PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 6/01 c/CIRC/DateDue.p65-p.15 a sSPfUDY OF HS CONTRACTION AND EXvaNolON O¥ OaxrBON STEELS. A fFHEBSIS submitted to the Faculty of MICHIGAN AGHICULIVUHaAL COLLEGE c Ge Ue. Shumway Ke O. kernen Candidates for the Degree of Bachelor of soience. Jane, 1920. THESIS - (37 (59 or Lt (rif IT Tilt VI vir VIII TABLE OF CONTSNTS Object and scope of Investigation. Theory apparatus and Material. Instruments and Measurement of samples. Description of Tests. Oondlusions. Dats. Curves and lrawings. ‘ C q ye 103 , ‘ X OBJECT aND SCOrE OF THe INVESTIGATION. With the increased price of steel and its increased use throughout every branch of induetry the importance of salvaging parts, which have been machined oversise, has begun to be realised. There hes been no accurate estimate of the value of parts which, due to careless mohine work, are nearly a total loss each year, but it is needless to say that it is a large one. In view, therefore, of the increasing importance of salvaging these parts the object of the present investigation is an experimental atudy of the @mplication of heat treatment to the problem in order to furnish men in charge of such work exact information as to the limitations and results attain- able. With a view to making the remilts of this investigation available to those who have not had special technical training, the authors have mesented the results by a series of ourves which shonld be elear to the average reader, however, in order to mke a eomlete report, all experimental data has heen in- Gluded so that readers may familiarise themselves with the methods employed in investigating the various questions which have been considered. Il THEORY We find that in quenching a piece of steel, during the process of heat treating, if the piece is not handled properly it is liable to warp due to the unequal contractions and expansion in the steel. This characteristic is partieularly emphasised in forgings or tools of unequal thickness. If it were possible to cool each particle af the piece in the time and order desired the resulting dilatation could be regulated with regard to direotion. we know as a result of recent research work that when a piece of steel is heated through its critioal range the formation of Austenite takes place with a decrease in volume, and that a somewhat eorrespomiing and opposite increase in wolume takes place when it is oooled through the same oritisal range. Sauveur states - “that the metal which in cooling above the upper critical point was contracting, as ie the rule with all cooling bodies, on passing through the upper critical point undergoes suddenly a marked dilatation amounting to over one thousandth of its length, immediately followed again by a normal contraction. such dilatation implies that the change of Gamma into Beta iron tekes place with an augementation of volume or in other words that Gamma iron is denser, has a higher Specific Gravity, than Beta iron”. The dilatation oocurring at the upper eritical point, on cooling, for steals oontaining respectively 0.05 to 1.5 percent oarbon are shown graphically by Sauveur in plate IB. On heating at the upper oritical point a spontaneous contraction eccure of the same magnitude as the dilatation on eeoling. Benedick's determined with great care the dilatation of pure iron between 700° and 960° centigrade and some of his results are shom on plate Ia. The partionular point to be noted is the ocourence Of a marked dilatation at C, the upper eritical point. | There has been considerable diseussion as to whether or not a dilatation ocours at the second transformation point. It is trie that there is no such marked change, as ocaors at the upper critical point and it is qestionable whether there is any ehange. At the first transformation point a third €ilatation takes place on cooling. The maximum €@iletation which can Occur is in the cane of a eutectoid steel when cooling through ite singls critical point. all changes of course vary with the —- earbon aontent of the ateel. The preceding theory is shown grephically on plate II. In heating from a to »b there is an expansion (marked +). On passing through the critical range there is a contraction (marked -). On heating still further a positive action takes place again. On cooling through the same curve the ranges will ohange in sign. A theory which has bem advanced, but not published is that if the steel is heated too fast the positive forces will predominate ani the finished plece will be lerger then the eriginal and that if the steel is heated too slow the negative forces will predominate ani the finished piese will be emaller than the original. IIr APPARATUS AND MATiNIAL All laboratory work was done in the Michigan Agri cultural College heat treatmmt laboratory and forge shop. This is a modern shop equipped with several small furnaces and one large one. These furnaces are gas fired, using Lanaing city gas having a quality of about 680 3.7.U. air is furnish- ed by a Root's Positive Presmure Blower, with a constant pressure exhaust Valve to keep the pressure on the furnaces constant regardless of the number of furnaces in operation. In this investigation a surface combustion style 0. furnace was used exclusively. This type of furnace has an automatic valve for maintaining a eonstant mixture of gas and air. The flame enters the combustion chamber at the level of the hearth but inelined downward so that it would normally pass veneath it. The spsage between the hearth and the walls of the furnmmce is filled with loose pieces of carbon- undum so as to break up the flame and produce complete combustion. There is an adjustmemt on the mixing valve for varying the porportion of gas to air. Dpring this investigation a slightly rich aixture was used at all times to prevent oxidation of the samples. A Leeds and Northrup potentiometer attaohed to an iron-constantin thermocouple was used to measure the temperature. A quenching bath of water at a temperature of eixty degrees fahrenheit was used for cooling the samples. This temperature was measured by a meroury thermometer and was kept constant by removing waym water and adding oold water. The apparatus used in measuring the samples eonsisted of micrometers, telescope gages, thickness gages and scales which were borrowed from the sollege machine shop and are desoribeé in detail elsewhere. & pin gage, used in measuring the pitch diameter, was constructed by the euthors. The material tested consisted of automobile transmission gears, and were of Reo B-1l5 forged steck whose analysis is identical with S.A.& 1020. These gears had been completely mohined but not carburized. The countershaft drive gear was selected as being typical of all moh pieces, and because of its large Qiameter the change in sise would be greater. IV MEASURIEGNTS OF SALWLES. In order to ascertain the amount of shrinkage resulting from any heat treatment it is evident that en accurate Overall measurement of the gear met be made before and after the treatment. In addition, to ve of value, the measuremnts met be mde with the same instruments and at the same spot on the gear, since the variation in the surface might cause readings to be obtained which, if not taken at the same points, would be greatly in error. In order to insure the readings being taken at the seme point a drawing of the gear was made and the dimensions on whieh measure- ments were to be take, were indiceted and numbered. In addition to this the points on the rim were marked with a prick punch to locate the teeth on whigh the diameters were neasured. In order to eliminate errers due to taking the same measurement with several €ifferent micrometers, the same micrometers were used for all corresponding measurements. Numbers one to three inolusive were measurements of the Aiameter on the pitch circle. This showed the change in the outside diameter and was taken as a base rather than the top or bottom of the teeth, because of its greater importanse in determining the running qualities of the gears. Numbers four and five were the distances from the plane of the end of the hub to the piane of the rim of the gear. This gave a check upon Warping caused by the forces set up in shrinking. Humbers wix to nine inclusive were masurements of the inside diameter of the hub, the most important of ell the measurenents taken, since the objeot of the whole process waa to reduce thie diameter. This wae measured in two directions at rizht angles and at each end of the hub. It was necessary to masure at each end of the hub as the hole was counterbored near the center leaving a short surface near each mad for a bearing on the ahaft. It was necessary to measure at right engles in order to tell whether the hole grew out of round. SRumbers ten, eleven ani twelve were masurements on the thickness of the rim through specified points. These showed actual shrinkage of the metal. MEASURING INSTRUMENTS. In order to measure the pitch diameter it was necessary to construct a pair of gage pins which would fit between the teeth and bear upon the tooth at each side at the ociroumference of the pitch oirole. These pins were made from drill rod, hardened and ground to proper sise. By use of these pins together with a pair of six to seven inoh outelide micrometers, a very accurate measurememt could be taken checking to within five ten-thousandiths of an inch. an accurate measurement of the pitoh diameter could not be obdtained ‘ginoe there was an odd number of teeth in the gear, a condition which brought a tooth opposite each space. Sinee a oomparative measurement was all that was re- quired this did not cause a sensible error in the results. The dGietance from the plane of the end of the hub to the plane of the rim was measured by means of «a set of thickness gages and a scale. A heavy scale was usec and the gages inserted between the scale and the hub end. Keadings were taken at the ends of two diameters at right angles anf@ the readings for each diameter averaged. These readings would show any tendenoy to warp or dish. The measurement of the hole, which was the most important ané also the most difficult to acourately obtain, was secured by means of a telescope gage and outside micrometers. Due to the roughness of the in- side surface it was rather hard to get a elose cheok on this reading, onsequently it was necessary to take several readings and reoord the average. The masurements of numbers ten, eleven and twelve were obtained directly with a pair of one inch micrometer calipers. These readings were take at the base of corresponding teeth on all gears. Vv DESCKIPPION OF TESTS. It was necessery in order to nake the results useful, that gimilar conditions shonld exist in all testa, therefore the first step was to determine the mthod of heating the samples. This was done by heating several samples to the same temperature but at different rates. One extreme was to place the gear in a cold furnace ané bring both up to temperature together. another extreme was to bring the furnace up to temperature and thm place the gear in it. The last extreme was productive of the best results and was adopted for ell teats. In preparing the samples (transmiseion gears) for test the hole in the oenter was filled with moist fire clay and packed hard. ‘the olay was rounded well over the edge of the hole to prevent any oxidation near this important point of measurement. In placing the sample in the furnace great eare was used to locate each one in the same spot, which was adjacent to and partly covering the thermo- couple. This assured that the conditions, during heating, would be the same for all tests. The quenching water was kept at sixty degrees 15 by means of frequent changes. all gears were left in the furnace for a period of thirty minutes. the fumace temperatures were very dlosely regniated. all measurements were made and checked by the In order to distinguish between the pieces, after heating, they were numbered with steel dies. 14 VI CONCLUSIONS. a careful study of the data of these tests has 104 the writers to the following conclusions: 1- It is shown by curves 6, 7, 8 and 9 that it is possible to obtain a shrink, on parts mich as gears, of about ten thonsandths of eo inoh by successive heat treatments. 2. It is shown by ocurvwe number 3s that it is possible to obtain a large mough shrink, on parts mach as geare, in one heat treat- ment to salvage a large portion of moh parts which have been machined oversise, because the average oversize error is not over two thougandths of an ingh. 3. That, as has been proven in previous research, a postion of the shrink is due to a change in volume of the iron in going frc.c one state to anoth er. 4. That, beeause of the observed teméanoy of the web and rim to cool first, a considerable portion of the shrink is due to the shape of the sample. M¢é 1 0014 e011 00045 0015 0035 2004 00015 005 20018 0083 e001 e0113 0036 e00335 00146 006 0015 0045. 0018 e0039 200728 0006 eO1l 002 0034 e004 0052 0046 0105 e003 Vit DATA ~ Differances. Gear fi. 4 5 »00089.001 -06889.005 .006 .008 2008 .006 2002 .002 20045 .0045 Gear #2. Gear 43 e .001 .0015 0045 .006 Gear #4. 0018 .0025 Gear #5. 0028 .0005 Gear #6. 000] .0005 1350° 6 e002 0013 0017 0008 ~0000 e000 ©0001 7 0015 0015 0015 0005 e001 001 e001 1375°. .002 0025 .0008 002 1425°. e00S1 0015 1500°. 0002 0049 13500°. 0005 0008 1500° 0055 005 1500°. 0004 «0000 0009 0026 e0000 0005 004 003 e002 e0019 e0025 0008 eO0016 0014 9 0038 210 0022 e00115.0003 -00175.0043 004 e0005 e001 00045 0018 0014 0015 20012 e0015 -0013 0018 008 0012 0008 0002 0006 .0008 001 20009 -0004 12 -0006 0024 0024 .0004 ~0008 0035 e000 e001 e000 e0002 00007 - 0008 12 0015 -000B 0053 0033 -001 e001 e0002 e001 e000 2000 ©0000 -0016 ~0005 16 Gear #1. AB xeo. to 15360°. i 2 3 4 5 6 6.1211 61203 6.1185 -0085 eOll 1.578 Heat to 1350°. «. in water at 60°. 6.1225 6.1236 621195 009 e014 12570 Tem. of gear brought up gradually with the furnace. keheated to 138°. 601115 621090 6.1078 00811 .013 1.5687 Oven up to temp. them gear placed in for 35 minutes end quemahed in water. Heheated to 138°. 62107 6.1035 6.1025 e011 016 1.3668 keheated to 1550° 6.1055 6.0965 6.0973 0017 O24 1-366 xeheated to 1350° 6.102 6.097 6.095 e022 2030 1.566 Reheated to 1350° 6.098 6.0925 6.092 e023 038 1.566 6.0965 6.0907 6.0885 0875 0365 1.5659 17 Gear #1. «as neo. to 13° (Continued ) 7 8 9 10 11 12 1.371 1371 10371 e682 e884 e881 Heat to 1360°. «. in water at 60° 123695 1.367 123678 8842 8846 - 6825 temp. of gear brought up gradually with the furnace. “xeheated to 1350°. 1.368 1e3661 1256675 e8845 e886 e883 Oven up to temp. them gear placed in for 35 minutes and quenched in water. Keheated to 139° 1.3665 1.3635 1.365 6802 e 8836 08777 Heheated to 138° 1-366 1.56365 1.561 e882 e884 e881 steheated to 1350° 1.365 1.363 1.5605 «880 8835 «880 keheated to 138° 12564 1.389 1.3595 8788 880 0879 1.363 1.356 1.355 8786 e860 | 8788 Gear ¥F2. As Received. Heat to 1376° Cu@oh in HO e a 2 3 4 5 6 7 8 9 10 60118 6.1182 6.1184 0.0 90.0 1e372 1.572 1.5715 1.571 .882 6.113 6.116 6.115 0.002 0.0025 1.3725 1.573 1.3698 1.3692 .8818 fenperature of water = 60°. (Continued ) 41 12 e8825 .882 e6815 .881 ‘jemperatare of water = 60°. 19 Gear #3. «as received. i 2 3 4 5 6 7. 8 9 10 Li 12 6.1141 6.1165 6.1205 .0O11 .0105 6.3717 1.3724 1.3714 1.3714 .8788 .879 .879 Furnace heated to 1375° before gear was put in, then raised above 1425° for 35 minutes and quenched in water at 60° from 1425°. 6.1123 6.1126 6.1166 .010 .012 1.3686 1.3709 1.36965 1.370 .8782 .880 .879 Heated to 1500° and qtenahed. 6.104 6.1084 6.1110 .0145 .018 1.3684 1.366 1.367 1.3655 .6878 .880 .879 Gear F4. i 2 601185 6.1162 601175 6.1156 7 8 1.573 1.5715 125722 1.5706 48 neceived to 1300° and juenched. 3 4 601141 00075 6.1095 eQ09 (Cont inued ) 9 10 1.38715 .879 1.3706 878 5 00055 2008 di -8788 879 6 1.5725 10372 L2 «8788 -8786 Gear <5. 1 2 3 4 6.1371 6131 6.1315 0004 6.1159 62120 6e121 0066 (Continued ) 7 8 9 10 1.372 1-372 10372 28804 1.367 125704 103704 8796 aB xecelved to 1500°¢ e004 ~0045 11 8793 «8786 2i 1.3715 1. 366 12 28796 -878 Gear #6. «8S Keceived to 1300° i z 3 4 5 6 601196 6.119 6.120 .03945 .005 1.8715 60116 60117 6.117 .090355 .00465 1.37 (Conti nued ) Measure Numbers 8 and 9 near edge. 7 8 9 10 al L2 1.3715 1.87168 1.3718 .86804 .880 .8799 123715 1.5704 123705 .&80 -8603 .6794 eee terete gp LL Le ae Ta) - : | he loa am ! 4 oe - LLL eae Pa As a 208 (CATA “17574 x N Q iN R x N \ VS AS N S K u 8S NY i . ror MICHIGAN AGRICULTURAL, COLLEGE MICHIGAN AGRICULTURAL COLLEGE (SPYILL SY PPUPLIOY ZL. i 5 Sigssssesesveiss goa agestgeres pessnerage ee: err ‘ : MICHIGAN AGRICULTURAL COLLEGE Fe ee NA ee aaah dae ra as wre ple A | MICHIGAN COLLEGE OGPARTMENT OF MATHEMA MICHIGAN AGRICULTURAL COLLEGE Pea ibeee so pers aes: : sais : as cee wel cs cee } a ESgSEL Meee Y a CPLLAY @ AO assy Oo Wea EE Ae dh te VA Vij uss yo oard oy gny #9 PUD WOAY B240{SIp al ee ILS, \ t 4 aon 570 10+ 7 39577 EY Se Lae Ei Sheth Ok eh a od hn aS Rear : Rae Cae a ae ALS , ti ane se Drains Pasd Sige ne eS A ee eee é f as ne pd K€ ts / YH Kb 'N a Kae YN : SN Fe yleg ae g eee i Pj A ak. 7 fae" LH) FG A =- ae : ii x 3AN § il E LIBRARIES i WN 3 1293 02236 0832 “S & ~ eS | 7 ~ ag ee, - ain eal ey 7. ion = / iy : : @ ; a OD ; * . ee MICHIGAN STATE LIBRARIES HH ANAT AT TNH NSAI TA 3 1293 02236 0832 MICHIGAN STATE LIBRARI ES | | | i} 3 1293 02236 0832 MICHIGAN STATE LIBRARIES MINIT IN 1293 02236 0832 2 me <_& = a ——— gee il 6 083 I ~ = aS - hae et 7 ~ i . c A a MICHI | } \ | | 3 1293 02236 0832 1IGAN STATE LIBRARIES ATE LIBRARIES MICHIGAN = L > WOVE AUN I} il HTH Wi | A | WEN | | | | | HA 3 1293 02236 0832 ST | | |