI

7



NUMERICAL RESULTS 10:53:17PLOT NO. 9POSTI DISPL.STEP-:ITER-eFREQ-.455.324DXX -1.451

DSCA-8.425ZV -1DIST:122.287XF -9".53

Figure 16. Comparison of numerical and experancntal vertical mode shapes (fifth flexurai mode ?).

417


speculated that the truiinnons were exhibiting sonie resistance to rotation and a rotational spring was added at that location,which improved both predictions.

Finally, some of the parts that were not originally represented in the modcl need to he included. These parns include thethermal shrouds, bore sight, mantelet, and trunnion mounts. Future models will incorporate these pwILs and hopefully betterrepresent the whole system.

REFERENCES

1. Wilkerson, S. A., ct al. "Static Load Test on the M256 System." In print March 1993.

2. Rowekamp, B. J. "120-mm Tank Main Armament System Test Acord; Breech Mechanism InterchangeabilityDocumentation for Hybrid Cannon." Benet l•aboratories, Watervliet, NY, August 1987.

418

ad#: d - apps.dtic.mil · component part notice this paper is a component part of the following...

Documents