AN SNVESTIGATCON OF M fat flu Dow» M M. 5. W STATE COLLEGE Warren Jay Eding 999: 'FORGENG-HAMMER ANWL FAILURES. “ |ll|l|l|||lll|ll.ll" AN INVESTIGATION OF FORGING-HAMMER ANVIL FAILURES BY Warren Jay Eding A THESIS Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MAS TER OF SCIENCE Department of Metallurgical Engineering 1953 lilllllll' 6/19/53 Chapte r TABLE OF CON TEN TS I. Introduction to Hammer Anvil Failures .......... II. Selection of Testing Equipment III. General Test Procedure IV. Electrical Strain Gage Tests V. Metallurgical Investigation of Base Material . ..... VI. Conclusions and Recommendations ............. cg: x...) F1199"! ,- ,r C r.“ 12 19 49 60 63 64 CHAPTER I INTRODUCTION TO HAMMER ANVIL FAILURES One of the major problems encountered in the forging industry is the cracking of hammer base anvils in the region of the radii of the sow-block seat. These cracks generally originate at the front or back face and progress across the anvil and down the base to complete failure. The hammer base is twelve feet long, five feet high, and four and one-half feet wide, weighing approximately fifty tons. Figures 1 and 2 show several typical base failures. The sow—block, which is a separate portion of the hammer that supports the forging dies, is held in the anvil seat by a wedge type sow-block key. This key is driven in place by an air powered key driver. Figure 3 shows a 5,000 pound hammer with sow-block and key in place. This static stress, developed by the wedging action, in addition to the cyclic dynamic stress during forging, produces the combined loads causing failure of the bases. The purpose of this study is to develop a test procedure and method by which the magnitude of the imposed stresses can be eval— uated, to determine if the base material is metallurgically sound, and Figure l. ‘9 Typical hammer base failures. Figure 2. Typical hammer base failures. (R \ . . ‘H g‘owtrdua: u... Five-thousand-pound forging hammer with sow-block and key installed. Figure 3. to establish whether the imposed stresses are within the metallur- gical specifications of the hammer base material. This thesis is subdivided into sections concerned with selec— tion of equipment for evaluation of imposed stresses, general test procedure, electrical strain gage tests, metallurgical tests, and con- clusions and recommendations. The extreme expense involved in replacing a broken hammer base in addition to the productive loss of the equipment makes the solution of this problem of prime importance. A complete analysis of load distribution and intensity together with metallurgical data on present forging hammers will make it possible for both forging hammer manufacturers and users to incorporate metallurgical and design changes necessary to increase the operational life of present and new forging equipment. The experimental data for this thesis were obtained by the unlimited use of facilities at the Oldsmobile Forge Plant, Lansing, Michigan, through the courtesy of Mr. Thomas E. Darnton, Manu- facturing Manager. Appreciation is extended to Dr. Austin J. Smith and Dr. Roland T. Hinkle for their excellent suggestions and gui- dance throughout this investigation. Grateful acknowledgment is also due the personnel of the Oldsmobile Forge Plant for their cooperation, especiallyl Mr. Jack Gannon for his assistance in photography and rep roduction of graphs . CHAPTER II SELECTION OF TESTING EQUIPMENT To evaluate the imposed loads under forging conditions it is necessary to develop a nondestructive test method which is reliable and applicable to the severe conditions encountered in a commercial forging plant. The resistance wire electrical strain gage was se- lected as the most practical tool for study of this problem. ~ The resistance wire gage is composed of a grid of fine wire cemented between two sheets of treated paper. The basic principle involved in the operation of the gage is that, as the wire is deformed in either tension or compression, the resistance of the wire changes. This change in the resistance of the wire is, then, a measure of the strain to which the wire is subjected.1 In selecting instrumentation to use in conjunction with the strain gages“ several requirements must be considered. These requirements are portability, static strain indication, dynamic strain ‘M—V For a more detailed explanation of the wire resistance strain gage, see A. V. de Forest and H. Leaderman, ”The Develop— ment of electrical strain gages,” Natl. Adv. Comm. Aero. Tech., Note 744. indication capable of recording strain waves within the base which exceed five hundred cycles per second, and means of accurate re- cording of strain indications for further calculation. To meet the static strain requirements the Baldwin—Lima- Hamilton SR-4 Type "L" Portable Strain Indicator was chosen. This instrument reads directly in microinches per inch, with no correc- tion factor when used with a one hundred and twenty ohm resistance strain gage. Compactness and ease of operation were other deciding factors in the choicevof this unit. The dynamic strain instrumentation selected included the Ellis Associates Model BS-6 Switch Balance Unit and the Model BA-Z Bridge Amplifier coupled with a Du Mont Type 304—H Cathode-ray Oscillograph. This system enables single channel readings of one to four arm bridge circuits to be taken with recording rates of five to twenty- five thousand cycles per second. Accuracy of five percent is main- tained. The Model BS-6 Switching Unit allows individually balanced resistance components in six channels and switches one channel at a time into single channel equipment. It can also be used with the SR-4 Type "L" Portable Strain Indicator. The Ellis Model BA-Z Bridge Amplifier incorporates a calibration system and chopper circuit which makes it possible to throw in a zero line while re- ceiving dynamic signals. The Type 304-:H Oscillograph is equipped with a triggered sweep and long persistence screen for ease of visual observation. For permanent recording of dynamic readings, the Du Mont Type 297 Oscillograph-Record Camera was selected. This is a Polaroid-Land camera with multiple frame exposure, which makes it possible to determine what signals have been accurately recorded at the test location. In addition to this recording equip- ment a Simpson Multi—meter Type 260-RT was used‘to check gage circuits and resistance. The selection of SR-4 strain gages is dependent upon the specific investigation at hand. A wide variety of types and sizes of~ strain gages are available from the Baldwin-Lima—Hamilton Cor- poration, and each has its advantage for certain applications. The necessity that gages be placed in the restricted fillet areas of the anvil seat, resulted in the choice of the A-7 and C-7 type gages. The A-7 gage is a wrap-around type with a 0.250 gage length and 120 ohm resistance. The C-7 is of similar type and dimension with 500 ohms resistance. The C type gages are unstable for static indication but have a higher resistance chawge for given elongation than the A gage. 10 With this choice of gage it is possible to place gages directly in the fillet of the anvil beneath the sow-block. The first two tests on the hammers were run using both types of gages. It was found that the amplitude of the dynamic indication was sufficient to per- mit the use of A type gages for both static and dynamic indications on subsequent tests. This procedure saved considerable time in application of gages and preparation prior to testing. This collection of equipment provides a complete testing unit capable of recording both static and dynamic strain loadings over a wide range of application. It can easily be moved to the test location. and, combined with proper gage application technique, will provide accurate results. Figure 4 shows the combined strain gage ins trumentation. Figure 4. Strain gage instrumentation. ll CHAPTER III GENERAL TEST PROCEDURE The successful use of electrical strain gages is dependent upon correct application of the gage to the structure being tested and correct wiring procedures from gage to strain-recording instru— ments. The use of strain gages to test steam forging hammers dur- ing operation presents many special problems. Surrounding conditions in a forging shop require careful waterproofing techniques to protect strain gages and wiring from the steam, die lubricants, and dirt around the hammers. Extreme care in preparing surfaces for strain gage placement and complete shielding of lead wires from electrical interference are also a necessity. It is the purpose of this chapter to describe the procedures and techniques which were found to be most successful in combating these problems. I. Surface Preparation. a. All pits, scale, rust, et cetera, are removed from the gage area with a hand grinder. b. The surface is degreased and cleaned with gauze saturated with carbon tetrac hlo ride . 13 C. Metal conditioner is brushed on the surface and wet—lapped with silicon carbide paper. (1. Residue is removed from the surface with tissue. 8. Metal conditioner is again applied to the surface and wiped dry with tissue. f. Entire gage area is cleaned with gauze saturated with Acetone. g. The gage area is dried with a small hair dryer for two minutes to remove all moisture from the surface. Cementing. a. Trim gage paper to desired size and place a small quan- tity of Duco Cement on forefinger. b. Place gage between thumb and forefinger and soften for forty-five seconds. c. Place a small quantity of Duco Cement on prepared sur- face. d. Press gage into place with forefinger, squeezing excess cement from around gage edge. e. Gently apply pressure with pencil eraser to grid area and force grid into intimate contact with the surface. f. Gently work out the excess cement from under the entire length of. the gage. IV. 14 After the gage has dried for approximately two minutes, apply a smooth even coat of cement over the entire gage area. Gage should be allowed to dry for two hours before attaching lead wires and twenty—four hours at room temperature be- fore using. Le ad Wire P reparation. a. Clip the lead wires of the strain gage to three-quarters inch length. Us‘e Beldon 8014 wire or equivalent for connection from gage to terminal board and cut to length. Strip ends of the lead wire and tin both gage and lead wires. Clip ends of tinned lead wires to one-eighth inch and solder to gage wires with a lap joint. Le ad Wire Ins ulation. 3.. Bring .both lead wires back over the gage and hold in place with Cellophane tape.“ Place one inch of Mystik tape at lead wire end of gage so it overlaps the insulating paper flap of the gage. Bring one lead wire back over the gage, forming a small loop running the lead wire out from the gage area 15 perpendicular to the gage axis. This is necessary to pro- vide sufficient clearance for installation of the sow—block. d. Cover with one inch length of Mystik tape and repeat the procedure with the remaining lead wire. e. Cover with Mystik tape and coat the entire area with Duco Cement except where tape joins the gage. Electrical shielding of the strain gage and lead wire is ac- complished by placing Aluminum Foil tape over the gage and. wires running along the anvil fillet. Wires outside of the anvil are wrapped with this tape from the anvil to the terminal board and grounded to the common ground terminal at that location. Care must be taken to prevent breaks or splices in the shielding, and it must be grounded to a common ground terminal. This terminal is in turn grounded to a fixed ground source in the vicinity to prevent pick-up of electrical disturbance-in the amplifying units. Gage installation is completed by waterproofing the gage area and lead wires on the anvil with three coats of Synthetic Coating Type A-l78-‘B which is produced by the B. F. Goodrich Company. After the second coat of Synthetic Coating, a one inch piece of Koroseal tape is applied over the gage area and serves as additional mechanical attachment and waterproofing. Petrosene wax was used on the first test but did not offer sufficient protection against the steam and oil from the hammer. A special terminal board with provisions for twelve strain gage connections and two compensating gages was constructed. This board was located at the hammer and connected to a switch box at the instrument location by Beldon 8424 shielded microphone cable. The switch box at the instrument location was constructed of micro- switches to contain two series of six channels each, making it pos- sible to obtain readings from twelve strain gages with the BS—6 Switching Unit. Shielded microphone cable and micro— or radio switches must again be used to prevent interference pick-up in the amplifying units, and the shielding on the cable and the individual switches must be grounded to the common ground terminal. . Instrument operating techniques and procedure varied little during individual tests and a general description follows’. A two gage circuit was used on all tests, composed of active and compen- sating arms. The compensating gage was mounted on a section of material identical‘ to the base material and placed in the vicinity of the sow-block to compensate for temperature changes at the active gage locations. During static strain readings, an initial reference reading was taken and subsequent readings after each key blow. No calibration 17 is necessary during these readings because the Type ”L” Strain Indicator reads directly in microinches and the increase in strain is computed by the difference in successive gage readings. During dynamic strain readings, the amplitude of the impact signal was adjusted within the range of the oscillograph screen and the calibration switch set to approximate this amplitude as nearly as possible. Record of the calibration was made by sensitizing the film to bring out the screen graph and then photographing the cali- bration trace on the screen. The procedure for recording the actual impact signal was the same with the exception that a zero reference line was thrown across the screen with the chopper switch on the BA-Z Unit, and the sweep control on the oscillograph was set on triggered sweep. This procedure produced a photographic record of the calibration and impact signals with a graph and zero reference line in the background for simple evaluation of true strain values. Figure 5 pictures a typical gage installation on a five thousand pound hamme r. Figure 5. Typical gage installation and hammer terminal board. 18 CHAPTER IV ELECTRICAL STRAIN GAGE TESTS A series of four individual tests was conducted on two differ- ent forging hammers. The hammers used for these tests were five thousand and fifteen thousand poind steam hammers, Numbers 503 and 153, respectively. As each test was completed, certain points and questions for further evaluation presented themselves and were consequently investigated on following tests. Each test will be de— scribed and analyzed individually, and a collected discussion of the combined test results will follow. The first test was conducted on hammer 503 to determine the magnitude of the static and dynamic stresses imparted to the hammer base anvil in the region of the fillet radii. Strain readings were taken after each blow as the key was driven, during die heat and for the forging blows producing maximum stress signals. Figure 6 shows the gage locations during this test. The static stresses imposed by driving the key show an in- teresting relationship between key travel and key fit. As the key was progressively driven and readingstaken, it can be noted that the 20 I ~ 3‘ , .. I .- /: [I , I / . 9 >1; 7 #503 flA/M/l/f/P Figure 6 21 bearing area between the key and anvil definitely moves. By refer— ence tovaigure 7, it can be seen that as the key is driven, the stress at certain of the gage areas increased much faster than in others and in the early stages of driving actually decreased in some locations. This nonuniform loading would indicate poor key and sow—block fit. With a very accurate fit, the bearing area would remain constant, and the stress rise would also be constant. The actual numerical value of the static loads was much higher than had been anticipated, and the very high stress of 30,000 p.s.i. indicated by gage S-Z, which approaches the yield point of the base material, make the accuracy of this individual gage questionable. _However, certain important factors must be considered before elim- inating completely the indications of this gage. The gages S-l, S-Z, S-4, and S-5 are all located in the fillet radii, where it must be assumed the stress will be considerably higher than on the face of the anvil.. The close similarity of loads for gages S-4 and S-5 and the over-all continuity of the readings further strengthens their validity. Noticing that the readings for gage S-Z were excessively high, an attempt was made to verify the readings at the location of S-Z by removing the key and replacing a new gage in the same loca- tion. However, this could not be carried out because the key had ZZ (f‘ I / V 27 I :7 i' '1 AWL-W ( ,1/ . I l l' / (A '7. E)” +H- ' "-- v' / ~“'. - III II»- Ill 1 ., - . 1 . J. . . _ 3/. a... 1. . - - : 12/qu .. \ l _ I r. r I 4 _ , _ a. nil-2:34:11. 2|. 1.III,'I Z7” 2:” 71.6" K J k/ ' fire/F, “I“. 10 .4 0/41 .. 2.; /’/J / . A Y _ /, I / s . II a ) -- --+l.|.l.LJIi_IIu i y, u . _ 0v; . _ x _ p? r I . V {A '1 A, . I r . . I _ / ‘III ll 1-1!! lilo lllivll . \ l» ..... r _ . ‘IL . / . _ _ L / // L p I 1*. . cu _ v a _ . z . . . ./ ‘ Iiil'l_lvt i II I \ .ll :2”? /(_). _, ,,4',"/’ ,7 #9b 3 V L. ’v- /'.,'/L'./'"—/ I 1 / 60C QVQV bxk Vke U .. (7th.? ‘/ Ix / 0,16) ()0 23 become wedged so tightly that it was impossible to remove it. This in itself confirms the high stress imposed on the base. On closer examination of the key, it was found that it had raised seven—eighths of an inch on the front side. The only way possible for this key to raise without backing out was by imposing an additional stress on the anvil while moving upward. If the original stress on the key was near the yield point, plastic deformation, or permanent set would certainly take place which would then allow further repetition of this cycle. Checking both key and sow-block fit, feeler gages of from 0.004 to 0.019 could be .placed between both key and anvil, and the side and bottom of the sow-block and anvil. This nonuniform seating of the sow—block on the anvil would account for some of the sow-block failures which have occUrred. Considering these factors, the accuracy of the gage S-Z is perhaps not as questionable as may at first be imagined. Nevertheless, elimination of gage S-Z still gives a residual stress of 19,000 p.s.i. which is far too high in view of the dynamic stresses involved. Uniform bearing of both sow-block and key on the anvil should lower the residual stress necessary to hold these units cor— rectly in place. Dynamic stresses recorded were very uniform as can be seen by reference to Figure 8, and the maximum stress produced was A. /. /' (5 ”/7; 457- z 5.4 fl / //L/ /- 24 8/54/97 2 725.5711 ”1/846 7 4 a4 0.9 fld/P/A/é FO/Pé //V6 424/ c} /V/6/‘/7' «EMF/1f #503 r‘ -. - I . N ._.-.___--.._._...._. -M... -........1 -. ._ .. / '! V VJ p» —- ~ I r'FIw-H—_--—- -.._ (15/ 7 1 I . 2.2900 9- *1; .-- 4. , . / ’ : r-fiL ' L i /éi§ "LU .443“ 5’0 I -a—A «~- 1 n - . . ‘ _ ..- .__........-._-__._...._..L_ /‘£.’I56)! -__- *‘fly—f ~-.—.-- . u. . 2’ 49.200 9-~-- ~---- ---'-~“- “r“ r I / _.- “II . 4..., 0 j~9————~9~--— , __,. i ' 0 03 4.3.90 r———- - ——-—-~-—-~—9.—~~-- 'r - --,._.: 1' £)/J:00%_«_--_ I I r I i i l C 2 4‘ If 7 6 Eda/V /V///W5£_/P .03.. SAC/f 4/", ,f '5 ’ --- -. v a A]. —% LI .L --—_-.. _ .. -fi ..__ A ‘3' - -- — —- -— --—- A, [3“; 0 4:”, A{ 0—0. x +» w. a - I uo —— — _——-- Figure 8 25 35,000 p.s.i. These values of maximum stress occurred on the final blow of the finisher where there was a very small amount of flash area, and the energy of impact is almost completely absorbed by the dies and base. These dynamic stresses are again higher than necessary be— cause of the localized bearing areas through which the dynamic load must be transferred to the base. Uniform load distribution should lower these stresses a considerable amount. The poor bearing sur- face was observed when the sow-block and key were removed from the anvil before installation of the gages. Definite galling at local- ized areas could be seen on the anvil face, caused by heavy loading at these spots and movement of the key. No great difference in dynamic loads between day and night shifts was observed. IThe main difference was that on the night shift the hammerman used an extra blow on the second impression which subjected the base to one extra cycle of 20,000 p.s.i. for each forg- ing. It was not possible to obtain readings from gages D-4 and D-5 because steam was accidentally discharged in this areaOand removed the wax protective coating, shorting the gages and making them unstable. Readings on the night shift were taken only on D-1 and D-Z, as the other gages had shorted out during the day shift. 26 No individual forging blow of excessively high impact stress was observed. Maximum stress values, both observed and recorded, varied little from average values. The results of this test indicate that the effect of key and sow—block fit on the imposed stresses should be investigated more thoroughly and that Petresene wax does not offer sufficient water- proofing protection for steam hammer testing. Test number two was conducted on hammer 503 to determine the effect of improved key and sow-block fit on the static and dy- namic stresses imparted to the base. Gages were placed in the lo- cations shown in ‘Figure 6. Strain readings were recorded during the key drive, as the dies were brought up to heat, and during the forging operation. A very marked decrease in the static loads was found in this test. Referring to Figure 9, the maximum load encountered was 6,150 p.s.i., which was only 20 percent of the maximum load in the first test. The number of blows required to drive the key was thir- teen, as compared to twenty on the previous test; the progression of key travel was very regular. The fact that this key fit was still far from perfect, even though greatly improved, was shown by the decrease in load by all 27 07/1/4971 7—!qu filiI-II-Wl-i I-.. M _ _ h . p D K my I vii IIU iiiiiiii \J 1) 1J4... _ _ 3&3” .\ \ ./ .7 \\ . _ I -I -é.y.n..¥. --\- -.I . .... I I - . ‘1‘ LIVIVII‘II. t _ _ r ....., «I... Irlfr. _ - _ .I. ...-.-.m..-J.?I1L i. . . ./ 2r _ _ m A . .. TIE-ii. . Elli-i .. i _ _ . 1 _ - all-III-- - .. .II -I -iI II. n . r N F . _ n a _II, - - _ _lIl-II- WI - -- -w I I ._ c- _ a . _ _ _ V .I ii-IIIII-I-l _ -I-I i m .. It ---I I _ _ _ L i _ _ . . llIlI- .. _- I- -_-- . . m _ _ _ Iiich-I- .fi I. I -. - ill III- TI .- - _. i L _ __ _ - -I- .I- -.__- _ r. , W . -l I , - I- -I-- -JfiI-lI-Il|l_ - - a i m _.III I - II in , II- A III||L - u - w _ H _ . _ - ‘l- I - II _ P III- - .. 4/ a. z /.1. .J. A 4 . . II II IIIr- .i-v‘II-I .Il . . .. IJ .2.— , 1 . . . _ .1/r4 . .... u ._ .. ..I...,..,..H4 / -.. Illa-41.7.. 7,. re "7 ’I' x? T. .‘i’ff/ A7;- “— 9—_—94' ‘ ~~ H-a—‘t-flt‘ (V /O .x; p.\-v Figure 9 Z8 gages on blow five, with gage S—l actually varying between tension and compression throughout the drive. The over all range of varia— tion, however, is much smaller than that encountered in the first test. Gages S-3 and S-6 indicated the highest stress. These gages located on the face of the anvil would point out that with the fit of this key, the wedging force was initiated at the front edge of the anvil, and this was further supported by the readings of gages S—2 and 8-4, which were located in‘the front edges of the fillet radii. A one thirty-second inch shim was placed between the key and sow-block in order to determine the effect of this shim on load dis- tribution. Before readings were taken the hammer was operated for one complete shift and at the end of this time it was noticed that there was displacement of the key. To correct this the key was driven out and replaced. The readings of Figure 10 were taken through this procedure. The key was examined when it was re- moved, and it could be seen that there was a much larger area bear- ing than before, although there was still evidence of localized bear- ing areas. In two areas both the key and anvil were gouged quite deeply, presumably from holes left when a key had been burned out. This was the second time that the key had been driven since the original fit, and the combination of localized bearing areas and gouging ~\ . ‘-- \\ .15 ff ') (I LIL/7 5.“ o 54 A 29 5439/97 Z 7255‘?- Z SIP/474’: Z 0/41) 17(72- Xl/J Xf)’ D/P/V/NG A’cy wz‘f/fi g2 Mr» ’50.? l-- . \ Etw‘w' <9 \ .HW / / . +'—'-r—H— -I—1 -y—— )\ fi-fi—d—r d—<—- \ 772% t ,1?) Figure 10 x . '§\ { h 9 9 r- 11‘. \ 5‘ 5N 30 could possibly have been great enough to destroy the degree of fit that had been obtained. This would account for the much greater loads obtained during this operation. However, the locations of high stress and the general trends were the same as previously, which help to verify the accuracy of the data. Another interesting observation was that immediately after driving the key, gages S-3 and S-6, with the highest loads, indicated very slow relief through localized distribution through the shim. After one hour and three hours of forging, all of the gages indicated a relief of about 3,000 p.s.i. The key did not begin to raise after two days of operation, and it became evident that it is not necessary to drive the key this tightly to hold it. The dynamic loads were also lower than in the previous test as can be seen in Figure 11. The maximum load was 30,000 p.s.i. as compared to 35,000 p.s.i. previously. The individual gages indi— cated a stress approximately 5,000 p.s.i. lower, a definite improve- ment. It was not possible to compare the night operation because of failure of the amplifier unit. Considering the combined static and dynamic loads, an operat- ing range of 29,000 to 37,000 p.s.i. average can be determined. This range, although vastly improved over the previous test, is still at, or above, the endurance limit of the material. The results of any 31 ’G/NG 0/? 4., . f- / -’ / (5 fl J U 7- //i ”/W / (J K‘. 1“. , , . [\J L/ / fig v 2 'iI/ ___--.___ -4. d. I. .4... R941; 29-“ /1- 11/1 *JOJ I / A. ,j/ I (/ / / 'ZJ /_ ,/ // /. I win-9c _ * -—-- ~—— 0 a... Md 20 .2 .J . HI. ., a ... 1 »._..._. ‘ :L) f:- "‘ 4.. 3. Figure 11 .47) 2 32 one test cannot be cbnsidered conclusive. However, if this degree of improvement can be obtained with the small amount of care that was taken to fit the key and sow-block, a corresponding improvement should be.possible with a perfectly machined fit. It becomes appar- ent that it is necessary to remove the human error involved in hand grinding the anvil base, and substituting mechanical control to do this job. Test number three on hammer 503 was the last of the tests that was conducted for this problem, but the discussion is placed at this point for continuity in comparison with the earlier five thousand pound hammer tests. Before conducting the test, a detailed study was made of all previous tests and the configuration of the hammer itself, in order to determine the specific distribution and origin of the dynamic loads on impact of the hammer ram. From this study it became apparent that the principal stresses at the fillet radii were being initiated by the effect of Poisson's ratio which causes the sow-block to ex- pand against the anvil sides during the forging operation. Further, if this is the origin of the load on the fillet radii, the principal stress at. this location will be unidirectional and at right angles to the anvil seat. Another factor noted was that the temperature variation 33 in the anvil area produces a major effect on the static stress levels. To get a further picture of stress distribution around the anvil seat, a scale model of a hammer base was made. By use of brittle lac- quer techniques, areas of stress concentration under static and dy- namic loads were located. Under dynamic loads this model indicated a maximum stress area approximately three inches from the fillet, toward the center of the anvil. With verification of these points in mind, the final test on the five thousand pound hammer was carried out. Strain gages were located in the positions shown in Figure 12, and readings were again taken as the key was driven, during die heating, and under impact conditions. Prior to this test, the anvil seat of hammer 503 was com- pletely remachined and a new sow-block and key prepared for in- stallation. Theleffect of improvement in fit is indicated in Figure 13. The increase of stress with each key blow is very uniform up to the point where the dies were heated. After the die warmers were aPplied, localized expansion and contraction reduced the compressive load on all of the gages which were under compression. Evidence that this key fit is still not perfect is shown by 0—2., 6-5, and G-6, Which indicated a still further decrease of compressive loading during ’4 ‘ £334»)? C77: .CA 985/ f/(7/15‘ 3‘ ' / q ‘ v r‘ / A— ‘\ C‘v‘IC‘ _ s.‘ I-./i ‘ , -’ V- —v— A I‘ ’2 __ ’ I I "I“ ' t—fi \ ( 0 ‘x . .3 k . . .‘1 I ’ -- c '1 Figure 12 35 A25)" DAD/r“; xix: W J - ..|1+IIIIm.'LII .~.\\$vfl\ ,vmra (J J 4 v I Q .L.__ , .-._ , Alli 4 fl a II Ifilnilqli. 5‘ 4.? kl” Np ”I?” J .l w ...V .\\\\NN.N\. .Wu _n\ _ , t. ’- I!‘ .\ u I A vii”! . F [7/]; X ill 1r 1} _ .- t¢-WK ,IA._ 9Yi L .2. ..._ a A , L L . - _ _ . a . . . .IIIIuLvIIIIII-14!|Illl.% \- I :4. (k..# :4- ”Li’-.. Ill LT /+~ v.” I '16 , - : .VI ll. I.l \..§n r“- .\\.fl\ \ L . .- L ,L .2. __ +/ L _ \ , 1. H . ~\ . r L . _ .T. I I..- . / .L a... \ Lek“ .L 3.” i? r. L ... T a L a w . Ayn a ...A. L x... A . e . . . ... L x: x.” f \ a? W _ 0 . IJ‘I + w ./» «I xxxkxlnwi 17: 4. L ¢ ._ L ._ . _ .. . . If . , L . , ..r i. .m. -i§LiEe;i.- Liiti-ée.. t L” L L 4" ~ ‘- ‘— '7 1 £21 / I .q— / l 1‘1/4‘1 I]; 1 017/1? II .JILL a . a.; y I- i k.\ L L. A .4/ILLI 4.4L. _ __ \\. i c _W F._—# *4 TI‘MIJ‘ Yl 1:“:- ~\ .3 . l v T- ,,t r . 3‘ fr: \ .1‘1 , L L .. . .x .. . ‘4. l n.1, J abut~l “7 ._« \‘ l.n‘...l| .wu _ - L _. .. _ _fi . «L. I -i a 6734/76 4(24fi-5 Je/ /6000 L _ _a I L _ _ — L 0 C r. A. 'J n A \. .. _L o. A a \. 0 v Ii, 0 a. 1. a. .L a r. a. ‘. 3V. 0 C 0 A... 4, A , A... 3. A. f. A C «k \. M 2 16 .C c .4, A. 2 . o o ,/ / ._ J. . c 8 J n _. “SCAN...“ .afixww-Nku‘exzb . x \ Hid QTQ .4 u\k Fe .0 J 9 , ‘5' Se? 7‘— v c/ .r Figure 13 ”ftiilllfflflf’f’ see—094943— 6.: I as .- 6+ 0 6.1 - 0v... 36 the final tightening of the key. Final static loads with this sow-block and key were lower than those of the original test, which also helps to verify the importance of good sow-block and key fit. Temperature variation over the face of the anvil during the die warming procedure is shown by Table I. Through the original heating stage, temperature variation was very erratic due to the po- sitioning of die warming blocks. After a period of twenty—four hours, the temperatures around the anvil area were fairly well equalized. This indicates the importance of not driving the key fully tight until the temperature at the anvil area has equalized, thus preventing high localized stresses. Dynamic readings, which are shown in Figure 14, indicate a maximum of 28,000 p.s.i. which was slightly lower than readings in earlier tests. The distribution of these readings confirmed the hy- pothesis that the imposed loads originated from Poisson dimensional change of the sow-block. Absence of any noticeable indication on G-5 and G-6 indicates that the dissipation of the stress wave at the anvil seat was widely distributed, and that a stress wave was propagated along the upper surface of the sow—block traveling across the block to the sides of the anvil. This wave was then multiplied by a can- tilever action at the fillet radii, causing the high strain indications at tho se locations . 37 TABLE I ANVIL SEAT TEMPERATURE DISTRIBUTION, HAMMER BASE 503 Temperature (° F.) Time Die Warmers On Sow- Point Point Point Point Block 1 2 3 4 30 minutes ......... 95 75 7O 70 100 1 hour ............ 105 75 7O 70 100 1.5 hours .......... 100 80 70 70 100 2 hours ............ 100 80 75 75 95 24 hours ........... 105 95 80 80 95 Note: Temperature check positions shown in Figure 12 are num- bered counterclockwise from the left side. CH/WT 2 7557‘ J fl/A/A/wc .4 5.4.0.? f/Alll/ffl *503 25000, 240 00 20000 k» / . .9) ((24/90/9555/a4/ E Q» 8 8 E § N (r) ‘b in \0 k (D (.D Q) (D K!) (D 8 SWIM/N 6246/53 61 Figure 14 69 6/0 39 The principal stress directions were determined by means of G-A, G—B, and G—C placed in the anvil fillet. The graphical deter- mination by Murphy's method is shown in Figure 15. The direction of this stress was as anticipated, completing the verification of the effect of Poisson's ratio on imposed loads. The results of this test gave a fairly complete picture of the distribution of imposed stresses in the area of the anvil seat. One test was conducted on a fifteen thousand pound steam hammer toobtain a comparison in working loads and general stress distribution with the five thousand pound hammer. Strain gages were placed in the locations shown in Figure 16, and readings were taken following the same procedure as in earlier tests. Figure 17 shows the build-up of stress in the base as the key was driven, dies heated and at four different time intervals during forging. The procedure in this case was to drive the key only partially, bring the dies to heat, and finish driving the key. However, due to a misunderstanding the key was driven during the second shift and readings were not taken for the final tightening of the key. Therefore, the section on the chart between die heat and the first forging reading represents the total build-up during this f final ke y drive . G .9 ea = SEW/.4” 0A" 64 8,, ssrwlz/ cw GB 8c s S 7434/;1/ 0/1/ 60 1’4 3 1‘: = 45.0 7‘??? ”i +~—-—I 20.70 ,/.J"'(,.1"O-/,1 C .1155 viz—‘2'? ' /1 ICII¥ h CM? can}? {5" 1.2.791) x949. I I C. ”.9, 41 5,7}?ESS A9 {GP/M "/ 2'34 720/1” Figure 15 40 7/53 544,441.41? Figure 16 41 (L, ,.\ r-‘T; \. I / I Q A i‘) r‘ \. TATI S 'CHART I QTATIC LOADS DUF- T I c I: 3 KEY DQ1VIN HEATIN610F‘ DIES; ADID FQ‘RGINQ '* HRIVlMEQ $5.3it _L,_ I l T I ‘ ‘ 1 ILLI- i I I x 2 3 4 o 9*75T/ / 3 4 KEY DRIVER BLows [ Cw; I AFTER I Sit—M I FORGHNG] GAGE. LEC‘EN D q I — —————_ q 2 .3 —k- w} —“-«+~ ~—s—-‘-—‘«—>'—~r f-Vhfi'v-‘fl-W C1. 2- — — —- —- -— — - (:1 _l" ‘ -6\ 11 D ,4). ”- 4p“ (em-4}“5—0-9- (4‘) 3“ “‘01— STIfi BIL—E ’1‘ h' “ .2751; '7 "-fi-DD-EP—(N—Q F.‘8 (.4- L.--L_.-_L-._ 1-- . - g1 ._ VH‘V->~1~Qr“‘ -- Figure 17 43 The initial build-up was quite uniform with the exception of G-4 which was on the outside face of the anvil. This would indicate that with this key fit the wedging force of the key was taking effect on the front ege of the anvil. Gage 5, also on the face, confirms this. When the dies were heated and the key fully tightened, the load was transferred to the center of the anvil opening. This was due to the shape of the die which was located in the center of the anvil. The maximum load was recorded on Gage 2 at the center of the anvil on the key side. The value was 23,000 p.s.i., with the other gages averaging about 12,000 p.s.i. The first dynamic readings recorded while forging a large turbine disc are shown in Figure 18. The maximum loads were noted on the same gages with maximum static loads. Both day and night shifts were recorded with a maximum on the day shift of about 32,000 p.s.i. One peculiarity noticed in this forging operation was that after the first two blows, each blow imparted about the same amount of stress to the base. As to the number of blows, the night hammer- men hit four more blows than the day men--the day men using twenty blows.to complete the forging. After completing these forgings, dies were changed for a smaller turbine disc. The key had loosened during forging in the GHART-Z; DYNAMIC. LOADS ON BAsF. DUQINQ FORGHNG, HAMMERHSS 32000 30000 l I. 28 ZeoooIM —-—- , ’ ‘ \ I 7 \ A24CIOO m (1 2200: 0 Zoo 0 ' ‘ A, J18 ocO YI IoCOO n. ‘4. 0 2| OD D IZOCO ICOOO \" i LEQEND NIGHT—SHIFT DAYSAIFT 3* 4 qaqss Figure 18 a: 0‘5- DYNQMICQILY UNSTRBLJ; 44 45 other die and it was retightened with the results shown in Figure 19. The relationship of the gage readings remained unchanged but the final load reached was much higher, with G—Z reaching 31,000 p.s.i. after forging. The readings, ten to fifteen on the chart, represent five different time intervals covering a total period of three days. During this period the hammer had been run intermittently and this is reflected in the variations on the graph. This variation was due to temperature change in the dies. Gages 4 and 5 which were on the face of the anvil still showed very little increase in load through this final drive. The dynamic loads (Figure 20) on this disc were somewhat higher, reaching a maximum of 33,000 p.s.i. on Gage‘Z. The general pattern was the same as before. In forging this disc, there was no reheat between the blocker and the finisher which might account for the higher stress which would result from working the colder metal. Considering both the static and dynamic loads, forging of the first disc resulted in a working load of 55,000 p.s.i., and the working load on the seconddisc exceeded 60,000 p.s.i. Though it is difficult to compute accurately a true combined stress level, the values in- dicated in this test exceeded the yield strength of the base material and undoubtedly approach the ultimate strength. u '_ -———_—.___—._ —-—. .—.. _- ._ __h ,"I 8, q IO II I I6: calf-‘3 | 1740qu106} I 4 — ~ C 7 l —N—~«— ‘— #H—i—firfi—fi—W - A ,‘ ("n «n axes-0W E‘. Figure 19 46 {IV 75,"."9 - I'D“1"I\U:.M"C LIV/5‘31?» 0 I3. 73) ASE :1; .. DLIQIHGI FUR 6.‘ I. a" II MIR-NEE I53 / I .. -—--——-—~.——_ ,.\.-.-__.-.._LLLLL I I ._ _ ‘——-~—~v ___ - _‘.L__ - _. / . \ , . __-._._:. _L_ ./L_.Li__L I I I \ I [.L.. IL_L-\L__; LL L. _ ‘. -LLLLL LL- 1 V . I / I L) I I L; I I /1//6//7' 5////7‘ ‘ c 'e g ~ ,3»: mwa ,-,/ wry/m7 ~ — — — —- g ' s ' 54.5 Figure 20 48 Before this key was installed, certain high bearing areas were evident, and these showed up in several gage areas during the key drive. The general trends in static and dynamic loading in both five thousand and fifteen thousand pound steam hammers are sim— ilar. The intensity of the static loading in both hammers appears to be dependent upon the same factors while the over-all working load is slightly higher for the larger hammer. CHAP TER V METALLURGICAL INVESTIGATION OF BASE MATERIAL The determination of imposed stresses on the hammer base anvil during forging established the degree of loading which the base material must be capable of withstanding. The purpose of these present studies was to determine if the base material was metallur- gically sound, meeting the specifications under which it was pur— chased, and if these specifications were capable of withstanding the service stresses during the forging operation. Test specimens were cut from hammer bases which had failed within the past year. From these samples a tensile bar was machined, chemical analysis determined, and microsamples produced. Table 11 lists the chemical analysis and physical properties of the various samples. All samples were taken from cast steel bases, both five thousand and fifteen thousand pound. Samples 543l-A, B, C, D were core-drilled from a new base at the foundry. The loca- tions were as follows: 50 TABLE II METALLURGICAL ANALYSIS OF HAMMER ANVILS Lab- Chemical Analysis oratory W Sample C Mn P 5 Si 7449 0.20 0.39 0.020 0.033 0.34 7331 0.22 . 0.44 0.023 0.034 0.40 7331-A 0.19 - 0.43 0.018 0.025 0.36 7949 0.26 0.66 5431-A 0.19 0.48 0.030 0.022 0.33 5431-B 0.33 0.52 0.046 0.048 0.37 543l—C 0.16 0.48 0.035 0.035 0.34 5431-D 0.17 0.49 0.030 0.030 0.33 51 TABLE II (Continued) Physical P rope rties Yield Ultimate Ni Cr Strength Strength % E1. % R'A' 0.48 0.10 25,000 55,500 24.5 34.1 0.44 0.05 .25,000 55,500 31.0 49.2 0.35 0.05 0.14 0.03 26,000 60,700 25.0 34.1 0.40 0.34 21,250 54,000 25.0 46.3 0.38 0.34 37,500 71,900 5.5 6.2 0.35 0.31 20,375 47,900 27.0 56.9 0.32 0.31 23,000 54,550 18.0 '21.0 52 Location Right side. Left side at the sprue. Front face. Back face. If X All samples were taken twenty—four inches below the anvil seat on the vertical center line of the base. All other samples were removed in the vicinity of the failure with the exception of sample 7331 which was removed from a solid section of the same base as sample 7331-A. Hammer bases are purchased on carbon specification only, with no specification as to physical properties. They are pit cast, poured from one end at a pouring temperature of 2850 to 2900° F. The anvil opening is partially cored and machined to size later. Castings are left in the pit one day for each five thousand pounds, approximately twenty days for a fifty ton anvil, and when removed are at a temperature of 200 to 300° F. They are subjected to no further heat treatment. Figure 21 illustrates the sulfide inclusions and the stringers which were evident in all of the microsamples in an excess amount. These stringers and inclusions are very detrimental to the impact properties of the material, serving as stress concentrators throughout the section. Figures 22, 23, 25, 27, 28, 30, and 31 are typical 53 ' -LLLLL-_ ____L__-, L -M -._.A.._.--_—L - _ Figure 21. Sample 7449; picric acid etch; X 200; C - 0.20%. Figure 22. Sample 7449; picric acid etch; X 75; C - 0.20%. Figure 23. Figure 24. , Sample '7331; picric acid etch; x 75; c - 0.22%. Sample 7331; picric acid etch; X 75; C - 0.47%. 54 55 Figure 25. Sample 7331-A; picric acid etch; X 75; C - 0.19%. Figure 26. Sample 7331-A; picric acid etch; X 75; C - 0.49%. 56 II III . L C - 0.26%. ;X75; id etch ; picric ac Sample 7949 Figure 27. Sample 5431-A; picric acid etch; X 75; C - 0.19%. Figure 29. Sample 5431-B; picric acid etch; X 75; C - 0.33%. Q 57 Figure 30. Figure 31. 58 Sample 5431-C; picric acid etch; X 75; C - 0.16%. ' Sample 5431-D; picric acid etch; X 75; C - 0.17%. 59 examples of the structure of a low carbon cast steel which has been subjected to a very slow cooling rate in a large section size. This structure is pearlite in a matrix of free ferrite. Severe carbon segregation is evident in samples 7331 and 7331-A (see Figures 24 and 26). Two microspecimens were cut from each test sample re- moved from this base. The microsamples showing the high percent— age of pearlite were rechecked for carbon content and indicated 0.47 and 0.49 percent carbon in samples 7331 and 7331-A, respectively. The carbon segregation in this area could possibly explain the fail- ure of this base in the side-frame fillet instead of the usual anvil seat area. The high carbon content of sample 54315B is due to the segregation of carbon at the sprue during cooling of the casting (see Figure 29). ‘ The physical properties of the samples were quite scattered and generally low for this range of chemical analysis. This further points to segregation and inhomogeneity throughout the casting. Dis- regarding those deviations from a perfect casting which were pres- ent, it can be seen that a perfect casting with the maximum physical properties possible with this analysis and section size would be taxed to its maximum to withstand the imposed stresses recorded during the forging ope ration. CHAPTER VI CONCLUSIONS AND RECO MM ENDATIONS The investigation carried out to determine the imposed stress levels on hammer base anvils during forging operation has established definite levels which may be expected in the region of the anvil seat. Static loads that may be expected with the presently designed key and sow—block units are in the range of 15,000 p.s.i. to 20,000 p.s.i. with the degree of fitting now used in sow-blocks and keys. This range varies with the accuracy of fitting these units, and smaller tolerances in placing both key and sow-block in the hammer should result in a substantial decrease in static loads. Static stress levels are also aggravated by die warming and the key should not be fully tightened until temperature variations in the anvil seat area have equalized. Dynamic stress levels are also influenced by the degree of sow-block and key fit, and stresses up to 30,000 p.s.i. may be expected in the fillet radii of the anvil. Perhaps as important as the determination of degree of stress is the determination of distribution and origin of the strain loading. The transfer of load to the sides of the anvil seat is definitely 61 caused by the effect of Poisson's ratio on the sow-block under im- pact. A theoretical analysis of this effect applied to the hammer base reveals that a unit increase in sow-block width will produce a proportional decrease inntransmitted load to the anvil sides. A. fur- ther investigation of wave propagation across the top of the sow- block is necessary to determine the exact benefit this type of de- sign change would produce. The principal stress in the fillet radii is unidirectional and at right angles to the anvil seat. Modification of key design to produce a key which would act as a stiff spring would also reduce the strain transmitted to the anvil sides.‘ A key of this design would be comparable to two springs in series result- ing in smaller deflection per unit load applied. Metallurgically there is evidence of carbon segregation and excessive inclusions in the present base material. However, with the present type of casting the maximum physical properties are not sufficient to withstand the imposed loads for satisfactory periods of time. An alloy steel casting combined with a normalizing heat treat- ment to reduce, the relative blocky ferrite and elimination of segre- gation would produce a marked improvement in the physical proper- ties of the base. Because of the orientation of the principal stress in the fillet radii, a correlation of improvement in impact properties 62 would be possible by means of impact testing with a specially de- signed specimen to evaluate the effects of changes in heat treatment and chemical analysis. Recommendations for additional work are: 1. Determination of wave propagation in the sow—block by additional strain gage tests. 2. Photoelastic tests with variations in anvil fillet, sow-block and key design. 3. A more extensive examination of one broken base by core- drilling samples from several selected locations. 4. Consultation with hammer base manufacturers to develop a heat—treated alloy steel casting. 5. Impact testing with a specially designed specimen to eval- uate proposed changes in chemical analysis and heat treatment. BIBLIOGRAPHY Battelle Memorial Institute, Prevention _o_f_' the Failure of Metals Under Repeated Stress. New York: John Wiley and Sons, Inc., 1941. 240 pp. Bridgman, P. W., Studies in Larg Plastic Flow and Fracture. New York: McGraw-Hill Book Company, Inc., 1952. 355 pp. Bullens, D. K., and Battelle Memorial Institute, Steel and 1.3.5. Heat Treatment. Vol. III. New York: John Wiley and Sons, Inc., 1949. 598 pp. Gensamer Maxwell, Strength 31; Metals Under Combined Stresses. Cleveland: American Society of Metals, 1941. 100 pp. Hetenyi, M., Handbook of Experimental Stress Analysis. New York: John Wiley and Sons, Inc., 1950. 1077 pp. Lee, George Hamor, An Introduction to Experimental Stress Analysis. New York: John Wiley and Sons, Inc., 1950. 304 pp. Nadai, A., Theory of Flow and Fracture 9}: Solids. Vol. I. New York: McGraw-Hill Book Company, Inc., 1950. 559 pp. Thomson, William Tyrrell, Mechanical Vibrations. New York: Prentice—Hall, Inc. 1948. 212 pp. APPENDIX 65 (gage type A-7, resistance 119.5 ohms, gage factor 1.93) TABLE III STATIC STRAIN DATA, HAMMER 503, TEST 1 Load Re adings Ins trument Re adings s—1 s—z s-3 s-4 s-5 S-l Zero ........ 12-1875 1240720 10-0770 10—1375 12-0155 Truck drive 12-1910 12-0720 10-0800 10-1400 12—0155 + 35 Key driver 1 12-1960 12—1740 10-0755 10-1380 12-0150 + 50 Key driver 2 12-1970 12—0845 10-0775 10-1420 12-0210 + 10 Key driver 3 14-0030 12-0930 10-0820 10—1470 12-0255 + 60 Key driver 4 14—0120 12—1070 10-0850 10—1540 12—0330 + 90 Key driver 5 14-0147 12-1130 10-0890 10-1555 12-0360 + 25 Key driver 6 14-0230 12-1270 10-0935 10-1660 12-0440 + 85 Key driver 7 14—0290 12-1360 10-0970 10-1700 12-0490 + 60 Key driver 8 14—0310 12—1400 10-0990 10-1720 12-0515 + 20 Key driver 9 .. 14-0450 12—1590 10-1060 10-1840 12-0630 +140 $i;:u}:eat¢d . .. 14-0480 12—1720 10-1140 12-0030 12—0700 + 30 Before forging . .14—0440 12—1675 10—1130 10-1970 12-0640 - 40 Total ...... .. +565 After key is substantially tight, average key with stress build—up of 900 p.s.i. per blow. movement is 3/32 inch TABLE III (Continued) ———;7 Strain (mic roinche s / inch) Stress (p.s.i.) S-Z S—3 8-4 8-5 S—l S-2 S-3 S-4 S-5 0 + 30 + 25 .0 +1050 0 +0900 +0750 0 + 20 - 45 - 20 - 5 +1500 +0600 -l350 -0600 -0150 +105 + 20 + 40 + 60 +0300 +3150 +0600 +1200 +1800 + 85 + 45 + 50 + 45 +1800 +2550 +1350 +1500 +1350 +140 + 30 + 70 + 75 +2700 +4200 +0900 +2100 +2250 + 60 + 40 + 15 + 30 +0750 +1800 +1200 +0450 +0900 +140 + 45 +105 + 80 +2550 +4200 +1350 +3150 +2400 + 90 + 35 + 40 + 50 +1800 +2700 +1050 +1200 +1500 + 40 + 20 + 20 + 25 +0600 +1200 +0600 +0600 +0750 +190 + 70 +120 +115 +4200 +5700 +2100 +3600 +3450 +130 + 80 + 90 + 70 .+0900 +3900 +2400 +2700 +2100 — 45 — 10 — 60 - 60 -1200 -1350 -0300 -1800 -1800 +955 +360 +595 +485 +16950 +28650 +10800 +17850 +14550 67 TABLE IV DYNAMIC STRAIN DATA, HAMMER 503, TEST 1 (gage type C-7, 500 ohm resistance, gage factor 3.29) Cali- Cali- , Strain Stress Gage Blow , , Signal , , , bration bration , (microm. (p.s.i. No. No. . , Units , Setting Units per 1n.) avg.) Dax Shift D-l 2 10 20 5.3 399 11,950 4 5 15.3 10.2 50. 15,200 6 10 20 12.3 930 27,900 D-2 2, 5 15.3 7.1 35. 10,600 4 5 15.3 10.5 52. 15,700 6 10 20 14.2 1070 32,100 D— 3 2 Negligible 4 Negligible 6 . 1 l3 14 164 4,900 D-6 2 5 15.3 3.5 176 ‘5,300 4 5 15.3 10.8 541 16,250 6 10 20, 14.6 1108 33,250 Night Shift D-l 4 10 20 9.5 716 21,500 5 10 20 8.8 663 19,900 6 10 20 15.0 1133 34,000 D—2 4 10 20 5.5 412 12,350 5 5 15.3 11.5 575 17,250 6 10 20 15.2 1150 34,500 D-3 Gage unstable D-6 Gage unstable Note: All stress Values to the nearest 50 p.s.i. 68 TABLE V STATIC STRAIN DATA, HAMMER 503, TEST 2 119.5 ohm resistance, gage factor 1.93) (gage type.Ar7, Instrument Readings Load Readings G-l G—Z (3-3 G-4 G-S G-6 G-l O ...... 12-1950 12-1520 12-1200 12-0030 12-0430 12-0230 1 .. 12-1950 12—1525 12—1205 12-0035 12-0435 12-0235 0 2 ...... 12-1970 12-1540 12-1220 12—0040 12-0440 12-0240 + 20 3 ...... 12-1965 12-1540 12-1225 12-0045 12-0445 12-0250 — 5 4 .. — 12-1965 12-1545 12-1230 12-0045 12-0445 12-0250 0 5 ...... 12-1950 12-1530 12-1220 12—0030 12-0430 12-0235 — 15 6 ...... 12-1960 12-1550 12—1240 12-0050 12-0450 12-0255 + 20 7 ...... 12-1955 12-1550 12-1240 12-0050 12-0450 12-0260 - 5 8 ...... 12-1960 12—1555 12-1250 12-0055 12-0450 12-0260 + 5 9 ...... 12-1945 12-1560 12—1260 12—0060 12-0450 12-0260 - 15 10 ..... 12-1950 12-1560 12—1265 12-0060 12-0450 12-0275 + 5 11 ..... 12-1950 12-1565 12-1275 12-0065 12-0450 12-0285 0 12 ..... 12-1955 12-1570 12-1285 12-0070 12-0455 12-0290 + 5 13 ..... 12-1945 12-1570 12-1290 12-0075 12-0455 12-0300 - 5 Dies heated 1 hour 12-1970 12-1620 12-1390 12-0480 12-0440 + 25 12-0140 TABLE V (Continued) 69 Strain (microinches/inch) Stress (p.s.i.) G-Z G—3 G-4 G-s (3-6 G-l G—Z G-3 G-4 G-5 (3-6 +5+5+5+5+5 0+150+150+150+150+150 +15+15+ 5+ 5+ 5 600 +450 +450+150 +150 +150 0+ 5+ 5+ 5+10 150 0+150+150+150+300 + 5 + 5 0 0 0 0 + 150 + 150 0 0 0 — 15 - 10 - 15 - 15 - 15 450 - 450 — 300 - 450 - 450 - 450 + 20 + 20 + 20 + 20 + 20 600 + 600 + 600 + 600 + 600 + 600 0 0 0 0 + 5 150 0 0 0 0 +150 + 5 + 10 + 5 0 0 150 +150 + 300 + 150 0 0 + 5+10+ 5 0 0 '450 +150 +300 +150 0, 0 0+5 0 0+15150 0+150 0 0+450 +5+1o+5 0+10 0+150+300+150 0+300 +5+10+5+5+5150+150+300+150+150+150 0+ 5+5 0+1‘0 150 o+150+150 0+300 + 50 +100 + 65 + 25 +140 750 +1500 +3000 +1950 + 750 +4200 70 TABLE V (Continue d) Ins trument Re adings Load Readings G-l G-Z G-3 G-4 G-S G-6 G-l Before forging 12-1955 12—1590 12—1375 12-0120 12-0485 12-0435 - 15 Key out .. 12-1445 12—0600 12-0465 10-0665 U” 10-1440 stable Shim in g-tfght 12-1405 12-0915 12-0630 10-1130 10—1610 - 40 Key tight . 12—1655 12-1270 12-1140 12—0270 12-0110 +250 1 hour forge 12-1610 12-1260 12-1080 12-0285 12-0090 - 45 3 hour . forge 12-1540 12-1180 12—1025 12-0110 12-0000 - 70 TAB LE V (Continued) 71 Strain (mic roinche s /inch) Stress (p.s.i.) G-Z G-3 G—4 G—S G-6 G—l G-Z G—3 G—4 G—5 G-6 - 3O - 15 - 20 + 5 - 5 - 450 - C)OO - 450 - 600 + 150 - 150 +315 +165 +465 +170 -1200 +9450 +3950 +13950 +5100 +355 +630 +140 +500 +7500 +10650 +18900 +4200 +15000 - 10 - 60 + 15 - 20 -l350 - 300 -1800 + 450 - 600 - 8O - 55 -175 - 9O -Z-100 -2400 -1650 -5250 -2700 DYNAMIC STRAIN DATA, HAMMER 503, TEST 2 (gage type C-7, resistance 500 ohms, gage factor 3.29) TABLE VI 72 Cali- Cali- Strain . , Signal , , Stress Gage Blows bration bration Units (microm. ( s i ) Setting Units per in.) p. ' ' D-l 2 10 20 7.5 283 8,500 4 5 20 6.1 460 13,800 6 5 20 12.5 943 28,300 D-Z 2 10 20 7.8 291.7 8,750 4 10 20 5.8 216.7 6,500 6 5 20 13.5 1020 30,600 D—3 "2 10 20 5.2 196.7 5,900 4 10 20 4.9 186.7 5,600 6 5 20 11.1 840 25,200 Note: All stress readings rounded to the nearest 50 p.s.i. 73 TABLE VH STATIC STRAIN‘DATA, HAMMER 503, TEST 3 (gage type A-7, resistance 119.5 ohms, gage factor 1.93) Instrument Re adings Load ____. Readmgs G-l G-Z G—3 G—4 G—5 G-l 0 ....... 14-0010 14-0555 12—1590 14-0520 12-1550 1 ....... 14-0010 14-0555 12-1475 14-0540 12-1550 0 2 ....... 14-0030 14-0550 12-1320 14-0565 12-1540 + 20 3 ....... 14-0045 14-0500 12-1255 14-0590 12-1535 + 15 4 ....... 14-0060 14-0400 12-1235 14-0585 12-1530 + 15 5 ....... 14-0060 14—0310 12-1210 14-0595 12-1530 0 6 ....... 14-0070 14-0250 12-1210 14-0600 12-1530 + 10 Dies heated 15 nun. .. 14-0060 14-0200 12-1275 14-0640 12-1540 - 10 30 nun” ... 14-0075 14-0200 12-1280 14-0655 12-1560 + 15 45 rnin. L. . 14-0090 14-0235 12-1310 14—0670 12-1575 -+ 15 1 hour 14-0100 14-0285 12-1355 14-0670 12-1570 + 10 1 hour, 15 nun. 14—0105 14-0310 12-1385 14-0670 12-1580 + 5 1 hour, 30 rnin. .. . 14-0110 14-0340 12-1420 14-0675 12-1620 + 5 1 hour, 45 nfin. ... 14-0120 12-1460 14-0660 12-1605 + 10 14-0375 TABLE VII (Continued) 74 Strain (mic ro inche 5 /inch) Stress (p.s.i.) G-z G-3 G-4 G-5 <3-1 G-z G-3 G-4 G—S 0 -125 20 0 0 0 -3750 + 600 0 - 5 -155 ._ 25 10 600 - 150 —4650 + 750 - 300 - 50 - 65 25 5 450 -1500 -1950 + 750 - 150 -100 - 20 5 5 450 -3000 - 600 - 150 - 150 — 90 - 25 10 0 0 -2700 - 750 + 300 0 -— 60 O 5 O 300 -1800 O + 150 O - 50 + 65 40 10 300 -1500 +1950 +1200 + 300 0 + 5 15 20 450 0 + 150 '+ 450 + 600 + 35 + 30 15 15 450 +1050 + 900 + 450 + 450 + 50 + 45 0 5 300 +1500 +1350 0 - 150 + 25 + 30 0 10 150 + 750 + 900 0 + 300 + 30 + 35 5 40 150 + 900 +1050 + 150 +1200 + 35 + 40 15 15 300 +1050 +1200 - 450 - 450 75 TAB LE V 11 (Continue d) Instrument Readings Load ____ Readings G-l o—z G-3 G-4 G-5 G-l IKey'fight .. 14-0420 14-0640 12-1340 14-0770 12-1650 +300 Total +410 G-6 G-7 G-8 G-9 G-lO G-6 0 ....... 12-0790 14-0185 12-1885 14-0645 12-1110 1 ....... 12-0790 14—0185 12-1880 14—0645 12—1150 0 2 ....... 12—0780 14-0195 12-1885 14—0640 12-1255 - 10 3 ....... 12-0775 14-0210 12-1885 14-0645 12-1260 - 5 4 ....... 12-0775 14-0220 12-1875 14-0645 12-1300 0 5 ....... 12-0770 14-0225 12—1865 14-0645 12-1335 - 5 6 ....... 12-0765 14—0235 12-1860 14-0645 12—1365 - 5 Dies heated 15 nun" ... 12-0780 14-0235 12—1890 14-0790 12-1285 + 15 30 nun” ... 12-0790 14-0245 12-1910 14-0815 12-1315 + 10 45 nfina ... 12-0800 14«0260 12-1920 14-0820 12-1310 + 10 1 hour 12-0805 14-0260 12-1940 14-0820 12-1300 + 5 1 hour, 15 nfin. 12-0805 14—0255 12—1940 14—0805 12-1280 0 TABLE VII (Continued) 76 v Strain (mic roinche 5 /inch) Stress (p.s.i.) G-Z G-3 G-4 G-S G—l G-Z G-3 G—4 G-5 +265 -120 +110 + 45 +9000 +7950 -3600 +3300 +1350 + 85 —250 +250 +100 +12300 +2550 -7500 +7500 +3000 G-7 G-8 G-9 G-lO G-6 G-7 G-8 G-9 G-lO 0 - 5 0 + 40 0 0 - 150 0 +1200 + 10 + 5 - 5 +105 — 300 + 300 + 150 — 150 +3150 + 15 0 + 5 + 5 - 150 + 450 0 + 150 + 150 + 10 - 10 0 + 40 0 + 300 - 300 0 +1200 + 5 — 10 0 + 35 - 150 + 150 - 300 0 +1050 + 10 — 5 0 + 30 - 150 + 300 - 150 0 + 900 0 + 30 +145 + 20 + 450 0 + 900' +4350 + 600 + 10 + 10 + 25 + 30 + 300 + 300 + 300 + 750 + 900 + 15 + 10 + 5 - 5 + 300 + 450 + 300 + 150 - 150 0 + 20 0 - 10 + 150 0 + 600 0 - 300 - 5 0 — 15 - 20 0 - 150 0 - 450 — 600 77 TABLE VII (Continued) Instrument Readings Load Readings G-6 G—7 G-8 G—9 G-lO G—6 1 hour, 30 rnin. .. . 12-0820 14-0260 12-1940 14-0800 12—1265 -+ 15 1 hour, 45 nun. 12-0830. 14-0260 12-1940 14-0800 12-1265 + 10 Key tight.. 12-0910 14-0370 12-1880 14-0790 12-1620 + 80 +120 To tal TABLE VII (Continue d) 78 Strain (mic roinche 5 /inch) Stress (p.s.i.) G-7 G-8 G-9 G-lO G-6 G-7 G-8 G-9 G-10 + 5 0 - 5 — 15 + 450 + 150 0 — 150 - 450 O 0 0 0 + 300 0 0 0 0 +110 - 6O - 10 +355 +2400 +3300 -1800 - 300 +10650 +185 - 5 +145 +510 +3600 +5550 - 150 +4350 +15300 DYNAMIC STRAIN DATA, HAMMER 503, TEST 3 (gage A-7, resistance 119.5 ohms, gage factor 1.93) TABLE VIII Gage C2111- Cali— Signal Strain Stress No. bration bration Units (microm. (p.s.i.) Setting Units per in.) 1 10 8 11.8 +915 +27,450 2 10 12 13.2 -682 -20,460 3 10 12 7.1 -366 -10,980 4 10 20 10 +310 + 9,300 5 Negligible 6 Negligible 7 10 12 7.5 +387 +11,610 8 10 20 6.5 —202 - 6,060 9 10 12 8.5 —440 -13,200 10 . 10 8 10 +775 +23,250 A 10 13.5 10.5 960 28,800 B . 10 20 9.5 465 13,950 C 10 20 0 0 O Note: All stress values rounded to nearest 50 p.s.i. 80 TABLE IX STATIC STRAIN DATA, HAMMER 153 (gage type A-7, resistance 119.5 ohms, gage factor 1.93) Load Re ading s Instrument Readings G-l G-Z G-3 G-4 O ..... . ......... 12-0900 14-0630 Unstable 10-1490 1 ............... 12-0900 14-0625 10-1490 2 ............... 12-0905 14-0630 10-1490 3 ............... 12-0920 14-0630 10-1500 4 ............... 12-0910 14-0630 10-1570 5 ............... 12-0920 14-0630 10—1610 6 ............... 12-0920 14-0630 ~10-164O 7 .............. . 12-0920 14—0630 10—1660 8 ............... 12-0915 14—0630 10-1665 9 ......... . ..... 12-0920 14—0635 10-1695 10 ....... . . 12-0920 14-0635 10-1730 Die heat 1 ......... 12-0960 14-0690 10-1725 Die heat 2 ......... 12-0980 14-0710 10-1730 Forging 1 ......... 12-1350 14-1390 10-1630 Forging 2 ......... 12-1200 14-1230 10-1490 Forging 3 .......... 12-1250 14-1110 10-1410 Forging 4 ......... 12-1300 14-1270 10-1410 Key retight 1 ....... 12-1350 14-1280 10-1500 Key retight 2 . ..... 12-1375 14-1300 10-1540 Key retight 3 ....... 12—1405 14—1320 10-1565 Key retight 4 ....... 12‘1430 14-1340 10—1600 Key retight 5 ....... 12-1440 14-1355 10-1620 TABLE IX (Continued) 81 Strain (mic roinches /inch) Stress (p.s.i.) G—l G—Z G-3 G-4 G-l G-Z G-3 G-4 Unstable Unstable O - 5 0 0 - 150 0 + 5 + 5 O + 150 + 150 0 + 15 0 + 10 + 450 0 + 300 - 10 O + 70 - 300 0 +2100 + 10 0 + 40 + 300 O +1200 0 0 _ + 30 0 0 + 900 0 0 + 20 0 0 + 600 — 5 0 + 5 - 150 0 + 150 + 5 + 5 + 30 + 150 +150 + 900 0 0 + 35 O 0 +1050 + 40 + 55 - 5 +1200 +1650 - 150 + 20 + 20 + 5 + 600 + 600 + 150 +370 +680 -100 +11100 +20400 - 300 -150 -l6O -1-40 -4500 -4800 74200 + 50 -120 - 80 +1500 --3600 --2400 + 50 +150 0 +1500 +4500 0 + 50 + 10 + 90 +1500 + 300 +2700 + 25 + 20 + 40 + 750 + 600 +1200 + 30 + 20 ‘+ 15 + 900 + 600 + 450 + 25 + 2.0 + 35 + 750 + 600 +1050 + 10 + 15 + 20 + 300 + 450 + 600 82 TABLE IX (Continue (1) In strument Re ading s Load Re adings G—l G—2 G—3 G-4 :Key reught 6 ...i... . 12-1460 14-1370 10-1640 Key refight'? ....... 12-1475 14-1385 10—1660 :Key refight 8 ....... 12—1490 14-1390 10-1675 Ifiey refight 9 ...,... . 12-1500 14-1390 10-1695 :Forging 1 ....... 12-1415 14-1420 10-1610 Forging 2 ....... 12-1450 14-1570 10-1610 .Forging 3 ....... 12—1390 14—1540 10-1565 Forging 4 ....... 12-1490 14-1610 10-1630 Forging 5 ....... 12-1565 14-1670 10-1670 G-S G-6 G—7 G-8 0 ............. . 10-1960 12—0720 12-0660 12-1985 1 ............... 10-1955 12-0720 12—0655 12-1980 2 ............... 10-1950 12-0720 12-0660 12-1980 3 ............... 10—1955 12—0720 12-0660' 12-1985 4 .............. . 10-1975 12—0745 12-0660 12-1990 5 . .......... 10-1995 12-0760 12-0665 12-1990 6 . ...... . 12-0015 12-0765 12-0670 12-1990 7 ............ 12-0020 12—0765 12-0670 12-1990 8 ............... 12-0025 12-0760 12—0665 12-1990 9 ............... 12-0040 12-0770 12-0670 12—1990 10 .............. 12-0060 12-0785 12-0675 12-2000 [no heat 1 ....... 12-0055 12-0765 12-0725 14-0040 the heat 2 ....... 12-0070 12—0770 12-0740 14-0050 83 TABLE IX (Continued) Strain (microinches/inch) Stress (p.s.i.) G-l (3-2 G-3 (3-4 (3-1 G-Z G-3 G—4 + 20 + 15 + 20 + 600 + 450 + 600 + 15 + 15 + 20 + 450 + 450 + 600 + 15 + 5 + 15 + 450 + 150 + 450 + 10 0 + 20 + 300 0 + 600 - 85 + 30 — 85 —2550 + 900 -2550 + 35 +150 0 +1050 +4500 0 _ 60 - 30 — 45 -1800 - 900 -1350 +100 + 70 + 65 +3000 +2100 +1950 + 75 + 60 + 40 +2250 +1800 +1200 G—5 G—6 G-7 G-8 G-5 o-6 G-7 G-8 +20 +20 +20 +15 +20 0 0 0 +15 +15 +5 0 -5 +10 +15 -20 +5 — 5 + 5 0 0 + 5 + 5 0 - 5 + 5 + 5 +50 +15 + UTOOOOOWUIOW - 150 - 150 + 150 + 600 + 600 + 600 + 150 + 150 + 450 + 600 i 150 + 450 + 450 + 450 + 150 - 150 + 300 + 450 - 600 + 150 - 150 + 150 + 150 + 150 - 150 + 150 + 150 +1500 + 450 84 TABLE 1x (Continued) In strument Re ading 5 Load Re ading s G-5 G-6 G-7 G-8 Forging 1 ........ . 12-0160 12-1100 12-1110 14-0400 Forging 2 ......... 10-1990 12—0880 12—0920 14-0230 Forging 3 ......... 10-1935 12—0815 -210755 140240 Forging 4 ......... 10-1935 12-0890 12-0910 14—0270 Key retight 1 ....... 12-0010 12-1045 12-0990 14-0310 Key retight 2 ....... 12—0040 12-1070 12-1015 14-0320 Key retight 3 ....... 12-0040 12-1105 12-1040 14—0340 Key retight 4 ....... 12-0060 12—1130 12-1060 _ 14-0355 Key retight 5 ....... 12-0070 12—1145 12-1080 14-0370 Key retight 6 ....... 12—0080 12-1165 12—1100 14-0380 Key retight 7 ....... 12-0090 12-1180 12-1115 -410385 Key retight 8 .. . . . .. 12—0100 12-1195 12-1130 14-0395 Key retight 9 ....... 12—0105 12—1200 12-1130 14-0390 Forging 1 ......... 12-0070 12-1200 12—1270 14-0350 Forging 2 ......... 12-0070 12-1260 12-1250 14-0400 Forging 3 ......... 12-0040 12-1180 12-1225 14-0340 Forging 4 ......... _12—0125 12-1380 Unstable 14-0480 Forging 5 ......... 12—0165 12-1360 14-0540 TABLE IX (Continued) 85 Strain (microinche s / inch) Stress (p.s.i.) G-S G-6 G-7 G-8 G-5 G-6 G-7 G-8 + 90 +330 +370 +350 +2700 +9900 +11100 +10500 -170 -220 -190 -l70 -5100 -6600 —5700 -5100 - 55 - 65 —165 + 10 —1650 -l950 -4950 + 300 0 + 75 +155 + 30 0 +2250 +4650- + 900 + 70 +155 + 80 + 40 +2100 +4650 +2400 +1200 ‘ + 30 + 25 + 25 + 10 + 900 + 750 + 750 + 300 0 + 35 + 25 + 20 0 +1050 + 750 + 600 + 20 + 25 + 20 + 15 + 600 + 750 600 + 450 + 10 + 15 + 20 + 15 + 300 + 450 + 600 + 450 + 10 + 20 + 20 + 10 + 300 + 600 + 600 + 300 + 10 + 15 + 15 + 5 + 300 + 450 + 450 + 150 + 10 + 5- + 15 + 10 + 300 + 150 + 450 + 300 + 5 + 5 0 - 5 +150 +150 0 —150 —- 35 0 +140 - 40 ~1050 0 +4200 -1200 0 + 60 - 20 + 50 0 +1800 - 600 +1500 - 30 - 80 — 25 - 60 - 900 -2400 - 750 «1800 +105 +200 Unstable +140 +3150 +6000 Unstable +4200 + 20 - 20 + 60 + 600 - 600 +1800 TABLE X DYNAMIC STRAIN DATA, HAMMER 153 (gage type A-7, resistance 119.5 ohms, gage factor 1.93) Cali- Cali- , Strain , , Signal , , Stress Gage bration bration , (microm. , _ , Units , (p.s.i.) Setting Units per in.) Day Shift, First Key Drive 1 20 20 13.75 830 25,500 2 20 20 16.25 1006.7 30,200 3 Unstable 4 20 19 5.25 350 10,500 5 20 19 4.5 293.3 8,800 6 20 19 7.75 505 15,150 7 20 19 16.25 1060 31,800 8 20 19 16.25 1060 31,800 Night Shift, First Key Drive 1 20 20 11.75 726.7 ‘21,800 2 20 20 14 866.7 26,000 3 Unstable 4 20 20 3.9 241.6 7,250 5 20 20 3.5 216.7 6,500 6 20 20 6.75 417.3 12,520 7 20 20 15.25 943.3 28,300 8 20 20 13.25 820.0 24,600 Day Shift, Key Retightened 1 20 19.5 10 635 19,050 2 20 19.0 12 783.3 23,500 3 Unstable 4 20 19.0 4.5 293.3 8,800 5 20 18.5 4.5 301.3 9,040 6 20 18.5 7.0 470.0 14,100 7 Unstable 8 20 18.5 13.0 ' 873.3 26,200 TABLE X (Continue d) 87 Cali- Cali- , Strain _ , Signal , , Stress Gage bration bratlon Units (micr01n. (p.s.i.) Setting Units per in.) Night Shift, Key Retightened 1 20 19.5 12.5 793.3 23,800 2 20 19.0 17.0 1106.7 33,200 3 Unstable 4 20 19.0 4.5 290 8,700 5 20 18.5 4.5 316.7 9,500 6 20 19.0, 6.0 390 11,700 7 Unstable 8 20 18.5 13.0 873.3 26,200 88 SAMPLE CALCULATIONS G.R. = Gage resistance in ohms. G.F. = Gage factor. C.R. = Calibration resistance in ohms. e = Strain in microinches per inch. Calibration setting = Calibration setting on Ellis BA-Z amplifier. Calibration units 2' Number of oscillograph screen graph units covered by the calibration signal. Signal units = Amplitude of impact signal on oscillograph screen graph. 6 E 2' Modulus of elasticity for steel (30 x 10 p.s.i.). Computation of strain per unit of oscillograph screen graph: G.R. = .500 ohms G.F. = 3.29 Calibration setting = 10 C.R. = 100,000 ohms. Calibration units = 20 Signal units = 12.3 [G.R.] + [G.F.(C.R. + G.R.)] (D II [500] + [3.29(100,000 + 500)] 1512 mic roinches /inch. Strain per graph unit [Strain] é [Calibration units]. 1512 + 20 = 75.6 microinch/inch/unit. Strain per graph unit Stress per graph unit = Strain/Unit x E. . \. \- ' 7 ~ _6 6 \ - Stress per graph unit = 75.6 x 10 x 30 x 10 = 2268 R."s.i./Unit.7_f M Signal units x stress per graph unit. Z: ' 9 ~\ ‘ ‘“ \ Impact stress = 12.3 x 2268 = 27,906 p.s.i. ‘. Impact stres s \ k}.3rl‘ 1“ ,li (.tk’lfli .3 ll 1‘ .l { t (. ‘ t 1100M USE—UW ‘ ‘ \ g 5-4 l 13‘) nEd. if! .5») c. MICHIGAN STATE UNIVERSITY LIBRARIES 1741 III 3 1293 03071