-A163 909 INTERFRANE DUCKLING OF ALUMINIUM ALLOY … · the yield behaviour in compression, but a total reduction of 4% [ has been been included allowed in for the hold values time

-A163 909 INTERFRANE DUCKLING OF ALUMINIUM ALLOY STIFFENED v/1PLATING(U) ADMIRALTY MARINN TECHNOLOGY ESTALISHNNNTDUNFERMLINE (SCOTLAND) J D CLARKE ET AL. OCT 65

UNCLASSIFIED ANTE(S)/R85iG4 DRIC-OR-97439 F/6 1116 ML

nlu 1. Ij3..,

MICROCOPY RESOLUTION TEST CHART "INATIONAI BUREAU Or STANDARD, IQ6 3-,A '

*1

p..-

-.

t:

" ij

" " " ' " ... " ";' " "" ' " " -" "- -'-'"""".'i i,' - .' ii-'i .'' -.. / '- "" "" " ...t

" ". , .,. ' , - .i ' : .i . '. . .- : ' - -.- "- . -..E.-. ,., . .. . . ...- .-M . -.-. .: .. i

.- '.- '. ';';L.'-,_'_.'._,'_,'-'.."."-'_ ,, --.-..-.-.. '.: .'.L i._'... ' --- -,' -.-.- -' '.-'-. . . '.-'_ -1z.L 1-36-':. "_.'H'_."-

-. .. -.- * .. . , . - " "_ __ % . _;& .-.1_: . ,-

UNLIMITED S B 749-li

o. e b -Sq

REPORT AmTE(W R 8 5 104

C iCOPY No 37.. ADMIRALTY RESEARCH ESTABLISHMENT

.. ARINE TECHNOLOGY)

I-I

INTERFRAME BUCKLING OF ALUMINIUM ALLOY

STIFFENED PLATING (U)

J. 0. ClIark eJ. W. Swan

Q"TIFE iO

fi fiLE COPYWA R E IMunfermlinel

St. Leonard's HillDUNFERMLINE Fife KY11 5PW

962 6 121OCTOB ER 1985

UNLIMITED

- 2" -.. --- ,--

~ ~-_'. "°"

* - ..-. ;. b9 UNLIMITED

A"TE(S) R85104

INTERFRAME BUCKLING OF ALUMINIUM ALLOY

STIFFENED PLATING (UL)

BY

J D CLARKE AND J N SWAN

Summary (ULL):

Compression tests on five stiffened N8 aluminium alloy platescovering typical full scale warship scantlings are described. Thepanels were manufactured using normal shipyard production methodsto ensure typical initial distortion and residual stresses.

Maximum loads and post-buckling load/deflection results arecompared with theoretical predictions using the ARE computerprogram Nl06C.

AcceSsion For

Avilbiit _oe L:,.-"' S

ARE Dunfermline

F St Leonard's Hill -

..-...,.., DUNFE RI]NE Fife K,11 5PE TWC

CopyDright "i"

Controller HMS0 London "-E1985

I.-

- 1 - Octoberm9e ZDUNFMUNLIMITED

".

TET

.MITED

"' "T I T L E ... ... ... ... ... ... ... ...I .

DISTRIBUTION 3

INTRODUCTION .. ... ... ... ... ... ... 4

SEXPERIMENTAL DETAILS AND PROCEDURE 4-7

Details of Test Grillages .- 5... ... 4-5

Yield Stress Measurements 5-6

Residual Stress and Distortion Measurements 6

Loading Rig and Instrumentation 6-7

Test Procedure .. .. .. .. .. 7

RESULTS 7-9 -

Load/Deflection Behaviour ... ... ... 7-8 .

Plate Deformation 9 F

Plate Stresses ... 9

COMPARISON WITH THEORETICAL PREDICTIONS 9-10

RECOMMENDATIONS FOR FURTHER WORK ... ... ... 10

" CONCLUSIONS 11

ACKNOWLEDGEMENTS ... ... ... ... ... ... 11

REFERENCES 1-10 12

TABLES 1-5e ... .. .. ... ... 13-26

FIGURES 1-33

IF APPENDIX A ... ... ... ... ... ... ... 27-28

ABSTRACT CARDS

-2-

UNLIMITED -

7777777 777 7

UNL IMITED

Copy No i

1 MD(MTr), ARE Teddington

2-6 CNA, Bath (Attention Mr D W Chalmers)Copies 4-6 f or onward transmission to NAVSEA,DTNSRDC and DREA under IEP ABC36)

7 DSc(SEA) (Attention AD2), Main Building, London

8 CS(R)2e(Navy), Hayes, Middlesex

9 -20 DRIC, Orpington

21 SCNO, Washington

22 D(S), ARE Dunfermnline

23-27 ARE Internal Loan (Dr Smith, Mr Gass, Mr Dow,Mr Clarke, Mr Swan)

28-30 ARE Spares

-3

UNLIMITED

VL% UkhLIM"

INTERFRAME BUCKLING OF ALUMINIUM ALLOYSTIFFENED PLATING

INTRODUCTION

1. Although the only recent major use of aluminium alloys inBritish warship hulls has been the superstructure of the Type 21frigates, they have potential applications for advanced navalvessels eg SKATHs, hydrofoils and Surface Effect Ships. For this

reason it was considered that a series of measurements and testsshould be conducted to raise our understanding of the factorsaffecting the buckling of stiffened aluminium plating nearer toour current understanding of steel structures.

2. Four panels 3.7 m x 2.7 m with different plate and stiffenersizes were ordered from Yarrow Shipbuilders Ltd. Normal ship-building practice was requested for welding to ensure typicalimperfections. During construction extensive measurements toobtain in-built residual stresses were taken by ARE and aftermanufacture the panels were surveyed to obtain plate and stiffenerdistortion (M)A. Each panel was then cut into two and end andside supports were welded on to give eight compression test pspecimens each having three longitudinals and two full framespaces. Compression tests have now been carried out on five ofthese and the results are presented here and compared withtheoretical estimates.

3. The tests were similar to those previously reported forstiffened steel structures (2) except that the sides of thespecimens were supported only at the frames so that the failuremode was interframe buckling, and secondly the applied loading wasdisplacement controlled so that the post-buckling behaviour couldbe monitored. A secondary objective of the tests was to check thetheoretical estimates of post-buckling behaviour calculated by theARE program N106C (3).

4. In parallel with these tests theoretical and experimentalwork was carried out on the development of weld induced residualstresses and distortion at Cambridge University (4) and on platebuckling at Imperial College (5) and Cambridge University (6,7).

EXPERIMENTAL DETAILS AND PROCEDURE

Details of Test Grillaes

5. The stiffened panels from which the test grillages were takenwere fabricated in NS (5083) aluminium alloy. The first threepanels had identical longitudinal stiffeners (38 x 76 mm x1.65 kg/m T bars) with nominal plate thicknesses of 5, 8 and12 mm. The fourth panel had 8 mm thick plating with 78 x 8 mmflat bar longitudinal stiffeners. In all cases the transverseframes were 64 x 127 mm x 3.85 kg/m T bars at a spacing of 1000mm.

*( ) References on Page 12

-4

UNLIMITED

UNLIMIL £2

6. The panels were divided into two test grillages, each withdimensions as shown in Figure 1. This subdivision resulted ineach grillage containing two different plates joined by alongitudinal butt weld offset from the centreline. The ends ofthe grillages were reinforced with doubler plates to avoid localfailures and steel beams were used to distribute the load from thefour jacks at one end and the six load cells at the other. Thewidth of the plating along the outer edges was chosen to have anelastic buckling stress with one edge simply supported equal tothat for a full width panel simply supported on both edges. Underelastic interframe buckling the grillage can therefore be con- .sidered to represent the combined behaviour of three longitudinalstiffener/plate sections. In the absence of buckling, ie foruniform elastic or plastic stresses the total cross sectional arearepresents approximately 2.75 stiffener/plate combinations. An

::"estimated allowance for this difference has been made below when.comparing measurements with theoretical predictions. The error inthis estimate is thought to be small, but can be checked later

• using three dimensional elasto-plastic finite element analysis if*. considered necessary.

7. Average grillage dimensions and material properties aresummarised in Table 1, and Table 2 gives plate slendernessparameters and upper limits to the failure loads based on platebuckling with no allowance for stiffener deformations.

Yield Stress Measurements

8. Tensile stress/strain measurements were made on at least sixspecimens from each plate and group of stiffeners. In order toestablish the effect of strain rate and hold time a number ofthese were tested according to the recommendations of the DOE-TRRLPanel for Testing Procedures for Steel Models (8), ie at a strainrate of 300 pc/min with a hold of 2 mins at a strain of 0.005.Since it was found that testing at 3000 PE/min gave very similarresults most of the testing was carried out at this rate. Theaverage reduction in 0.2% proof stress at 300 pV/min below that at3000 p0/min was 2%. There were further load reductions during the2 min hold of 2.0% at a strain of 0.005 and 2.6% at 0.01.

9. Attempts were made to measure the compressive yield but thesewere not satisfactory because of buckling effects. CambridgeUniversity (6) have developed a technique using small coupons withPTFE anti-buckling supports. They report compressive values forN8 aluminium alloy but there are no corresponding tensile values.

*. In steel, comparisons (8) have given static yield values 3% higherin compression but the CP118 minimum specified proof stresses forN8 aluminium alloy are 5.6% lower in compression. In the absenceof any other information it has been assumed in the theoreticalanalysis reported below that there is no significant difference inthe yield behaviour in compression, but a total reduction of 4%has been allowed for hold time effects. This reduction has not[ been included in the values given in Table 1 or in the "stress/strain curves for the stiffeners shown in Figures 2 to 4.

It can be seen from the proof stress values in Table 1 that thevariation from plate to plate is much greater than the hold time

-5-

UNLIMITED "

"- -

UNLIMITED

correction. Within each plate the results are fairly consistentbut results for the stiffeners showed a large scatter as can beseen from Figures 2 to 4. In fact the average proof stress of87 MN/m 2 for the flat bar stiffener is below the minimumspecification value of 125 MN/m 2 for N8 alloy (9).

Residual Stress and Distortion Measurements

10. Full details of the measurements on the original panels aregiven in (1). Residual stresses were determined from accurate7 :- -- distance measurements between indents on the plate surface before

and after welcing. Average and maximum values are given inTable 1. The stiffeners all exhibited a combination of direct andbending stress but there was no systematic correlation between the

.'., magnitude of the bending stress and the stiffener distortion.

11. Plate deformations are given in Table 1 and stiffenerdistortions in Table 3. The stiffener distortions were measuredusing three different methods. The original panels were surveyedwith a displacement transducer mounted on a bar, and measurementswere also made using a swept laser beam as a reference plane asreported in (1). As a further check some measurements were alsomade after the test grillages had been manufactured using thelaser deflection method described in (10). All three sets ofvalues, where available, are listed in Table 3. It can be seenthat there are quite large differences in the results althoughthey show similar trends. The initial measurements were all madeon the plate side and some of the variation could be due to localdistortions in the plate surface caused by welding. The lasermeasurements on the final grillages were made on both the plate-side and on the stiffener flanges. The results again show the ..

same trends with differences of typically 0.5 mm between measure-ments on either side. An exceptionally large difference wasrecorded for stiffener 6 of grillage 4A which also has the largestdistortion measured on either side. In this ca'e there may havebeen a significant difference in the weld gap during fabrication. 4'

12. Table 3 also gives the values assumed in the theoreticalanalysis. Any apparently inconsistent values have been eliminatedbefore taking averages.

Loading Rig and Instrumentation

13. The test frame is illustrated in Figure 5. The grillageswere loaded by four 500 kN Dartec servohydraulic jacks operated indisplacement mode. These incorporate load cells, but as a checkon the longitudinal loading at the supports the load was alsomeasured at the reaction end using six load cells. The supportsat the frame ends were double bottle screws connected to barswhich enabled the top and bottom connections to slide longitudi- Lnally under load. The upper supports can be seen in Figure 6,similar supports were attached underneath.

14. For grillage 3A theoretical estimates of the failure load(Figure 30) suggested that this might exceed the 2000 kNavailable, and two 250 kN pressure controlled jacks were added to

.- ...

'[ ~UNLIMITE ="

37- 17~' -7 7- Nu. - I-- * -- VX*... - -

4 UNLIMIT -

the rig for this test. When the load on the servohydraulic jacks.,- reached 600 kN these jacks were switched in. Although the failure r

load did not exceed 2000 kN they were useful as it was found thatthe pump pressure was low and the servohydraulic jacks could only

- provide a total load of 1680 kN.

". 15. The strain in the plating and longitudinals was monitoredusing foil resistance strain gauges distributed as shown inFigure 7. These were recorded via a Peekel strain gauge loggerfor subsequent computer analysis. Plate and stiffenerdeflections were monitored by ARE deflection transducers arrangedas shown in Figure 8. These were supported on a Dexion datumframe which has been removed in Figure 6 but can be seen in Figure .9a. The deflection signals were also recorded by the Peekel .logger.

16. An additional measure of plate deflection was obtained from atransducer which was scanned along the centreline of the plate

*' panels on the unwelded side (see track in Figure 8). The outputwas recorded on an X-Y plotter and subsequently digitised toproduce scaled plots.

Test Procedure

17. The height of the test grillages within the rig was adjustedwith the bottle screws until the centreline of the jacks and loadcells coincided with the calculated position of the neutral axis.The flexure plates were than clamped in position.

18. The load was applied by the four jacks operating together indisplacement mode so that the post buckling behaviour could bedetermined. At a number of points over the load/deflection curvethe strains and deflections were logged. The deflection was heldfor at least three minutes before logging to minimise short termcreep effects. At a number of load increments prior to bucklingthe load was returned to zero and the strains were logged so thatstresses could be determined from the elastic relaxation. Theplate deflection was scanned at each load point using the trans-ducer mentioned above.

.. 19. At the higher deflections care had to be taken to ensure thatthe grillage did not move the datum frame, and deflection trans-ducers which were nearing their limiting range were removed orreset. The tests were continued until the overall behaviour couldno longer be accurately monitored.

RESULTS

Load/Deflection Behaviour

20. The loading sequence and average deflections for eachgrillage are given in Tables 4a to 4e. The loads quoted are theaverage of the values given by the load cells and the jacks. Themaximum differences between this average and the jack load were2.6% at the maximum load and 3.8% at the maximum deflection. Thelongitudinal deflections quoted are the average values of the

-7-

UNLIMITED

- - - - - - - ........-* ..... .... . . . . . .

UNLIN

differences between the appropriate four pairs of transducersillustrated in Figure 8. 61 is the length change over the framespace nearest to the jacks, 62iS the corresponding value for theother frame space, and 63 is the change in overall length of thegrillage. In each case a contraction is shown as negative. As ingeneral the measuring points did not coincide with the neutralaxis a correction has been applied to allow for the effect ofrotation of the transverse frames.- The method used is describedin Appendix A...- :.

21. w, and w2 are average values of the vertical deflections ofthe three stiffeners in each frame space relative to their ends(positive upwards). The deflections of each stiffener are listedseparately in Tables 5a to 5e. The values for stiffenersdeflecting downwards are less accurate for large deflectionsbecause of the effect of stiffener tripping, which tended to occurnear the centre of the frame space, ie near the location of thevertical deflection transducer on the table.

22. The permanent deformation after the final load is illustratedin Figures 9 to 11. In many cases the maximum deflection did notcoincide with the location of the deflection transducer so theresults have not been presented here. In all cases the trippingonly became noticeable in the post-buckling region and did notappear to influence the maximum load.

23. Figure 12 shows the load/deflection curves across both framespaces for each grillage, and Figures 13 to 17 show the values foreach frame space separately. For grillage 3A there was a suddenjump in deflection just past the maximum load and the probablepath is shown dotted. In addition to the measurements taken afterthree minute holds, continuous load/overall grillage contractionplots were taken as an indication of the behaviour of the grillageduring the test. For other grillages the load scale was theoutput from one jack only and the results are not included here.For grillage 3A, which was the last grillage tested, the jacksignals were summed to give total load and the trace for the finalloading is illustrated in Figure 18. It can be seen that near themaximum load the reduction during the three minute hold isconsiderable. The dynamic jump does not occur at constantgrillage displacement because the frame acts as a spring and thereis an additional contraction as the stiffness of the grillagereduces. The stiffness of the frame can be determined from thedifference between the jack displacement and the overall grillagedisplacement. Results from all five tests fall within the bandshown in Figure 19. The non-linearity is thought to be partly dueto the initial closure of gaps, eg bolt clearances, in the rig.This plot can be used to define the limiting post-bucklingstiffness which can be controlled in the frame.

24. It can be seen from Tables 4a and 5a that the stiffeners ingrillage 1A did not behave similarly. In one frame space theouter stiffeners deformed vertically in opposite directionsproducing a considerable amount of twist in the structure. It wa7decided therefore to test the second grillage fabricated fromYarrows panel 1 as well. This grillage (IB) behaved much moreuniformly, but as can be seen from Figure 12 the load/shorteningcurves were very similar.

-8-T

'[.iUNJLIMITED

- . .. -'." "-

. -- .- .) - . . .J . . . .li .

- J ... Hm... ... i . .. !..... . -- . . *[- .i .

UNLIMITED

Plate Deformation

25. Figures 20 to 23 show the plate deflection traces measuredalong the centreline of the plate panels on the unwelded side (seeFigure 8). It was not possible to obtain a complete trace at someof the higher loads because the deflection transducer was out ofrange. In some cases therefore the permanent deformation showncorresponds to higher load values than for the deformation shown

* under load. Tables 4b to 4e give the load levels corresponding tothe scan numbers.

26. A large part of the deflection, particularly in the case of I-

grillage 3A which has the thickest plate, is due to the stiffenerdistortion. It can be seen that the plating generally bucklesinto three half waves and that the initial deformation, particu-larly for grillage 1B has a three half wave component.

Plate Stresses

27. The longitudinal stresses derived from the elastic strainrelaxation is shown in Figures 24 to 26. Account was taken oftransverse strains using a Poisson's ratio of 0.3. In the platethe values are the average of the top and bottom gauges. At thestiffener positions the values were interpolated to give thestress at the centre of the plate.

28. The reason for the low values on one gauge for grillage 3A isnot known but could be due to partial failure of the adhesive.The average plate stresses have been derived by integrating thesevalues and are plotted against edge strain as a fraction of yieldvalues in Figure 27. These plots have been used to check the Lvalidity of the stress/strain curves assumed in the theoreticalanalysis discussed below.

COMPARISON WITH THEORETICAL PREDICTIONS

29. A number of calculations have been carried out using the _computer program N106C (3) to give an indication of the accuracyof this program in predicting maximum loads and post-bucklingbehaviour. A single stiffener and associated plating wasrepresented and symmetric behaviour was assumed beyond the midf;-ame positions. Each half frame length was subdivided into tenelements each with twenty fibres representing the stiffener andone variable width fibre representing the plating. Since measuredresidual stresses in the stiffeners were low and variable in signthey were ignored in this analysis.

30. Following the recommendation of Dier and Dowling (6) andMoflin and Dwight (6) it was assumed that the non dimensionalisedplate behaviour could be represented by that for steel with the

same a value 8 = b/t ,rJ. Appropriate curves were therefore

selected (via the parameter NEW in N106) which corresponded to ab/t value which for steel would give the same 8. This was notpossible for grillage lB since the equivalent b/t in steel (160) F

-9-

UNLIMITED

• _ ~ p .-. - - . - *.-----. ..a . , - , : - - -. ---. --- _

P7 w- -7 -7- 17 1.--I-T T-......

UNL7

was outside the range of values incorporated in the program. Fromthe residual stress and distortion measurements (1) it appearedthat the appropriate curves would be between the "nearly perfect"and "moderate" sets given in (3). Since the effective b/t valuesdid not coincide exactly with the values available calculationswere carried out for two NEW values which bracketed the range, ie.nearly perfect", low b/t and "moderate", high b/t. For grillagelB the best approximation available was to use the b/t = 90 curve vfor "severe" imperfections. This was slightly above the curveshown in Figure 27. Calculations were also carried out for thisgrillage using the Faulkner formula (NEW = 3). In each casecalculations were made using the average, and maximum vertical rstiffener distortions given in Table 3. The average values areprobably more representative as there will be an interaction inthe buckling behaviour of adjacent stiffeners.

31. Results of the calculations are compared with the measuredload/shortening behaviour in Figures 28 to 31. Except in the caseof grillage 3A where the calculated load/displacement curve has asharp peak, the estimates of maximum load bracket the measuredvalues. In each case however the post-buckling load carryingcapability is underestimated. One contribution to this differenceis the rotational restraint at the ends of the transverse frames -provided by the supports. The rotational stiffness of thesesupports was measured after the tests by applying a known momentand results are given in Figure 32. Over the initial part of therange the plot is linear with a stiffness of 2.0 x 107 Nmm/radianwhich corresponds to 1.4 x 107 Nmm/radian per stiffener. Figure 33shows the effect of rotational constraint on the bucklingbehaviour of grillage lB (NEW = 31, Average stiffener distortion).The measured value of 1.4 x 107 Nmm/radian gives a significantincrease in the post-buckling strength, but a value ofapproximately 5 x 107 Nmm/radian would be needed to match theobserved behaviour. A possible explanation is the effect of theshort end frames which are reinforced with doubler plates. Theseare not represented by the simple two half span model used forthese comparisons.

RECOMMENDATIONS FOR FURTHER WORK

32. It is important to establish the reason for the difference :nthe predicted and measured post-buckling behaviour. Some N101Ccalculations should therefore be carried out using a four spanrepresentation which includes the effect of the stiffer endsections. The effect of residual stresses in the stiffenersshould also be investigated to see if these are a possible rea:oenfor the over-estimate of the peak load for grillage 3A.

33. There are three remaining test grillages 2B, 3B and 4.These have welded edge supports but have not been strain kLauged.Consideration will be given to testing these with fully rotationalframe supports to eliminate this uncertainty in the comr.atincn.Attention will also be given to the accurate measurements oflongitudinal displacements, but it should be possible to nimplifythe instrumentation by omitting the strain measurements and someof the deflection measurements.

- 10

UNLIMITi D

___

t- _r '_ ,:_..................................."....-...","."....,...-..."."..-".--..-..'... ... . .' .

UNLIMITED

Nj.

CONCLUSIONS

34. One of the most significant facts to emerge from these testsis the variation in yield strength in material which is supposed Lto satisfy the N8 specification. If aluminium alloy is used in %future ships in critical locations where strength is important itwill be necessary to have adequate quality control.

35. It appears that in most cases the maximum strength ofaluminium alloy structures can be calculated using N106C with theplating behaviour represented by an equivalent steel b/t valuegiving the same B. It should be noted that the peak load forsharply varying load/deflection curves may be overestimated.

36. The post-buckling strength measured during the test washigher than that predicted by the simple two half frame finiteelement model used. This was partly due to the rotationalrestraint at the frame ends but could also be due to theadditional rotational restraint at the ends provided by the shortsections with doubler plates. It should be noted that sincedesign calculations are usually carried out using the two halfframe representation, there will be a similar underestimation ofthe post-buckling strength for stiffeners close to hard cornerswhere the frame rotation is constrained and in cases where thecollapse only occurs over one or two frame spaces.

37. In future tests care should be taken to minimise errors inlongitudinal displacements caused by frame rotation.

ACKNOWLEDGEMENTS

38. The authors would like to acknowledge the assistance of manyARE personnel in conducting these tests. In particularMr Somerville for the procurement of the panels and measurement ofimperfections, Mr Penman and Mr Aitcheson for tensile testing,distortion measurements, instrumentation and assistance with theanalyzis, Mr Hugill for some of the N106 calculations, andMr Adamson and Mr Duncan for carefully setting up the grillagesand their instrumentation in the test frame.

U- II -

UNLIMITED

.. . . . . ......._ ,...,:.. ,... _.._.. ...- , .- ... -:¢~'¢- :¢'LJ¢i-

UNLIMITED

REFERENCES

.. 1. Somerville W L. 'Welding Stresses and Distortions in N8(5083) Aluminium Alloy Stiffened Plates'. AMTE(S) ReportR85103, To be published.

2. Smith C S. 'Compressive Strength of Welded Steel ShipGrillages'. NCRE Report R611, May 1975.

3. Dow R S. 'Nl06C: A Computer Program for Elasto-Plastic,Large Deflection Buckling and Post-Buckling Behaviour ofPlane Frames and Stiffened Panels'. AMTE(S) R80726,July 1980.

4. Wong M P. 'Theoretical Analysis of Longitudinal ResidualStresses and Distortions in Welded Plates'. CambridgeUniversity Engineering Department Report NoCUED/D-Struct/TR86. 1981.

5. Dier A F and Dowling P J. 'Aluminium Plated Structures -

Final Report for AMTE'. Imperial College, Department ofCivil Engineering Report CESLIC-APS2, July 1981.

6. Moflin D S and Dwight J B. 'Tests on Individual AluminiumPlates under In-Plane Compression'. Cambridge UniversityED/D-Struct/TRl00, 1983.

7. Moflin D S. 'A Finite Strip Method for the Collapse Analysisof Compressed Plates and Plate Assemblages'. CambridgeUniversity ED/D-Struct/TRI01, 1983.

8. Recommended Standard Practices for Structural Testing ofSteel Models, TRRL Supplementary Report 254, 1977.

9. British Standard Code of Practice CPl18, The Structural Useof Aluminium.

10. Brown J C. 'A Laser Based Distortion Measuring System'. LAMTE(S) TM82450, August 1982.

" 12 -

UNL IMITED

. .- .. . -.

UNLIMITED

' + + .

L. ... v I r

-4-

0 Co. .'PJ-'-

WC4

o j NNLf 'U c

GJ-- -' N N N

-~a All~t N .

0 u- .

(u e

a. tjI..g

-4 CU Ve)& - W r I

L)L

U).~ -40 A

_____in 4.) M

e La0.

<~~ N Oc Wr1I

Kc=i '0 (N ~ ~ ~ ~ L )n --.)r-L 0 coo

0 00 a)C7r r D

LL.

4. Ij U.NL j .1 4j .E L

to 14 M5 W" cm-. m r% 5- - es - ,

-. ~~~~~ ~ >- > **-*%.;'~.*. *I*** ~ V~~.~ '--.. *.

UL IIMI TED

4. . 1 0 0o 0 0 0SU Z LA -t - .4 -1

m ~ ~ d mr4(n -- r.

4 0-

*. 14- rI 1-4 Vi 4

fd ~ ~ -#d It N4--

:1 0 % G N -

o ( t co4 N co

.4J 0 0 0) 0 0 -

44 N LA o

00

f.

~r Ca.

Q) 0 IN N c Crw % 0 0r- r o Lo

21, ci '!W.

~~T7 N - L

L L

C) -4 'oC)C

I - Lrn C14 n)

ufd -C. Il I) c

L.J

..- Fn .

14N N

~* - ~ .-4UNLIMITED'

N 077--

w..

+ + +

U3 En u.

11 - e IN n "-i 0 a 4N p

-i N14 . 1N 0 m m r- r -40

+ + + + +

1-4 43

IZ ) 4 0.4-4 0 0 o.-40 0 00-04.4-400000

P -. 4 P

lz 0 .+ II

- , UG, Co . . . . . . . . . . . . .U) Gu4 0 ,-4 0 -4000 --0C)0 0 e N0 -00 00

o +4. 0+- ++ + +-4+ +1 1+1

E-4 o2.- E-44E-44r c n, % r -4

H 4>0 - - I_

+ +. + + + + i+i + +1+

410z r1-N4y rmm - N 0 i .n m w . 0 0 ,m w q m.- N - 14 N m

44441+ 1 + + + 4 1

u 0r-4 N ~ I 00 00 00 0 N0

.)

IF W 17o u

Ci14

'15UUNLIMIT-D

0I4i 440-

I L .UNIMITED

-'11~to

N U)44.

CflL 04

~004

0)0

4o 4

4.)

01-4 + 11

*~ EsE- 0'C14 0. N 0

44)

U

0d oU41 0 s

40) 4. 1

1-4 e 44:. w .

0i >

-4 4 Utr-L .-.- .)

cn ~ U w 1 L) Q

Li t U 0

N 4d 0 ) 44 4

0 tn. C0. 0 U C.

044~ 01 $- a) (01 9c. 0 0

0 ~ ~ (dL ..

01 ~ '4-4 4) >

N ~ Li 4) 1 0

UNLIMITE

UNLIMITED

*~~~V -H4

Cuv 0u .0-1444.0 -

.0 0 u d lb 4sw -d9d Ln o

C4 r4 U)~

0.. d p 4. . 4)-

1-4 .. 0 04 ..-4

000

44+

0) ~~~- .- 4 0 it4 0- l urOi NO.

E4:Cl L C N~. O

(d L, 1 r4

- ldL

00)

4. -0 1- 0 Cl m ~ c n

9 0) 0 0 4 .-4 4i 4 u l r: C4 '. b

H A N r -0 000O

E4 'V LLzn 0-4 Cl br. ~ 0 A2-9- .

0 M C0

0'~~~~ 0 N- Nn m~ 4 lJ~D~Enl~ L ~ r ~ -r LL

-4~-MI N 0) E-4Im 0 F-4 le , rq m m-4 %

C4 m d. u r, 0 03r- I. W A .

A.q

u m

*. . . . . . . .,. . .

7. D

um

0 I

040

I- I) f '-4 4r- -4

.- 4.4~00M00U~tA~.-4 C~m -4'.4ON0N-m

14 -4

-4 N -4

E-4I I tftfI

t4J 0~0- W 40 M ~ N

00000000, 40.-40N4 NN0NN LALA

lz lu r 0 fN I Iko Il (n10 I N vl i ll %Da 4 i

r-4n0-4r-00d.0 N NO -4N N NLM

oIN l ai l o u)11 o l mom mf r o o

u, O'-4Nr- r-4H r-4 r-4H r N N N N N Nu N

-18-UNLI1MITED0

44

0i 0

4)4A) 0

U 4.)L

C44a)04)

44

03r- I. -

UU~~~ 00-0-4~L

(U0S: H-l r-0 q I) mm r-0N 0 mo o a mr -11 U w

as .0 4 0 . . . . . 40 31- c J%%- 000000 000 c;c;c c Mc;c rNrM q*10 U 4

--- Ia

*~~ '00Oz0

00-4 r-4N m 1- q14r4 1 X Oq'"DW0 0- 300C 4-tMr4C4

+' + - C(

i4-0

5: . .

N N0 00 00 0000 c 0 0 0N,0-imm 0

-4 ~ ~ r- ->-

tU0.

UNLIMITED

++++11411 i-

0 03 OOD-0iNO04~WC~n(

1"-

-0 ' )0 - C)OOC>mr-4O0 0 N NI.O. 4 -t 4 -41

4-).0 -)

Cl) wrm O o04~ N - - 4 tNu2 ylNc oC; OOOOOOOO.-4O.-CO r4OO'-4U) '4NU) C

0 .. 4

0~ L) O

~~-I~~ ~4 ai~0)L

u a

CV (n d'~ ) r-r C 0 N M -dU 0 -W HN Lar

0 20

UNLIMITED

S..w -

UNL.d, rEt

+ ++ + I . .+ . .+ ..+. 4 4 -4 N m

.-4.- N 0 .-0.4 0 M MA 0 N N- 0 10 M 4 C4NK-

Sc d. qT al 03 r N NW a) i CN4 O 0. co N 0 d. N

0 0 0 a~ 0 m c~ OW q N n -ir -4 qd NW 01 o0 N

w (~~~fl 0 000 NMm 4 10~n 0rz 000a0 - 0r4- In N11 1 m 11n1 dIr %D-4

I~~~ +

0-C4.to O.4-,r%0 m 4W0W N - mm

~0

00

0 044:

40 0-

0 0)

21

UNLIMITED

7

mo &

A. . . .~***

UNLIMITED

TABLE 5a

GRILLAGE 1A VERTICAL DEFLECTIONS OF STIFFENERS

+ve upwards

SCAN LOAD CHANNEL NO. (SEE FIGURE8L

NO. (kN) 116 117j 119 1123 1124 1126

Deflection relative to ends (mm)

10 542 -0.3 -0.2 0.0 0.0 -0.2 +0.313 648 +0.3 +0.1 +0.1 +0.1 -0.3 +0.515 722 +0.5 +0.3 -0.1 +0.2 -0.3 +0.717 834 +1.3 +1.2 -0.9 -1.1 -2.3 +0.818 836 +2.2 +2.0 -1.5 -6.5 -6.5 -0.2

-22-

UNLIMITED

. . .

UNLIINI T.

TABLE 5b

GRILLAGE 1B VERTICAL DEFLECTIONS OF STIFFENERS

+ve upwards

SCAN LOAD CHANNEL NO. (SEE FIGURE 8NO. (kN) 116 117 119 123 124 126

Deflections relative to ends (mm) .

1 02 22 +0.05 +0.14 +0.03 +0.11 +0.22 +0.063 62 +0.08 +0.07 +0.03 +0.23 +0.19 -0.054 119 +0.16 +0.06 +0.07 +0.37 +0.31 -0.195 0 +0.04 +0.01 +0.06 +0.02 -0.03 -0.016 172 +0.35 +0.13 +0.10 +0.47 +0.43 -0.377 0 +0.02 0.0 +0.05 +0.02 -0.05 +0.028 295 +0.77 +0.30 +0.07 +0.55 +0.65 -0.359 0 +0.02 -0.02 +0.04 +0.02 -0.07 -0.01

10 422 +1.16 +0.47 +0.22 +0.61 +0.79 -0.2111 0 +0.02 -0.03 +0.03 +0.01 -0.08 -0.0312 543 +1.57 +0.73 +0.38 +0.61 +0.92 -0.1313 0 +0.03 +0.04 +0.02 +0.03 -0.12 -0.0314 653 +2.01 +1.22 +0.71 +0.54 +0.93 +0.0515 0 +0.11 0.0 -0.06 -0.11 -0.19 +0.0516 0 +0.28 +0.01 -0.10 -0.05 -0.13 -0.1217 713 +2.83 +1.93 +0.95 +0.18 +0.66 -0.0218 787 +4.30 +3.39 +1.79 -0.81 +0.13 -0.2519 0 +1.20 +0.93 +0.32 -1.06 -0.81 +0.8520 785 +9.69 +7.33 +4.91 -3.17 -2.60 -2.2121 686 +18.31 +16.70 +14.78 -6.83 -8.10 -6.2822 0 +11.41 +10.36 +8.98 -5.95 -6.16 -2.7523 586 +23.77 +23.01 +21.82 -15.73 -13.74 -11.3824 553 +25.77 +26.62 +26.73 A A -14.2125 490 +32.61 +35.55 +36.08 A A -19.1326 0 +25.40 +28.80 +26.90 -15.50 -13.60 -13.00

A Transducer out of range

-23

UNLIMITED

- z.~ .. . . .. * ... *1

UNLIMITED

TABLE 5c


+ve upwards


Deflections relative to ends (mm)

1 02 54 -0.01 +0.06 +0.06 +0.06 -0.03 +0.053 99 -0.09 +0.05 +0.10 +0.03 +0.05 +0.024 202 -0.03 +0.13 +0.20 +0.09 +0.13 +0.055 297 +0.05 +0.23 +0.24 +0.14 +0.20 +0.176 0 -0.08 +0.13 +0.11 -0.03 +0.20 +0.067 0 -0.15 +0.15 +0.14 0.0 +0.25 +0.078 455 +0.12 +0.35 +0.34 +0.24 +0.35 +0.079 585 +0.22 +0.47 +0.51 +0.34 +0.36 +0.13

' 0 741 +0.38 +0.54 +0.65 +0.52 +0.37 +0.26

11 0 0.0 +0.21 +0.22 -0.01 +0.33 +0.1112 883 +0.68 +0.43 +1.02 +0.84 +0.66 +0.7313 961 +0.95 +0.54 +1.48 +1.24 +0.85 +1.0414 0 +0.23 +0.29 +0.51 +0.31 +0.56 +0.3715 1023 +1.23 +0.75 +1.91 +1.82 +1.14 +1.2716 1063 +1.54 +0.90 +2.74 +2.87 +1.64 +1.6017 0 +0.82 +0.48 +1.69 +1.51 +1.05 +0.8318 1064 +2.50 +2.56 +4.43 +4.62 +2.56 +1.2819 1057 +2.28 +4.90 +8.41 +9.71 +4.62 +1.2320 0 +1.20 +3.08 +5.98 +6.44 +2.96 +0.8821 1034 +5.33 +9.23 +11.77 +10.52 +4.35 +0.0922 1012 +5.51 +12.01 +15.79 +14.19 +5.60 -0.5823 985 +8.67 +16.55 +19.76 +15.08 +5.09 -1.8024 927 +15.36 +22.70 " +13.96 +3.35 -3.4925 836 +22.43 A A +12.26 -0.69 -5.9626 0 +16.21 A A +9.87 +1.31 -2.21


.24 -

S~~~~~UNLIMITED ~'. .

UNLIMITED

TABLE 5d


+ve upwards

SCAN LOAD CHANNEL NO. (SEE FIGURE 8

NO. (kN) 116 117 119 123 124 126


1 0 :42 198 -0.08 -0.33 -0.20 +0.03 -0.21 -0.023 398 -0.14 -0.58 -0.33 +0.04 -0.27 +0.034 600 -0.20 -0.85 -0.44 +0.07 -0.30 +0.115 0 -0.05 -0.08 -0.16 0.0 -0.13 -0.146 996 -0.37 -1.56 -0.79 +0.13 -0.57 +0.197 0 -0.03 -0.11 -0.14 +0.08 -0.13 -0.158 1185 -0.43 -1.97 -1.03 +0.15 -0.72 +0.249 0 +0.01 -0.15 -0.24 +0.13 -0.18 -0.17

10 1377 -0.59 -2.39 -1.34 +0.30 -0.51 +0.6211 0 +0.12 -0.16 -0.21 +0.13 -0.13 -0.0912 1574 -0.85 -3.33 -1,93 +0.58 -0.35 +1.2913 0 +0.03 -0.59 -0.59 +0.28 -0.19 +0.0514 0 +0.08 -0.60 -0.59 +0.28 -0.21 +0.0915 1663 -1.44 -3.91 -2.85 +0.88 -0.11 +1.9616 0 -0.28 -1.38 -1.16 +0.49 -0.12 +0.4017 0 -0.28 -1.51 -1.22 +0.51 -0.18 +0.4518 1685 -10.33 -2.82 A +3.08 +2.32 +6.6019 842 A -4.74 A +12.34 +15.65 +19.9120 808 A A +13.94 +18.01 +22.06 121 648 At A A +15.19 +20.31 +24.97

1 22 0 . A +10.40 +13.57 +15.94

Transducer out of range

- 25 -

UNLIMITED

UNLI..:

TABLE 5e


+ve upwards:..5 .. .. .



1 0 •2 96 -0.11 -0.07 -0.13 +0.15 +0.09 +0.203 166 -0.16 -0.14 -0.08 +0.23 +0.12 +0.354 254 -0.24 -0.19 -0.20 +0.34 +0.09 +0.445 0 +0.05 0.0 -0.20 +0.03 -0.01 +0.056 441 -0.61 -0.39 -0.30 +0.59 +0.15 +0.767 595 -1.30 -1.13 -1.38 +1.02 +0.42 +1.588 0 -0.40 -0.44 -0.87 +0.12 +0.20 +0.479 716 -3.00 -2.99 -2.40 +1.72 +1.12 +2.78

10 757 -12.5 -14.3 -12.2 +8.44 +6.15 +8.7311 0 -10.9 -9.4 -8.2 +4.77 +3.13 +4.2612 0 -10.8 -9.3 -7.2 +4.79 +3.39 +4.2913 572 -13.5 -12.0 -10.3 +7.00 +4.85 +6.8314 659 -13.2 -12.9 -10.9 +7.54 +5.35 +7.5615 0 -11.1 -9.5 -8.2 +4.91 +3.18 +4.2016 723 -16.9 -14.6 -12.1 +8.27 +6.06 +8.3017 619 -24.9 A -21.5 +16.6 +11.9 +12.118 577 A t , +19.0 +14.3 +14.419 535 A A A +21.7 +17.6 +17.620 488 A A A +25.4 +20.6 +20.621 426 A A A +30.5 +26.4 +26.822 386 A A A +33.5 +28.9 +29.323 0 A A A +27.0 +23.3 +22.0


rk:

-- 26 -

4 UNLIMITED-:'.-.-

A °

AILIMITE

r= 4f A 7*

.A

C4 06

Cy, L4U

oC

*i -0

Lii L

0 - JC3

z I u

uLiO CC -iui

Zn UZ_ n u-

Lu.

'Li 0EA I

E V) C

L~uA

L.A u-L U.

r=- Uo -UC.

FIGURE 1

UNLIMITEDAME () 850A... M T. E (S R851

r.. . .. . . . . . . .%. . . . . .. . . .

UNL Ii TED

z

LLJ

zLL.

LL.

L

10

oa LM LL

V)0

cx

L-

z L

-AJ7LU

CCo

C~~) '.

en c

A M 0 S 851U04I0ITE

.1641 .,'.-n

LUJZ ki

U-<LLJ

zz

040

W0

LL

-J

cr

LL Li

L/LM

zC

F0

U1LIMITE S)R851

.~'% r* '. .~ .Sb~~fl Ufl W).1 ~ ~) tJI~r~- v * ~ rJ. -Lai-

ui0

x La

.4L

IAIJ

Li..

Lx-J

C.44

F I UR

UNIIEDA T S)R 5 0

'lk, - li -, -I.

*:HrTIED

I-iI-

0i

LL-J

oczr

-

U..

I.-.

TEST FRAME -AFTER COMPLETION OF GRILLAGE 4A TESTAND REMOVAL OF DATUM FRAME

SL 841011 03

FIGURE 6

UNLIMITED

A MT E (S) R 8 510 4

UNLIMITED

0

I I I L I -

- C3r-L

o.A .j W - ~ 0 - L ~-

LI~~~~i Lo ' I ~ LLL I-- CD CD Z

LaJ- cc C

-~ U I

-- U i

C)) Z ZIr

L i z - -

U - * K

L M- C), ,

coo

cr ~

I --j

F I GU RE 7UNL IMI TEO A MTE (S) R 8510 4

-5 7 T 7 .... '* a..- * .. a

h - . Ile a.v q

CI

C,

Li IJ

z -z

-JJ

LL ujLU J

zz

LCL

L

U-X

Lo I.-

to c A

0. . <

-J 2L c

t-, -MJA

* a) GRILLAGE 1B SHOWING L. M FRAME

b) GRILLAGE lB PERMANENT DEFORMATION AFTER FINAL LOAD

F IG U RE 9r UNLIMITED

ANMT E (S) R 85 10 4

q7' '

GRILLAGE 2A PERMANENT DEFORMATION AFTER FINAL LOAD

b)

SSL 83 1/9

FIGURE 10UNL IMI TED

A M TE (S) R 8 510 4

a) GRILLAGE 3A PERMNEAT DE, .,IATIlN1 AFTER FINAL LOAD

S L 8 4110 5 3

b) GRILLAGE 4A PERMANENT DEFORMATION AFTER FINAL LOAD

- SL 4 132/103

F FIG UR E 11UNLIMITED

A M TE (S) R 85 10 4

-o - -- -.--. r---- -

UNLIMITED

TOTAL CONTRACTION OF TWO CENTRAL FRAMES

2000

LOAD-; (KN)

3A

1500 2A

I

Ft 500

4 A

0 5 10 isFRAME 1-3 CONTRACTION (51 +&2) MM

F I GU RE 12

UNLIMITED A M TE (S) R 8 5104

UNL IMITED

LLi

-I-

Cp Lfn

F RE1

UNLIITEDAM E () R 510

I TED

LMs

:z0E

LiJ

U3

i

LUn

F IE RE 1

UNIIE E ( )R851

-n- - 7

UNL IMI TED

CD,,

LUuLJ

LAV

u-IN

FIU- 1

zN 11TDAMTE()R851

ALIMIV

LA'it

-i-

LJ

-LJU-

LiJn

0 c

FIUR 1

UN M EDA T S)R 5 0

z7

U14L I IITE 2

%EE

LL. L.71-t L

LAJA

LU.

ICI

0 C00 40

F I GUR E 17

U1L IfI ITED AMTE (S) R 85104

U4L IMITED -

GRILLAGE 3A- OVERALL GRILLAGE CONTRACTION

2 000* LOAD

(KN)

-00

1500

0

00

FYAINAC ODNG SQEC

0I 5 0i

1000AL GRLAE CNRCIN03m

FI U E 1

IN.11 E TE( )R851

I777

-. --. L- 34 TIM,-a-* .- - -T- T,--

L uONP.'

STIFFNESS OF FRAME AND JACK SUPPORTS

1 500

LOAD(K N

1000

0 2 4 6 8 10 12

JACK DEFLECTION MINUS OVERALL GRILLAGE CONTRACTION (mm)

FIGURE 19

UNJLIMITED A MT E (S) R 85 10 4

-7 .

UNLIMITED 5

GRILLAGE 18 PLATE DEFLECTION(ALONG CENTRELINE ON UNWELDED SIDE)

(RELATIVE TO PLATE ENDS)

0- INITIAL DEFORMATION

C3 CV

ID 0

oL 0

-LJ

-3 0

Lu 0:

S 0

* 4''- ~0- DEFORMATION UNDER LOAD SCAN Nos S

L- CS

C3 0

uCLu

LAJ miTRANSDuCERS OUT OF RANGE rfOt SUBSEQUENT SCANS

S 0S 0

Lis

FIUR 20

,*L 0 L- : .*&c - .

UNLIMITED ,'

-.. . ,

GRILLAGE 2R PLATE DEFLECTION,-... (ALONG CENTRELINE ON UNWELDEO SIDE)

(RELRTIVE TO PLATE ENDS)

0 INITIAL DEFORMATIONz (r)

a-

0 0

S0-J

rw 0

z" 0

0. ;' - ; : DFOIWT ION UNDER LORD SCN NOB ..."5 " Z (" " '

- IC

ccs

Lii

.-

Li c

-0

>:] (Y) TRAlNSDUCERS OuT OF RANGE FOR SuOsEOU(NT SCANS "'

Li.

-3

U I ERMA TION AMDER LOR104

°-"• Z (I) $CA NOS 5

L -

- cc= CMT (S is81/,'

" - - " - , - - ,f- ' -- - ' "., ..

4 . .

.JLILIMI

GRVILLAGE 3A PLATE DEFLECTION(ALONG CENTRELINE ON UNWELDED SIDE)


0O INITIAL DEFORMATION

to-

C

ic-

C 0

1-- 0

0i 0

2: 0

0:

- 12

o) 0

U3 0

-LJ

TRANSDUCERS OUT OF RANGE FOR SUBSEQUENT SCANS

i~ 0

0- PERMANENT DEFORMATION SCAN NOS3 11z -is

I-

o3 0

UNIMTE

UNLIITE AMTE (S) R 8 51 0 4

*,** *** ,...- *** ~ ** *** ... .* * * * * .. * . *~ *~ ** * %

a UNLIMITED

GRILLAGE 4JR PLATE DEFLECTION(ALONG CENTRELINE ON UNNF'.DED SIDE)


IIcc

E 0

z- CD(TA EFRAIN.*LJS0

oi 0

MI

DEOMTO NESA O

ccI 0

C.4.,I

LL )i

Lu c

TRANSDUCERS OUT OF RANGE FOR SUBSEQUENT SCANS

S 0

0- PERMANENT DEFOR AT ION SCAN NOS 8

O 23

c-

Li

I Lu 0,. CD

FIG6U RE 23

UNLIMITED

AM TE (S) RB5104

* VWW -: -7.7 -

UNLIMITED

S r-4M.......

x EIE4+X*44(N

W

cr

-k-

w IL

w. z

UU

z

(W*OS/NW) SS3U.LS ]AIS9J3idWO3

cr.- Aa r

C3Czz

0

cc wE~ Ccl~fcc,

00-0 0 ''1 0 U-0 21 006 0 -S o -E 0 "

it UC/W SRS3IS3dO

F I RE2

UNIMTE

A SrR651

zA

L ALIM ITED

'-4 * zzzzzzz

0

x E3EA4+X*44(Fi

0

0 A

x 0L

Li

cr

cn CI C

z .

C)

a: w

cca:

jLL

z

ooor~ ooo~; (H*CS/NW) SS3UiS3AISS3UdW3OOd O

F I G U R E 2

UNLIMITEDz

rr

UNLIMITED

zzzzzzz

19 )-4n

x EMX+X*+0

0z

- r

(3 (n(ccU

wj x (-~.cccc c

_j .z

-j u.c

xDI:

Mr

_j

oo-so' 1 od 0036 0Os'L 00d 09 0.h oodoc 00-, Colo(WOS/NW) SUiS 3AISS3UdWO3

,~w .,t

L3L

0E)4C*

zcc

cr

P-

uj

cr cc

~cr

zC

U- %

'-sd OU. 0 0-, o-S 0 e o .i c

;W'O/NW SSUS2,G3dO

FIUR 26

UNIIE

AME()R8 0

-A0 -

AVERAGE PLATE STRESS / EDGE STRAIN

0-7

0-6

3A (':37-0)

Obit2A (bt 62-0

0-40

0*3/

0/ L/t9

0-2

0 0-2 0.4 0-6 0.8 1.0 1-2

-. e

F I GU RE 27

UNL rIITED AM TE (S) R 8510 4

* . UNL 1111I TED

GRILLAGE l B -COMPARISON OF LUAD ISHORTENINGWITH THEORY MN106)

800

LOAD3(KN)

600

400-

-0- EXPERIMENTAL VALUES

200 THEORETICAL CURVES1 NEW= 3 (FAULKNER) AVERAGE DISTORTION

2 NEW= 3 MAXIMUM DISTORTION

3 NEW= 31 AVERAGE DISTORTION4 NEW: 31 MAXIMUM DISTORTION A

0 10 1FRAME 1-3 CONTRACTION (mm)

FIGURE 28

UNLItUTED ANMT E (S) R 8510 4

- _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __77

UNLI111TED

GRILLAGE 2A- COMPARISON OF LOAD I SHORTENINGWITH THEORY (N106)

2a

1000

-44

800

600

400

-0- EXPERIMENTAL VALUES

rE THEORETICAL CURVES

2001 NEW= 16 AVERAGE STIFFENER DISTORTION2 NEW =16 MAXIMUM STIFFENER DISTORTION

3 NEW =24 AVERAGE STI FF!NER DISTORTI ON

4 NEW = 24 MAXIMUM STIFFENER DISTORTION

0 51015r-

FRAME 1-3 CONTRACTION

FIGURE 29

UI~L II IIT ED A M T E (S) R 85 10 4

GRILLAGE 3A COMPAR.jii ii. LOA4 upri~oootWITH THEORY (N106)

2 PEAK :2113KN

32000

4LOADIKN)

1500

0

0-*

500-394-0- EXPERIMENTAL VALUES

THEORETICAL CURVES

1 NEW = 13 AVERAGE STIFFENER DISTORTION

2 NEW= 13 MAXIMUM STIFFENER DISTORTION3 NEW: 21 AVERAGE STI FFENER DISTORTION4 NEW 21 MAXIMUM STIFFENER DISTORTION

0510 is

FRAME 1-3 CONTRACTION

F I GU RE 30UNI IIITED AMTE (S R P8 5104 U

- v

GRILLAGE 4A -COMPARISON OF IL. SHORTENING* WITH THEORY (N106)

800 32

LO AD(KN)

600

400

200 -- EXPERIM EN TA L VALUESL

THEORETICAL CURVES

1 NEW= 16 AVERAGE STIFFENER DISTORTION2 NEW= 16 MAXIMUM STIFFENER DISTORTION3 NEW= 23 AVERAGE STIFFENER DISTORTION4. NEW= 23 MAXIMUM STIFFENER DISTORTION

0 s 10 15FRAME 1-3 CONTRACTION (m m)

F I GU RE 31UNLIMlITED AMTE (S) R B5 104

UNL INII TED

ROTATIONAL STIFFNESS OF FRAME SUPPORTS

00 i6 4. Q

,Nmm/IRADIAN

APPLIED /0

-' MOMENT/

15 (N mm) 1

+/0

1.0

0

+0+

05 00

0/

00.0O0. 0.15ANGULAR ROTATION (RADIANS)

FIGURE 32UNLIIIITED AM TE () R 85104 F

-. ~~-.e--.--r

GRILLAGE JB- COMPARISON OF LOAD SHORTENING WITH THEORYC ~.>iEFFECT OF ROTATIONAL RESTRAINT AT FRAME ENDS

LOAD(K N) Re 5 10 Nmm IRAO 1

600

400

Re 0

200 - -EXPERIMENTAL VALUES

05 10 15F RA ME 1-3 CONITRACTION (mm)

FIGURE 33

UrJLINITED AMTE (S) R 85104

UNLIMITED

:'*:• . *APPENDIX A

CORRECTION OF LONGITUDINAL DISPLACEMENTS FORROTATION OF TRANSVERSE FRAMES

Figure Al shows the arrangement adopted for the measurementof longitudinal displacements. In order to avoid errors due tocontact between the transducer and the plating when buckling andvertical movement of the grillage occurred it was found necessaryto raise the pointer above the neutral axis for later tests. Evenin the initial tests the small vertical movement of the framesrelative to the transducer supports was sufficient to introduce anerror in the longitudinal measurement.

If the point is a distance h above the neutral axis rotationof the frame by an angle 0 introduces an error h sin 0.

If the ends of the frames were simply supported and thestiffener exhibited constant curvature the radius of curvature isgiven by

R L (for w << L)

8W

and the angle at the ends is given by

Sin e = Ll2R

where L is the length of the frame

and w is the vertical deflection at the centre of thestiffener (+ve upwards)

Hence the correction due to this rotation is given by

AL = h w

The fractional error in this correction associated withassuming w - L is equal to (2w/L)2 which is less than 1% for themaximum vertical deflection measured (35 mm).

The ends of the stiffeners are not simply supported but areinfluenced by the deflection in the adjacent frames. For thecentral frame it is assumed that if w, = -w2 the rotation will beas assumed above, but if w1 = w 2 the rotation will be zero

For w A w2 a linear relation between these two extremes isassumed giving corrections of:

S(W - w2 ) for the first frame space

and +2h - w2 ) for the second frame space.

-27-

UNLIMITED

I "

UNLIMT

The deformation of the outer frame spaces with doubler platesis small and is assumed to be negligible so that using the sameassumption the corrections for the outer frames are

2 h w,- for the first frame space

L2 h w2

and - for the second frame space.

giving total corrections of

- _ 2 hL (2wl - w2) for the first frame space

2 hand -- (2w, - w 2) for the second frame space.-''% L

The correction over both frame spaces is given by

2 hL (w + w )

which in most cases is less since w1 and w 2 tend to have oppositesigns.

These corrections have been applied to the measured longi-tudinal end shortening using the mean vertical displacements ofthe three stiffeners in each case.

-28 -

UNLIMITED

, .0 ° -.'

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

. . . . . . .. ... .... ... .. . . . . . . . . . . . . . . . . . . . . . . . . .. . . ... -. ...

UtILIMI ItJ

MEASUREMENT OF LONGITUDINAL DISPLACEMENTS

DE FLECTIONTRANSDUCER

.- * 7

h

STi- -NEUTRAL AXIS

FIGURE Al

UN~LINITED AMTE (S) R 85104

77.......7...............

* ~DOCWIENT CONTROL SMET ~I~O(Notse an ccsplatico overleaf)

UNLIMITEDOverall security classification of sheet ......................................................................

(As far as possible this sheet should contain only unclassified information. If it is necessary to enter-- classified information, the box concerned must be marked to indicate the classification eg (R). (C) or (S)).

1. DRIC Reference (if known) 2. Originator's Reference 3. Agency Reference 4. Report SecurityClassification

AMTE(S) R85104 UNLIMITED

(if known)

Admiralty Research Establishment DunfermlineSt Leonard's Hill, Dunfermline, Fife, KYll 5PW

Sa. Sponsoring Agency's 6a. Sponsoring Agency (Contract Authority) Name end LocationCode (if known)

7. TitleINTERFRAME BUCKLING OF ALUMINIUM ALLOY STIFFENED PLATING (UL)

7a. Title in Foreign Language (in the case of translations)

7b. Presented at (for conference papers). Title, place and date of conference

8. Author lSurnae.initials ga. Author 2 9b. Authors 3. 4... 10. Date pp ref

Clarke, J D Swan, J W 10.1985 66 10

11. Contract Number 12. Period 13. Project !4. Other References

15. Distribution statement

Distribution controlled by MOD Technical Policy Authority, Research Area Leader 1,CS, ARE Dunfermline

Descriptors (or keywords)

Aluminium, buckling, strength, test, compression, stiffened plate, interframe,-. F. :.,.yield ..

Summary (UL)

-- Compression tests on five stiffened N8 aluminium alloy plates covering typicalfull scale warship scantlings are described. The panels were manufactured usingnormal shipyard production methods to ensure typical initial distortion and residualstresses.

Maximum loads and post-buckling load/deflection results are compared withtheoretical predictions using the ARE computer program N1O6C.

'.77777r,, = Z 7 z.r" .7. .7.7...

NOTES ON COMPLETION OF DOCUMENT CONTROL SHEET

"nis Zoc-,;.nent Process Sheet is designed specifically for MOD reports and reports produced by Contractors.

Boxes marked * below need be completed only if the information is readily available.

',1. DRIC reference: Enter DRIC reference (BR number) if one has been assigned.

.. Originator's Reference: Enter the report number by which the document is identified by the originator of -Lthe report, in the form in which it appears on the cover.

W, 3. Agency reference: Enter reference number allocated by sponsoring agency (contract authority) in the cameof contract reports.

U. Report Security Classification: Enter security classification or marking which limits the circulation of -

the report, or enter UJLIMITED when this applies.55. Originator's Code: Code number for the DRIC-standardised form of the entry appearing in Box 6.

'h5a. Spcnsoring Agency's Code: Code number for the DRIC-standardised form of the entry appearing in Box 6a.6. Criginator (corporate author): Enter name and location of the organisation preparing the report.6a. Sponsoring Agency (Contract Authority): Enter the name of the monitoring MOD Branch or Establishment in

the case of Contractor's reports. If an MOD report covers work funded by an outside agency, enter the

nar-e of that agency.

7. Title: Enter the complete report title in capital letters but omitting initial definite or indefinite 6 "articles. If the report covers a specific period, enter this after the title, eg (1.1.1972-31.3.1972).

7a. Title in Foreign Language: In the case of translations, enter the foreign language title (transliteratedif necessary) and the translated English title in Box 7.

- , 7b. Conference Papers: If 7 is the title of a paper presented at a Conference, or a Conference proceedings, ""enter the Conference Title, where it was held and the date.

8. Author 1: Enter the name of the first author, followed by his initials.

9a. Author 2: Enter the name of the second author, folloed by his initials.

9b. Authors 3,4...: Enter tnird and following authors' names.

10. Date: Enter the month (in figures) and the year of the report (Dec., 1969 is written 12.1969). If thereport is undated but a period covered by the report is indicated, enter the date at the end of the period.pp. ref. Enter the inclusive number of pages in the report containing information, i.e. includingappendices, tables and illustrations, and the total number of references cited,

I1. Contract Number: Enter the number of the contract or grant under which the report was written.

1.2. Period: (always associated with the Contract Number). Only to be used for reports covo ing a specificperiod, e.g. quarterly, annual or final reports. Enter QR-l, AR, FR, as appropriate.

13. Project: Enter -mroject name or number.i4. Other reference: Enter any reference, other than those in Boxes 2 or 3, by which the report may be

identified.

'5. Distribution statement. Enter any limitations on the distribution of the document. If distribution islimited to particular groups eg MOD, MOD and its Contractors, etc. it should be stated. If the distributionis the responsibility of another authority eg a HQ Directorate (Techinical Policy Authority) enter"Distribution controlled by MOD Technical Policy Authority". Enter "via DRIC" after "Authority" whereappropriate and name the Technical Policy Authority if possible.

Descriptrrs: Amy number of descriptors (or key-words) can be entered. If selected from a published thesaurus,eg t e Thesaurus of Engineering and Scientific Terms (TEST), this should be indicated.

Abstract: The abstract should preferably not exceed 150 words, i.e. it can be considerably shorter than theAbstract to be provided on the Title Page of the Report. Information available in the report title need not beincluded in the abstract. !y.

ft.. .-..

,' ".. . -U

* - * . * *. a -- . ,... -.. 4.-.€-k.

v:-. 4%. "--r t-..-0 -

.. z .u o.. 0-. ..E .e-'-

I- CM inW 1- go = = aw

I-' - ICe.G - IC€ .U- 4-.- ( 4- 0 (A c

LCsI C 0-I 4 ~ r_4 .,J -GC LViC4

.=xf4f- . .. 0- &n r - 2 -..

v 4EA: 41 .4u, ,-" '3u 0

0 •~ .o V) -.. 0

-D 4 - C CD -,,--0 >, -' -_.C ,4 - . 0 ,4J-- & -M =-

: :0 0. >b10f-0 ,- 5. 6 9!-.t 0 c LmO

U 4. oL go-e0 DM 1 C o -dC

-- '0 Ci 4.J".O.E

C- -.- E iClD l . 0 00P

=D (1 ) CC U n U L L. ) a L 0 U r= I0L4

to to" 0 > L to. 1 0

• - 4- ' Ln _ o 1-, "0 44 .-- 4. -Ma..C

" €: : e.I-. u i--n r1m- > e p *O ~ - o . ~ j- n i-- *OVO .' - 000. ,,,%

.f ,Z= , "D 4-> W ,--,

.4- 4 0 0 u * CL &A 0%0_

I-r 06 CA 06 .-. ' E4bJ ~ . C 0,uw-C- W ..

= . 4,.0 C £ = U=C 41 C).

& c 0 0 " r-

-1 0 -,-,S(AI C 0 C - 1 1- C0. C 0-10

. o " 1 90 fu -I 3 0 M v E

4" 0 0 o 1 W- o X0) #A-

..,-. M -,, to c. go _= O 0 go r_ = ... to -, 0.,.(A C...EJ.- - 0 6 .. 4 - 0 .

C 0 0 0 C

-O._i 4LP J 04- 41 . c ,c j -c= 0 E-0) 04-' 4

C3 .. f 0J W 4 0 3 CL : - Ews ..) ....

3- 8C - W 2M 3C 00 - >in 1*0.0IVLA_21 *. W W 0. to 0 &A *uwoii .LL.a00WV0 n 4

LL 0L LA - 2 4- J4JL 4'0: L LA U- -' 4J 40 > 4.3 -

C0 i 0 :CL 4 n- C - .i 0 0 .. )>! ' 4! 0.- . 0 r_.i. c.- >% 0go--

X-' 4J. XC i.~ 4-'III tII LL __ _ __ _ __ _

CL - 06 i

0--V .I 0 0

4- o Ci~t L. 4J

FLME

DTIC

Documents

-A163 909 INTERFRANE DUCKLING OF ALUMINIUM ALLOY … · the yield behaviour in compression, but a total reduction of 4% [ has been been included allowed in for the hold values time