Fourth PROGRESSREPORT. (Proiedt SR-1, 19) WELDED REINFORCEMENT OF OPENINGS IN STRUCTURAL STEEL MEMBERS:

.,.-,1 SERAL No. SSC.56,, (...,.,, j,.,.,. 1, Fourth PROGRESSREPORT (Proiedt SR-1, 19) on WELDED RENFORCEMENT OF OPENNGS N STRUCTURAL STEEL MEMBERS: Cleavage Fracture and Plastic Flow in Structural Steel Plates with Openings,,,,by 1.. - -l.:..t. [ ;.1 D. VASARHELYJ R. A. HECHTMAN AND Y. T. Y05HlMl University of Washington,, Under Bureau of Ships Contract NObs-50238 1!, (BushipsProiect NS-731-034),,, for,,,,1 SHP STRUCTURE COMMTTEE 1 Convened by The Secretary of the Treasury, l, 1, Member A~et.acies-$bip Strkture Committee tidd~ess Con-espondence To:.>.L..,,., Bureau of.ships, Dept. of Navy Military Sea Transportation Service, Dept. Of Navy United States Coast Guard, Treasury Dept. Mariiime Administration, Dept. of Commerce American Bureau of Shipping, Secretary Ship Structure Committee U. 5, Coast Guard Headqudrfers Washington 25, D. C! -;..- _l ;.,,,,, March 1/ 1954,,,,, -.... -_..,.-.-..-

SHP STRUCTURE COMMTTE~ MEMBER AGENCES: ADDRESS CORRESPONDENCE TO: ~UREAU OF SHPS, DEPT. OF NAVY MLTARY SEA TRANSPORTATON SERVCE. DW7. OF NAVY UNTED STATES COAST GUARD, TREASURY DWT. MARTME AoMNSTRArlON. DEFT, OF COMMrRCK SECRETARY SHP STRUCTURE COMMTTEE U. s. COAST GUARD HEAOCUARTCRS WASHNGTON 2B. w. c. AMERCAN BUREAU or SHPPNO March 1, 19 3 Dear Sir: As part of its research program related to the improvement of hull structures of ships, the ship Structure Committee is sponsoring an investigation on the welded reinforcement of openings in structural steel meimbersat the University of Washin~ton. Herewith is a copy of the Fourth Progress Report, SSC-56, of the investt.~ation,entitled ~WeldedReinforcement of Openings in Structural.SteelMembers: Cleavage Fracture and Plastic FOVJin Structural Steel Plates with Openings rby i). Jasarhelyi,11.A. Hechtman and Y. T. Yos himi. Any questions, comments criticism or other matters pertainin~ to the Report should be addressed to the Secretary, Ship Structure Comw.ittee. This Report is being distributed to those j.ndividualsand agencies associated with and interested in the work of the Ship Structure Committee. Yours sincerely, K. K. COWART Rear Admiral, U. S. Coast Guard Chairman, Ship Structure Committee.

Fourth Progress Report (Project SR-XL9) cm WELDED RENFORCEMENT OF OPENNGS Hi STRUCTURAL STEEL MEMBERS: Cleavage Fracture and Plastic Flow in Structural Steel Plates with Openings by D. Vasarhelyi8 R. A. Heehtmn$ and Y. T. Yashimi JNWERSTYOF WASHNGTON under De~arbent of the Navy Bmsau d Ships Contract NObs-502~8 BuSM.ps Project No. NS-731-034 for SHP STRUCTURfi~&&ik!MT3ZJ.

CXJMVAGE FRACTURE AND PLASTC FLOW N &JWCT TJHALSTEEL PLATES WTH OPENNGS TABLE 0S?CONTENTS 000000 000QQO 0 00000 09000 Ofio o WizQ 1 2 111.3 Openings... ~..... 0 2. Methods of Piastic Analysis and Calibration Tests. o.. 0.. 0 0 0 0 0 0 0 0 0 Q o u o 3. General Behavior During Test and Fracture of Plates with Openings. ~ o ooooo* a 0 i 4 9 9 v. Elastic Stress Distribution in Openings........ 0 0 0 plates with 000000 m 10 Plastic Stress Distribution in Openings........... Plates with OOQOOO o 14 Plastic Energy Distribution in openings...... 000000 Plates with 000000 0 20 V. Conditions for the nitiation of Fracture. 0 25 V. conclusions. o 0 0 000000 000000 0 28 u. Acknowledgments.. Ouooo 0.5000 00 0 30 x. Bibliography... 000000 000000 0 31 Appendix B: Test ~f Specimen No. 85...... 62

LST OF FGURES Title m Plate with Circular Opening. Spec. No. 69..... 32 Plate with Reinforced Square Opening with Rounded COrners. ~p~cso ~JO. 70 and 71........ 32 Details of Specs. No. 49 5$ 17$ 37$ 38$ and 38A.. 33 Details of Specs. No. 95 and 96.......... ~~ Location of Arrangement SR-4 Gages and Slide-Wire Gages.... 34 of Grid-Wire Gages for Spec. No. 69.. ~% Jirrangementof Grid-Wire Gages for Specs. No. 95 and 96 O*-** *.*.* ec.e..o e... 35 Arrangement and 71 of Grid-Wire Gages for Specs. No. TO ***ee* Oot...om O..*. O Zoad-Average Elongation Curves. Specs. No. 37, 38, 69995and96.. 0....... 0......36 lhd-average Elongation Curves. Specs. No. PO and 710.....................36 Photographs of Plates after Failure. Specs. No. 69, 70and71....... a.. m..... a..37 Photographs of Plates after Failure. Specs. No. 95 and96....................~~ Particular Cases of Plates with Unreinforced Open- ings Analyzed by Elastic Theory. Plates [a], (b) and(c)..................38 Particular Cases of Plates with Reinforced Openings Analyzed by Elastic Theory. Plates (d) and (e).,....................38 Stress Concent~atfon Contours lrlv-llirectionfor Plate of nfinite Width with Circular Hole by Elastic Theory. Plate(a)..........39 0 35..-

Title Stress Concentration Plate of Finite Elastic Theory. Contours-in y-direction for Width with Circular ltole by Phte(b)m...039..039 Elastic Unit Strain Concentration Contours for Plate of Finite Width with Circular Opening Plotted from SR-~ Strain Gage Readings. tspec. No.69...... 0. 0.. 0.... 0.. 39 Stress Concentration Contours in y-llirectionfor Plate of nfinite Width with Square Opening with Rounded Cornersby Elastic Theory. Plate CC)......................39 Elastic Unit Strain Concentration Contours for Plate of Finite Width with Square Opening with Rounded Corners Plo%ted from 5R-4 Strain Gage R~adings. Spec. No. 38A............ 40 Elastic Unit Strain Concentration Contours for Plate of Finite Width with Square Opening. Plotted from SR-~ Strain Gage Readings. $pec. NO. 950 ~em. o. * o,.. o o * m. *.40 Elastic Unit E%r& Concentration Contours for Plate of Finite Width with Square Opening. Plotted f :~gqr-4 Strain Gage Readings. Spec. Nc?.g60 o.. 0 0. 0 m. e.. 0.40 Stress Concentration Cmdmurs in y-direction for Plate of nfinite Wftithwith Circular Opening with Face Bar Reinforcement by Elastic Theory. Plate(d)...................41 Stress Concentration Contours-in y-d&ection for Plate of nfinite Width with Circular Opening with ns@rt Plate Reinforcement by Elastic ~eary. p~ate(el..............42 Elastic Unit Strain Cbncentrati.onCurve for Plate of Finite Width. nsert Plate Reinforcement. Spee. No.17.................42

Comparison of Elastic and Plastic Stress Concentration Curves.forti-rf orplate with Square Ope,ning with Rounded Corners. 760F....m... O..!p!c.?.?7:... -. iv.-.-

StTess CminentTawith Square Open- *...O *em Stress Concentration Contours for~-flfor Plate with mq 46 47 47 47 48 48 Comparison of Elastic and Plastic Stress Concentration Curves forfi-~for Plate with Reinforced Square Op::m:: with Rounded Corn,ers. $pec. No. 71. 00..- Ge*om **C.0* 48 4-9 49 50

Title unit Unit unit Unit Unit lllitstrain Energy Contours for Plate of Finite Width with Reinforced Square Opening with Rounded Corners for Zoad of 1276 kipst Maxim Load. Spec. No. 70. 76@F..........S1 unit Strain Energy Contours for Plate of Finite Width with Reinforced Opening with Rounded Corners for Load of 1150 kips~ 90 Per Cent of Maximum Load. Speco No. 700 ~6@F...... ~1 Strain Energy Contours for Plate of Finite Width with Reinfore~d Square Opening with Rounded Corners for Load of 1150 kipsa$~~~r Cent of Maxim Load. Spec. No. 71. O* 52 Unit Strain Energy Contours for Plate of Finite Width with Reinforced Square Opening with Rmnd.ed Corners for Load a:k~:;6 kips. &i. mum Load. SPec. No. ~1......... 52, Contours of Equal Rate & Energy Absorption. speco No. 69. 76QF.................52 contours of Equa~ Rats of Energy Absorption. No. 37. 76QF.. 0 0. 0.. * w *. *.XC:53 Contours of Equa~ Rate of Energy Absorption. Spec. N0.38. -2OAF..........o.o,eo5~.. -.

!Htle 59. ~OlltOllrSof Equal Rate Of Erwrgy Absorption. Spec. No. 960 J&l?...... 0, 0 0 a o 0. * 53

Title -. Page 5b Photograph of s-pee.no. 85 after Failure..... 67 viii

LST OF TABLES Description of Specimens with 9 in. x $1in. Openings with l-l\8 in. Corner Radius..... ~ 111 v v Strengthand Energy Absorption of 36 in. x 1/2 in. and &8 in. x 112 in. Plates With Openings at Boom and at Low Temperature ~ 8 Comparison of Load Computed From True Stress Distribution and Testing Machine Load.....,19 Comparison of the Total Ensrgy Obtained by the Octahedral Theory and from the Load-Elongation Curweo o... 0..... 0 *..... e. 24 Amuendi~ Aa a Results of CaHbraticm Tensile Tests..... 59 H3 Description and Test Results d Spse. No. 85. 63 Kcb Geneml Yielding and Fracture of SpBc~ NJo~85. 64 ix -.

CZEAVAG13FRACTU ~ AND PLASTC FLOW TN STRUCTURAL STEEL PLkTES WTTH OPENNGS.-. o SYNOPSS of openings in structhe development of better design specifications. A thcrough investigation of: the plastic behavior of the plates was deemed necessary as a phase M this pro,ject. This avenue of approach led to energy and stress studies which utiliz~idexperimental tech-

-2-110 11 TTRODUCTON The tests of twenty-three l/4-in. plates in the First Progress Report (1) with and without welded reinforcement dmw attention to the importance of the behavior of these p~-ate~in the plastic range. A more detailed study of the unit strain energy and of the stress distribution in th~ plastic range was decided upon. The plastic analysis employed Nadaifisoctahedral theory (k) and.a method of stress determination developed by this investigation(210 The results of Wsts at both room and low temperature of thr=e 36-in. x 2~/2-in.plates with an unreinforced square opening with rounded comers were reported in the Second Progress Report (21~ and the following main conclusions wem drawn with respect to their behavior in the plastic range. 1 The octahedral theory proved to be a practicable means tm study the distributicm of the distortional energy in the ~12tStiC range of steel. 3 LO The true stresses in the plastic range were satisfactorily found by a method of analysis developed..

-2- > by R. A. Hechtman far the purpose. 3. Both the disto~tional energy and the true Wress analysis gave values in good agreement with the values 1 obtained by entirely different methods. 4. The maximum values of the true tensile stress occurred in the regions where the greatest unit energy absorption was found. 5. A drop in the testing temperature from 76 F to -20%, which changed the type of fracture from shear to cleavage, caused no significant increase or decrease in the energy absorption but made the unit energy gradients steeper in the region of the opening... program. These conclusions suggested the nature of the subsequent The present investigation has had these additional objectives 1. To analyze both the unit distortional energy and the true strssses in a plate with a circular unreinforced opening and to correlate these data with the elastic stresse~scomputed by the mathematical solutions available for this pro blem. 2. To determine bath the unit.strain en.er.gyand the true stresses in plates with a sharp natch~ the square opening, and tc correlate the data with the results obtained with much less severe notches~ the circular opening and the square opening with rounded corners.

?r --F -.:.. 111. TESTS OF PLATES WTH OPENNGS

TABLE DESCRPTON OF SPEGMEN8 WTH 9 in. x 9 in. OPENNGS WTH 1-1/8 in. CORNER RADUS Spec e Size of Reinforcement Percentage cross - Sectioti Testing Noe Of Rein- ~.-. Sq. Tneh Temp. n. forcement Gross Net deg F.m-..q Y-...-?-z ~...-.. -.. 38 None o 18.00 13.50-20 To Y2-3/4 x 22-3/4 x * 39 2%.00 21.38 76 71 12-3/4 X 12-3/~4X * 39 2koocJ 21038-46 96 Ncme 0 18.00 13,50-46 * nsert Plates

-6- TABLE 11 XJST OF PLATES SEDFOR FABRCATON OF EACH SPECMEN Spec. No. Plate No. Used?ora - Body Plate Reinforcement 37 38A 38 69 PO 71 95 96 26 4 4 3 3 3 1 1 --.-.- -- 10 10 -- --

7 co r- L1 8 l + H H 53 3

TABLE u Spec PerCent Test Fracture+$ GeneralYielding- UltimateStrength EnergyAbsorp.-w Natureof No. of kw u Per Cent kd Ave. Stress hd Ave* Stress T Final Reinf, c s un- Gross Net Gross Net U1.t&ateFa&re Fractmre. aeg.f ~ broken ksi ki 6 ksi ksi 1000~s Jil.ms 36M X 1/21 B* Plate. No Reinforcement. 37 0 76 0 54 M 450 25,0 33*3 8W 44.5 59*3 1700 2179 Thruopening 38A 0 0 87 13 0 SOO 27.8 37*O 898 49.9 66.5 2890 3470 ThruOpen.hg 38 0-20 91 90 @l 27.8 37.0 915 50.8 67.7 2778 2778 Mru opening 69 95 0 0 76 0 76 0 67 33 ~oo 27.8 37 o 845 47.0 62.5 89 11 5L7 26.$ 3594 710 39.4 52.6 1739 1100 2$33 1597 Thru opening & ThruOpening 96 0-46 100 00!!30 30.6 40.7 61@ 36.o l@ o 486 486 ThruOpe~ 48 X 2/2 BOdyPlate. ksert Pla@ EeinforceEnt. 70 39 76 1 50 b9 BOO 33.3 37.6 2276 53.1 59.7 3362 3699 Weld to Reti. 71 39-46 mo 00 800 33.3 37.6 1176 48.8 55*O 2084 208h Thr?lopening +i- Proportionh per centof totilnet cross-section areaat fractmresurfacetichdingfractureand unbroken section, if 131qrcC = Cleavage.S = shear. w 36-ti, gagelengthfar 36~t x l/2~tplates. 48-in. gage lengthfor 4@ x l/2t1 platis.

theory to the determination of the unit ener~y distributio~ in notched platesa as wsll as the tangent method of plastic stress analysis and its derivation (210 Progress Report was given in the Second Both methods utilize the measured strains in the plastic range of the material. The data of the calibration tests are given in Appendix A. 3. G+ Behavior during!&st and Fracture ~ Pates with. OneninQsO A comparison of the applied load and the average elongation on a gage length equal to the half-width of the plate for these plates is sh~wn in Figs. 9 and 10. This gage length extended vert~-tallyupwards from the transverse centerline of the specimen and enclosed the area in which the. grid of slide-wire resistanc~ gages was mounted. Thus the area under these load-average elongation curves up to a particular load represents the same quantity of energy as was obtained by the application CJfthe octahedral theory to the elongations measured by hh~grid s;ystemat that load. t is interesting,that a change in the made of fracture fron sh~ar to cleavage was accompanied by little change in the load-average elongatiorlcurves for the plates with the less severe strfiss-rais~rsasuch as tb.ecircular opening and square opening with round~d corners. n contrast a drastic change occurred in the shape of the load-average.

-1oelongation curves and there bya large reduction of energy absorption in the case of the sharp stress-raiser, the square opening. This observation suggests the manner in which the!, plastic energy absorption and the subsequent type of fracture were related to the degree of triaxiality Qf the stress con. dition at the notch. Photographs of Specs. No. 69? 707 71? 95 and 96 after failure are shown in Figs. 11 and 12. The deformation and fracture of SpeCSe No. 69? 70 and 71 were described in the T%irdProgress Report (3) ~ and the reader is referred thereto for a more complete discussion of these points. The data of these tests are summarized in Table V. V. ELASTC STRESS DSTRBUTC)N N PATES WTH QPE?KCNGS The elastic stress distribution was computed by theoretical formulas wherever a solution was available for a case similar to or like that d the test specimens. The theoretical stresses and the experimentally measured unit strains could then be compared. The stresses were computed for the following five cases% a. A plate Gf infinite width with a circular opening. b. A plate of finite width with a circular opening~ the width being four times the diameter of the opening.. The proportions of this plate were the same as those of Spec. No. 699 for which an experimental stress.

analysis was made in both the elastic and the plastic ranges. c. A plate of infinite width with a square opening with rounded corners? similar to Specs. No. 4 and 38A of the First and Second Progress Reports (1,2) ~ whose width was four times the diameter of the hole. d. A plate of infinite width with a circular opening reinforced with a face-bar reinforcement, similar to Spec. No. J of the First Progress Report (1) ~ whose width was four times the diameter of the hole. e. A plate of infinite width with a circular hole reinforced with an insert platel similar to Spec. No. 17 of the First Progress Report (1) ~ whose width was four times the diameter of the hole. Sketches of these plat~s are shown in Figs. 19 39 13, and 1%. The results are presented as elastic unit stress or unit strain contours in Figs. 15 to 259 in which only the stress or strain component parallel to the direction of loading is shown. The theoretical background of and formulas for the stress computation are not given in this paper. The reader is referred to References 5 to 9, inclusive, in the Bibliography. f the theoretically computed elastic stress concentration contours for unreinforced plates in Fig. lj1 16, and 18 are compared$ a number of points of similarity may be noted.

-12- The stresses some distance from the opening in these particular cases are not substantially affected either by the shape of tihe opening or the width of the plate. The contour for unit str~ss concentration lies at ai~ost the same angle and in almost the same location for all of these cases. At the Edge of the opening the maximum stress concentration of 3.00 for the p~at~ of infinite width with a circular opening in Fig. 1~ increases to only.3.23when the plate width is decreas~d to four times the diameter of the opening (Fig. 16}. The maximum stress concentration for the plate of infinite width with a square opening with rounded corners in Fig. 18 is 3.09y and for a plate width of four times the width of the opening would be somewhat greater. Thus for plate widths greater than about four times the diameter of the opening and fcm a corner radius of the opening greater than about one-eighth the dianeter of th~ openingq the theoretical elastic stress distri but.ionsare very Similarq and the maximum stress concmtration varies only between the limits of 3.00 and a maxhnum slightly greater than 3.23. This similarity with respect to botihstress distribution and stress concentration factor explains why both the energy absorption and the ultimate strength of the plates in the First Progress Report(1) with a circular opening and a square opening with rounded corners having a radius of D,/8were essentially of the same.. order of magnitude.

-13- The experimentally determined unit strain concentration contours in Figs. 17 and 19 were in good agreement with the theoretically computed stress contours in Figs. 16 and 18. Some allowance must be made ii comparing Figs. 18 and 197 - since the experimental results were obtained for a plate of finite width and the theoretical values for a plate of fnfinite width. When the experimentally determined unit strain contours in Figs. 20 and 21 for the square opening with the l\32-in. corner radius are compared with the similar plot in Fig. 19 for the square opening with the rounded corner, it may be seen that the high values of strain were concentrated more closely around the opening in the two plates with the sharp corner radius. The maximum value indicated was computed from the reading on a SR-4 gage of l/~-in. gage length located as close as possible to the point where the maximum was expected. Consequently the value shown here may be somewhat smaller than the true maximum. The effect of reinforcement upon the stress or strain distribution is shown in Figs. 22.-25. t would appear that the Beskin solution (6) fo~ these two cases gives a fairly good picture of the elastic stress distribution and the maximum elastic stress concentration. When Figures 22 and 24 are compared with Fig. 1~, it may be seen that the stress concentration contours in the

plates with reinforcement around the opening because of its greater stiffness restrains the boundary of the opening and develops transwrse tensile stress in the region of th~ weld. b~tween the reiu,~orcementand the body plate abme and below the openingj wher~ compressive stress wmld be present in an unreinforced plate. n the case of certain types of reinforcement which have relatively high rigidity~ a different approach to the analysis of the stresses in the body plate may be advisable. The reinforcing ring in such cases should perhaps be considered as a rigid inclusion restraining the circumferential deformation of the opening in a manner which according to Reference 8 in the Bibliography brings about very high sh~ar str~sses in the tody plate. These high shear stresses are located at the corners irithe case of a square opening with rounded corners. The fact that in previous ti.stsq~ the pkt~s with face bars having the larger per. centage of reinforcement broke in the weld in a ~ashion indicating high shear stresses in this location points to the n~ed for additional theoretical work along these lines. i.plastc STF~SS DSTRBUTON N PLATES WTH OPENNGS The stresses in the plastic range of the Steel were cow.putedfrom the measured strains by the tangent modulus Prog-.-

distributions in Figs. 26--43 show the ratio of the true stress at any point in the y-directiany the direction of loading, to the uniform true stress an the g~oss area of the specimen in a region remote from the opening. Contour maps at a number of loads in the plastic range were platted for Specs. No. 69, 37Z 389 70~ 71$ 959 and 96; butonly one or two of these are shown for each plate~ one of which is for maximum load~ the instant at which fracture was initiated. Whm the elastic stress concentration contour maps in Figs. 15--2J are compared with those in the plastic range, a number OT similarities may be seen. The pattern of the contours3 the distribution of t hehigh.and low values, and the location of the stress concentration contour of unit va,lueare very much alike for the same type of specimen. Moreover~ the shape of the opening affected the contours only in th~ vicinity of the opening. The general similarities between the elastic and the plastic stress distributions substantiate to some degrwe the common assumptions of the theory of plasticity that th principal.stress directions and the general stress pattern are not changed by t hetransition from the elastic to the plastic state. The effect of inc~easing tlw plastic stress level upon the values of the stress concentrat-ionswas a tendency of the stresses to approach uniformity. The maximum stress Concentration$ commonly called the stress Concentration factora

-16- is compared with the percentage of the ultimate load in Fig. 44. n these plots the experimental value of the elastic stress concentration factor has been plotted at the relative load at which general yielding began. t was found that the plastic stress concentration factors for the plates with the circular or the square opening with rounded cornersy Specs. No. 37, 38, 69, 70, and 71? plottiedas one family of curves, one curve for unreinforced plates and me for reinforced plates, regardless of the type of fracture. n Section V of this report in Figs. 15--195 it was found that the elastic stress concentration factor and the elastic stress distribution were quite similar for the circular opening and the square opening with the rounded corner. t is not surprising therefore that the stress concentration. factors in the plastic range were closely similar. However? a different curve resulted for Specs. No. 95 and 96 with the square opening. n the case of this sharper corner, the plastic stress concentration factor fell off much more rapidly than for the less severe corner radiia and this reduction took place closer to the maximum load. The stress concentration factor was always maximum in the elastic range% decreased as the plastic stress or load level increased~ and approached a constant and also a minimum value as the ultimate strength of the plate was reached. That is, fracture began when tha stress concentration factor approached a constant valuea which was also a minimum value. This observation

.~?. suggests that perhaps the ow-energy cleavage fracture of same welded elements, which is often accompanied by low ultimate strength? may result in part because the amount of plastic flow which has occurred is not large enough tq b~ing about a sufficient reduction in the plastic stress cmcentration factoro The plastic stress distribution shown in Figs. 26--43 was examined with the view of determining whether it may be correlated with the type of fracture in any way. A statistical analysis in the gaged area of the frequency of the various ~~al~esof stress ~oncentrati~n is shown in Fig. 45, which ~~~mpares.the results for a shear fracture with those for a predominately cleavage f racture$both for reinforced and unreinforced plates. The manner in which this analysis was developed will be explained. The gaged area referred to is the area of the specimen covered by the grid-wire system as shown in Figs. 6--80 For example, in the plot for Spec. No. 37 at the top of Fig. 459 approximately 3 per cent of this gaged area developed a stress concentration of 0.4~ 19 per cent 04P the area 0.8a and so on. Thus Fig. 45 is a distribution curve with respect to stress concentration and indicates what proportionate parts of the specimens were under either high or low values of plastic stress. Zn each comparison in Fig. 45 are shown the analyses for a room temperature specimen and an identical low-temperature s.pecimen~where the predominate mode of fracture was shear in the former-and cleavage in the latter. n this figure when

-18- the analyses for the two identical specimens tested at the two temperatures are compared, it may be seen that a larger portion af the area in the low-temperature specimen developed the lower values of str~ss concentration, while a correspondingly smaller area developed the higher values. The specimens sustaining a predominately cleavage fracture did not produce the same plastic stress distribution as those with a shear fracture. For the plates with a cleavage mode of fracture, the higher plastic stresses were concentrated in a smaller region around the opening; that isa the stress gradients were steeper. Cleavage fracture was accompanied by a less ef. ficient stress distribution in the plastic range than shear fracture. When the true stresses on any transverse cross-section of the specimen were summed up with due respect to the plate thicktotal force on the cross section. A ness~ the resultant was the comparison of the values obtained in this manner with the testing machine load is given in Table V. Agreement within fifteen per cent was attained for most of the computed values. t would be well to analyze the principal sources of error in the plastic stress analys~s. These are as follows: 1. The minimum of the two biaxial stresses frequently fell in the incipient yield range where the values. of the tangent modulus and Poisson~s ratio were uncertain.

.39. TABLE V COMPARSON OF LOD UM?UTED FROM TPJJESTRMSS DSTRBUTON AND TESTNG MACHNE LOAD - Spec. Machine c~mp~ted Load at Cross Section. kius~,. No o Load A B c D kips - -.- * Distance of cross section measured from transverse centerline of specimen.$ which is also the horizontal axis of the opening.

-20-2. The assumption that the x- and y-directions were principal directions was more in error> the closer.- the gage point was to the opening. The poorest correlation between the testing machine load and the computed load was usually found on cross sections near or through the opening where the deviation of the principal directions from the coordinate axes was greatest. 30 The selected cross sections of the specimen, which were initially straight lines$ became considerably distorted as the maximum load was approached. ntegration of the values along this somewhat curved 40 cross section for the shear The slid-wire produced an errar~ since no correction. stresses thus introduced was made. grid system$ which was design&d for large strains? was not sensitive to an elongation in any gage length smaller than 0.001 inches. The sensitivity of the system was therefore in the yield range of the material. After a review of the errors in the computed values, it appeared that the preceding reasons were responsible for the errors and not some inadequacy of the stress equations themselves. V. PLASTC ENERGY DSTRBUTON N PLATES MT~ OPENNGS The unit energy distribution in the vicinity of the opening.-.

g q -L-, - was computed by the octahedral theory experimental tandanalytical procedure Second Progress Report(2). of A. Nadai(4). The was described in the flmtour maps showing the unit energy distribution in the plastic range appear in Figs. 1+6--fi. Although this analysis was made for each plate at a number of load levels in the plastic ran.ge~only a few typical energy contours are shown in this report. t is interesting to point out that the contour line far the average unit energy absorption in the plastic range, the total en~rgy absorption in the gaged area divided by the volume corresponding to this areaq fell in almost the same locatign in each plate as contour line for unit stress concentration far both the elastic and plastic stress states. Moreoverq the higher values of the unit energy absorption appeared in tineregions where the higher values of the plastic and elastic stresses occurredt and vice versa. The maximum unit energy absorption~ which always occurred ad,jacentto the opening: was much greater in the unreinforced plates than,in the reinforced plat,es~as Figs. k6--fi show. t should be pointed out that the grid systen Gf me-inch squares was not fine enough to determine either the exact lgcation or the tme value of the absolute maximum. The maximum values in these figures are probably less than the unit energy which existed at the point where fracture was

22- fnitiated. Fracture started in these specimens at maximum load. The Second Progress Report showed that the unit energy absorption u at any point increased in the plastic range in accordance with the empirical equation, A +BPj u=e where A and B were numerical quantities and P the applied load. t was observed that A remained almost con,stant. The significant variable was B, the slope of the semi-logarithmic curve,., relating u and P. From semi-logarithmic plots of u against P for each point of the grid system, the values of B were obtained. A similar semi-logarithmic plot with respect to the average unit energy absorption ~Av for the entire gaged area gave the average a~ue f B r Av he atio \BAv has been called the relative rate of increase of the unit energy absorption. Maps showing the contours of equal values of B/BAv appear in Figs. 55--6. The major differences which were found in these figures with respect to the strain energy distribution in the plastic range were: 1. 2. The maximum value of the unit energy absorption at ultimate load was about twice as great in the unreinreinforced plates. forced plates as in the The distribution of the energy w~asmore nearly uniform in the re~.nforcedplates with less of a spread between the maximum and the minimum values.., 3. The distribution of the energy over the gaged area was

more n~arly uniform in the unreinforced plates with. circular openings or square openings with rounded corners than in the unreinforced plates with square openings. A concentration of t,h~high values in the vicinity of the sharp corners of the square opening was noticeable. When the unit strain energy values for a given load in Figs. 46--54 were integrated they could be compared, as shown in Table V13 with those values obtained from the load-average elongation curves. Reasonably good correlation was obtained so long as the plastic strains were fairly large. liowever~poorer agreement occurred in all t-heplates at,low loads where much of the gaged area had not begun to yield and in those plates with the square opening wb.erethe yielding was concentrated almost entirely at the corners of t heopening. The resistance-wire grid system used to measl~rethe plastic strains was designed to measure large values of strain and was therefore not sufficiently sensitive when yielding was just beginning. n the plastic stress concentration contour plots in Figs. 299 33? 343 ~b~ 39$ arid +27and in Fig. \59 it,was shown that the plasti~ stress gradients were steeper in the specimens with a p~edominately cleavage fracture than in identical specimens with a shear fracture. A similar type of analysis of the unit energy frequency distribution was made for the same specimens and is shown in Figs. 62.-6k. A simi2ar trend is revealed in that a larger proportion of the area of specimens with a predominately

-L 4-,,; 0.000.*. 1+ r-l 4 -A

deavage fracture developed the lqwer values of unit energy than of those with a shear fracture. The gradual development of this divergent behavior as the loadsincreased may be seen in Fig. 62 for Specs. No. 37 and 38 if the unit energy distribution is examined at the different loads V. CGNDTQNS FOR THE NTATON OF FWCTURE t would be well at this point to examine the experimental data for information pertaining to the initiation of fracture. The simplest of the common theories of fracture assume that fracture begins at a point at the moment a certain limiting value of the principal stressf principal strain, or unit energy characteristic of the material at the given temperature has been exceeded. &ch theories, it should be pointed out, do not differentiate between types of fracture, cleavage or sheary or take into account strain-aging and other metallurgical changes in the material. Data are available herein to examine the applicability of these three simple hypotheses of failnrer since the information from which the stress concentration and unit energy contour maps were computed give the observed maximum values of stress, unit strains and unit energy at ultimate~ or maximum load where fracture was initiated. The maximum true str~sses computed from the observed strains were as follow~l:

-26- Spec. Testing Percenta~e of Fracture Max. True Stress No. Temp.,OF Cleavage Shear ksi 37 76 ~ 90.0 38-20 9? 1:{.: 69 76 0 ;: 70 76 87; O 71 -&6 10: 8100 95 75 i; 9007 96-46 10: -- 6805 These While are principal stresses at the boundary of the opening. these values are of the same general order of magnitude~ it would not seem that a maximum stress theory could predict failure in these specimens with sufficient accuracy.. The maximum unit nominal strains developed in the unreinforced and reinforced specimens are shown in Fig. 65. The unit strains in the unreinforced plates were approximately double those in the reinforced plates. The plot in Fig. 66 for the unreinforced plates with that the maximum unit strain for related to the type of fracture. a square opening indicates this more severe notch was The maximum unit strain hypothesis of failure would not apply to these specimens. The maximum unit distortional energy in the specimens is also shown in Figs. 65 and 66. t is obvious that a limiting value of the maximum unit distortional energy not properly indicate the imminence of failure. Since the maximum stress occurred at the boundary would of..

--Jq L(*. the opening where the stress in the normal direction was zero and the stress in the direction of the thickness of the plate extremely sm.all~the maximum principal shearing stress would be a function of only the maximum stress in the y-direction. Since the maximum stress theory did not hold for these specimens? the maximum sheartng stress theory would not apply in this case either. t wo_jldappear that any theory of failure must consider other factors$ such as testing temperature the mechanical and\ or heat treatment of the rnetal~and the anisotropy of the metal~ as well as the geometry ~f the specime~l. The maximum stressa maximum unit strai.n~and maximum unit distortional energy are relat~d to all these factors and not Just to the geometry. The data in this report appear to establish the following facts concerning the conditions for the initiation of fracture: Fracture was initiated when the strsss concentration factor for a given specimen was approaching or reached a minimum and constant value. The plastic str~ss and unit energy gradients were steeper in,the specimens with a predominately cleavage fracture than in those with a shear fracture. h~en the stress-raiser became su~ficiently severe? the energy and the strain absorbing capacity of the plates was substantially less in the ease of a predominately cleavage fraature than for a shear fracture.

-28- The second and third observations would rule out the possibility that a specim~n developing a shear fracture could accurate~y describe the plastic stress and strain energy conditions 1 of an identical specimen at a temperature which would cause cleavage fracture. The theories which were used to develop the unit energy and stress distributions in the plastic range are based on the assumption that all stresses and strains are the result of the applied loading. Reasonably good checks were found between the values computed by these theories and the applied load and energy input. Therefore~ it would appear that the initial residual stresses from welding have no appreciable effect uponthe stress and energy distribution in the plastic range of the material. V. CONCLUSONS The following tentative conclusions have resulted from the investigation of plastic energy and stress distribution: 1. The maximum values of elastic and plastic stress, elastic and plastic strain and of unit distortional energy were located at the same point$ the point where failure started. 2. Apparently? no theory of failure based upon a limiting value of stress, strain or energy would yield a numerical value accurately indicating the imminence of failure. Such factors as the metallurgical characteristics

-::9- and testing conditions should be considered. 30 The effect OT increasing the plastic stress level upon the values of tha stress concentrations was a tendency of the stresses to approach uniformity. The stress concentration factor was always maximum in the elastic range,.. decreased as the plastic stress or load level increased, and approached a constant and also a minimum value as the ultimate strength of the plata was reached. k. The low energy cleavage fractme of some welded elements~ which is often accompanied by low ultimate strength, may result in part because the amount of plastic flow which has occurred is i~otlarge enough to bring about a sufficient reductior~in the plastic stress factor. concentration 5. The effect of low temperature was a steeper gradient of stress and unit energy in the neighborhood of peak values and the occuri enceof low values over a larger area of the specimen. Cleavage fracture was accompanied by a less Efficient s~ress distribution in the plastic range than shear fractzuw. 6. Tests of structural elements resulting in shear fractures would not predict the stress and strain energy distribution of identical elements undergoing cleavage fracture. 7 0The addition of reinforcement brought about a more uniform plastic stress and energy distributions with smaller

-30- differences between the extreme values. 8. The analysis of the stress distribution both in the elastic and the plastic ranges substantiated the theoretical assumption of no change of principal stress directions in the transition from the elastic to the plastic range. 9. The applicability of both unit energy and plastic stress methods of analysis was established in the use of plates with welded reinforcement. X. ACKNOWLEDGMENTS This investigation, at the University of Washingtonl sponsored by the Ship Structure Committee, is in progress in the Structural Research Laboratory of the Department of Civil Engineering, of which Professor R. B. Van Horn is head. This research program is directed by Dr. R. A. Hechtman~ Associate Professor of Structural Research. Dr. ).Vasarhelyi~ Assistant Professor of Civil Engineering$ the project engineer, was assisted by Mr. Y. T. Yoshimi, Mr. Robert McHugh, and Mr. P. Roy Choudhury. The authors express their appreciation to Mr. John Vasta of the Bureau of Ships~ Navy Departments Dr. Finn Jonassen of the National Research Counci~ and Dean H. E. Wessman and,.. Professor F. B. Farquharson of the University of Washington for their suggestions and encouragement.

Z.BBLOGRAPHY 3. D. Vaw%alyi and l. A. Hechtman~?WeldedReinforcement of Openings in Structural fitieelmembersrtqfirst Progress Report, Ship Struel!wm Committee, Serial Number SSC-39, s December 2951. ~. M. Greenspan9f~ffect of a Small Hole cm the Stresses in a Uniformly Loaded Plate;fQuart. of Appl. Math.$ April 1944

-32- H19N211 N3WlCJ3dS 679. - 94 3CU... 4 9NTQ! 40 NOCGMC H19N31 N3W193dS,,O-,L, 1,,02 ae - (u,,0 *,,91 - b 9N11OM 30 NOCEW.

A SPEC. NO. 4 L ~ R SPECZ3. NO. 37, 3S, 38A kifff ~ A A v D= 9 F 1 0! n SPECS. NO. 95, 96 /// ///// ////////// Fig. ~~0. 4, Details ofspecs. No.95and96. Fig.3. Details of Specs. No. 4, 5, 17, 37, 38, and 38A.

SPEC. 69 L 6 ~ x -d SPeC. NO.69 M!: mit l+il- +-l-t-t + GAGE POHTS SCALE; l=. 6- eaom@ SYMMETR{GAL ABOUT vatcal GEMTERLME i * Spec. Mo.70 ~ 7 LEGENDS SR- 4 STRAN GAGE TYPE A -7 F SR-4 STRAN GAGE TYPE A-12 + SR -4 STRAN GAGE TYPE AX -5 ---- SLDE - WRE STRAN GAGES ti Fig. G. Arrangeme.it of Grid-Wire Gages for Spec. No. 69. Fig, 5. Location of SR-4 Gages and Slide-Wire Gages.

,.., al -..

-36-1000..- 800 Cn ~ 600 G g n a q 400 200 E!-,,. > 7 -==U--L. ~ SPECMEN NO. 96 6 0 0.6 0.9 1.2 1.5 L8 AVERAGE ELONGATON OVER 18 N. GAGE LENGTH N N. 2.1 2.4?ig. g. Load -Average Elongation Curves, Specs. No. 37, 38, 69, 95 and 96. 1500 1200 # g 900 z n 4 3600. A SPECMEN N~ 70 SPECMEN NO. 71 1 300.<. 0 0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 2.4 AVERAGE ELONGATON OVER 24 GAGE LENGTH N N. Fig. 10. Load -Average Elongation Curves. Specs. No. 70and71.

SmeGo N9. 6.9,,,,,,:,,,,,,.,,!,,?,!, Fig. 12. Photo~raphs of Plates after Failure. spec. No. 95 and 96.

3$J- U.. (n t % h (n r ---+--1-+-+-$ - ; -t % t-- -F 4.n v x ra U 0-) u-l > A 1- O-J v m..

-39- Y 0 w)). 09 ~, -- 4--... / A/ ---..., -..,+ \ 0.90, --.,. \ \ \ ~ ~.. \ \ *-------:\.... \ /A 04-.. \ -: Y ~ ~,... ---- e -- 0 -- -i \ 0,20 0 SEE FG.10(a) Fi:. 17. Eastic Unit Str3in Concentration Contour6 FiE. 15. Stress Concentration Contours in >--Direction for Plate for k71ate of Finite Wi5th with Circular Openiim of nfinite Width with Circular Hole by Elastic Theory. Plotted from SR-4 Strain Gage Readin~s. Spe~. Plate (a). No. 69. 1: *\ / 0.... x SEE FG. 10(b) SEE FG. C(C) Fig. 16. Stress Concentration Contours in ydii-ec~io~ forplate of Finite Width with Circular Hole F!~.10, Stress concentration Contours in y-jlrection fop, by Elastic Theory. PJte (b). Plate of nfinite W idth with Square Opening wi{h Rounded Corners by Elastic Theory. Plate (c).

-40- Y Y -- x 0 m 0 Fig. No. 20. E astic Unit Strain Concentration COntourS for Plate of Finite Width with Square Opening. Plotted from SR-4 Strain Gage Readings. Spec. No. 95. 76 F. Y -. k R --=% #i N N - x Fig. i9. Elastic Unit Strain Concentration Contours for Plate of Finite Witith with Square Opening with Rounded Corners Plotted from SR-4 Strain G age Readings. Spec. No. 38A. 1.0 x! R+ la Fig. Jo. 2i. Elastic Unit Strain Concentration Contours for Plate of Fimite Width with Square Opening. Plotted from SR-4 Strain Gage Readings. Spe.F. io. 96. -46 F.

x 00 G w E w u 0 0.91 4 4 z T +s1 =kl

Y i t t o.soh,~f/i. \ OF OPENNG. & N 1. UNT STRAN CONCENTRATON FROM SR-4 GAGE READNGS. SPEO. NO. [7 i.42max, N BODY PLATE. 0 -.. THEORETCAL STRESS CONCENTRATON FOR SEE FG. lj(e) A PLATE OF NFNTE WDTH. PLATE (e) Stress Concentration Contours in y-dire ciion for Plate of[nfifite Width with Circular Openi~ with nsert Plate Reiuforce~ent by Elas*ic Theory. Pate (d), ~ig. 25. Elastic Unit Strain Concentration Curve for Plate of Finite Width, nsert Plate Reinforcement. Spec. No. 17. 1

300,3.23 3,23 / \ 3.CQ \ \ Fig. 26. Stress Concentration Contours for r, for Plate with Circular Opening for Load of 65o kipsj 77 Per Cent of Maximum Load. Spec. No. 69, 76F. 2.00 \ /, \\ L f!.20 2.10 \ \ 9_.00 1.*o.BO \ < w 1,70 \ / /,, \ \ \ #/,/ \ \ @ \ *%/ \. 80 $ $ik Y/ Y > \,40 (~0 ~., / - :-*o 1 =5 x 0.9 4 ----.......,....................... - - - FRACTURE > BY ELASTiC THEORY. 0.70 - - - x FOR LOAD OF 650K!= FoR LOAO OF S45 Kl~ {MAXMUM) Fig. Z8. Cmnparisonof Elastic and Plastic Stress COnCentratlOn Curves for G for Plate with Circular Opening. Spec. No. 69. Fig..z7. StresE Concentration Contours forff, for Plate with Circulal Opening for Load of 845 kips, Maximum Load. Spec. No. 69, 76 F.

\ \ \ =-, y) --0,s0- // 1,,s, #. x L- --- L- - -x Fi<.. 29. StrefiS CO,lC=:,tratiOn COJtOl.r~ fjr Cisfur Plate with square Opening, Rounded Corners. Load of BOOkips, Maximum Load. Spec. No. 37, 76F. StTess Concentration Contours for~ior Plate with Square Opening witq Rounded Corners, Load of 915 kips, Maximum Load. Spec, No. 38. -20 F. 3.00! 309 3.0$ 3.00-1.5C 1.0 1,,,, ~ l 3.09 3,W ~ 3 3,00! i 2.00 L2.00 --/, _o~ -o o, 2 L %*A,42!2,.-...>_,/ ----- - ----,00 r.oo O+-+=:-- -------. -.7,0. - - - - - - -... ~. W -~;v - r --BY Eb%STC THEORY o +OR LOAO OF S00 KtPS (MAXMUM] Fig. SD. Comparison of Elaatic and Plastic Stress Concentratim Curves for Plate tiith Square Opening with Rounded Corners.. Spec.?Ju, SJ>?,fiF- t \ 2.00 2.00,*, 0- -0-+ e, \ o~o o,, \ l., ),/ ~... // ----- { - K? -----.- - --- -------- )7 1.00 1.00 : E {.........,-.4?::: 0 w- 0. x / ---BY EwSTC THEORY.............. -. ~;iizii;..-...--. O -+OR LCMD OF 915 KPS lmaxmum)?7$. 32. Comparison of Zlastic and Plastic Stress Concentration Curves for ~J fm Plate with Square Opening with Roknded Corners. Spec. No. 38, -20 F. )

.-. ~+- - - - ---w.,.,... \,,~ -., \ \ A +. o 1,.4g-,.-,#- 1, >.-. / ; -- - +p- -- ---- %. 1 \o 1.,. ~ ---- ---- 0F2 \---- -. /.,.f / // /.+/,#,9...

( l 1.0 1.,? { 2 ( ( -3.0 T. -.. - ~- [- _._ x =5 Fig. NO.36. Stress Concentratiori Contours for Wy for Plate with Square Opening. Load of 648 kips. Max. Load. Spec. No. 96. -46 l?. 1s Y 4.0 1,.85L- 0,8C * 3.0 d 2.0 &* 0 da 1.50, k----30 \ o.= 2.0 t -..soq ----- o ---o-a =RAcTURE...-.+..,.................-... 0 52 ).05 - - -A- - - - -. x. ~ FRACTURE.+,++...3........... i4------ * FOR LOAD OF 648 KPS (MAXMUM) -- 0-- ELASTC UNT STRAN..y...:......(*T~)..................... FRACTURE..) ~~g.no. 37. -. Comparison of-elastic md F lastia Stress Concentration Curves ior ~~ for ~~atfi with Squars Opening. Spec. No. 96. -46 F.

l-- s 1, o. \ % \ - a

~., *.! + - - J= {..$s 1 H :3: -,. k =m ~ - - -. - ---. -. G T -Y F4 \ n%?? * --- 6UW..<\ \.-.+--2Z%U,;+ / 0 _? \ 0-;- ( %>.., \ \

J 5.0 u) u) w : ~ 2.0 0 50 60 70 80 90 Km PERCENT OF ULTMATE LOAD v 60 a id a 40 ( ( -46 F, CLEAVAGE 100% SHEAR O% { Sl%c. No.70 76 F, CLEAVAGE #a SHEAR 50% F~E.4d.Stress Coflcentrs}io~ l?apor inpladicflan~~ asfailura is ~~~roache~. ti~ o 0 0.4 0.8 1.2 L6 2,0 2.4 STRESS CONCENTRATON Fig.Ab. Effect of Testin~ Temperature Upon Plastic Stress Distribution at Maximum Load.

Y 1 % x Fig. 46. Unit Strain Energy Contours for Plate of Finite Width with Circuiar Opening for Load of 650 kips, 77 Per Cent of Maximum Load. Spec. No. 69; 76 F. Y Fi~.4&. Unit Strain Energy Contours for Elate of Finite Wic.th with Square Openin:, for Loai of 575 kips, 81 per Cent of the Load. Spec, No, S5. 7& F.,. ~ ----- x -. - - - f=~ Fig. 47, UnitStrain Energy Contours for Plate of Finite Width with Circular Opening for Load of 845 dps, Maximum Load. Spec. No. 69, 76 F. Fig, N0,49. Unit Strain Energy Contours for Plate of Finite Width with Square Opening, for Load of 710 kips, Max. Load. Spec. Mo. 95. 76F. 1,.,!,

-Q- J-J 1, + \\ %. : ~.1 \\,\ \ :./ + 000. -.,<~%e.,. <---,., - ~ >-.., /- /,,!,, // /, i!,$., _,....,, -...- _......-..., ~...<.... -- ~,,~,,-3A#., /. \\, (,,~,.. -.. i., /!..., _ool,,-... 0002.,,--

i x Fig. 53. UnitStrain Energy Contours [m Plate OiFinite Widthwith Reinforced Square Openingswith RoundedCorners for Load of 1150 kips 98 per cent cd Maximum Load. Spec. N.. 71, -46 F. Y x Fig. 55. Contours of Equal Rate Of Energy Ab. orpticm. Spec. No. 69. 76 F. -x Fig. %. UnitStrain Energy Contoursfor Plate of Finite W,dthwith Reinforced Square Opening with Rounded Corners for Load of 1176 kips. Maximum Load, Spec. No. 71, -46 F. :,.,,)

,.. -\k /----------- Wb--,-.. \ \\ l.., \.. \ \. \, \\ \\ o. ~ / \ \ \,,. h. \ \ \ \\ $, L,, \ f,,.,~!, \\\ \,, )1 (.- 1,X, m,,, P J + ~k \,::,,, \ ~., ~\ \ \.\. \,,. \, s \,:.. $., -..\@,\ Y..~_...: \,\... J,.0 L \ \\.+ %,:;., 9 ~,., 21 -.-.... ii \\\ \. --------- \l \,} \ \ 11? 1:, :,J, -+-i,--- : :;,, /t 1,, *,,/ 1.:,,0 ;,,,// 4 : ~ti /,/ j,/ / 0 /,/,, /,.. ~,/,(,/,{,-.,., /,/,/,, / %,, / - ; - - L -- -l-. g,,- c--u :. - - - - & - t ---- - --> +=>. 1 6J,.-.- ~.+.&-.-, -/- / ~=/. 2 1 / / /, 11 * { /// 0,/// {..,1/ f) /~ *j,/ /,,/,, // ;;;$O LA, /,, /.- - /,,0?, / / \ / Oe,, -\ 0- \ \ n,, \.

\ L o-. \ i, i \ \ \ / \ \ \ 1 / --0.s0 / \ \ d \ -.,;,P - \ \, ---0.9 0.95-> k\ 1{ / 07.-.. QS5, \ \ / -x Fig. 60. contours of Equal Rate d Energy Absorption. Spec. No. 7(11 76 F. Y c.1 log 0. 0 x = 5?ig. 61. Contours of Equal Rate of Energy Absorption. Spec. No. 7,1, -46 F..

,! f.: [: LOAD Z650KPS LOAD= 1000 KtPS /-1 in [ \ t \ \ \ \ 2000 Ut#tT ENERGY ABSORPTON ltd. LB.N3 SPEC, NO. 37 -- -SPEC, NO, 38 MAX. LOAD / o Z!ooo 4000 o 2000 4000 UNT ENERGY ABSORPTON N. LMN.3 SPEC. NO. 70 --- SPEC, NO. 71 MAX. LOAD < id a u o u :40 L5 k 1- Z -20 F, CLEAVAGE 91% w o SHEAR 9 % m: f?o w1 -Li%--tr. ~76 F. CLEAVAGE 0% - & : 60 K /) u : w < 40 1 w -46 F, CLEAVAGE 100 /! L o { SHEAR 0% b zd \\ o \ ~ 20 w n UNT ENERGY ABSORPTON N. LB. /ln.3 / / / - -. --- -- 0 2000 4000 6000 aooo 10000 UNT ENERGY ABSORPTON N, LB./l N.3 Fig. 62. Effect of Testing Temperature upon Plastic Specs. No. 37 and 38, at Maximum Load. Energy Distribution Fig. 63. Effect of Testing Temperature upon Plastic Energy Distribution Specs. No.!70 and 71, at Maximum Load.

-56- U -46 F, CLEAVAGE 100% < SHEAR OVO 60 1 \ i l\ i L 0 1-2 d u 20 o \ o \ \ \ \ \ 2000 4000 6000 8000 10000 UNT ENERGY ABSORPTON N. LB, /ln3 -! Fig.64. Effect of Testing Temperature upon Plastic Energy Distribution at Maximum Load. Spec. No. 95 and 96..-

\f J V 1 m to m 1 1 1 1 t T t & -4.: ~ 1-1 D i 3 t ui % 60 70 80 90 100 PER GENT OF ULTMATE LOAD Fig. &S. Stress Concentration Factor, MaYimum Strain and Maximum Unit Energy as Fracture is Approached. Fig.66. Stress Concentration Factor, Maximum Strain snd Maximum Unit Enery; as Frachure is Approached.

-58- APPENDX A CALBRATON TENSLE TESTS The data of the calibration tensile tests are given in this appendix. The details of the tensile specimen, the procedure far these tests, and the application of the resulting data to the plastic energy and stress computations were described in the Second Progress Report (2). The test results are summarized in Table a and plotted in Figs. la-ka..