~ ail anand aie mee 7 THESIS PBN RRNA OA DEES DS R. D. KEAN, A. L. MCCLELLAN Ds THESIS Preface. In this thesis we have endeavored to check previous investigations and comparisons of cool- ins media for hardening steel, acaing some interesting data in comparincs the heat absorbing power of cooling media, at different temperatures, for our particular test piece. The work was done by vractical tests under conditions which were analogous to those in modern shops. In our theoretical work ve used the followinz books and articles as references. ractors in Hardening Tool Steel, by John A. Matthews and Howard J. Stage, Jr. Handbook for Mechanical ungineers, by Kent. Journal of the Iron and Steel Institute, 1908, II p.<08. An article by Benedicks. Willie on Heat Treatment of Steel in The Proceedings of A.S.T.M., pe<8, Vol. XV. 101885 Part I Descrivtion of anpnaratus and method of running the tect. Part II Resulting curves and their analysis. Part III Comparison of absorbing power of cooling media at various temveratures, with illustrative curve. -L- DESCRIPTIVi APPARATUS AND PUTHOD OF RUNDING THs Trot. The test niece was turned from a viece of cold rolled steel to the dimensions given or the drawing. The piece was then bored for the thermal counle, the hole tavped and a piece of three-cuarter inch pipe, twelve inches lons, screwed into the hole. This pine acted as a handle and as a conduit for the thermal couple lsacs. It also admitted withdrawing the couvle for insyection, if necessary. All thermal couples used, were counles made from tron and constcntan wire. The source of heat was a cis furnace used in connection with a leau bath, which was kept at « temperature of 1200°F. rx} ive gallons of each medium wes used in each test. The followinc method was used through- out the test. Tne test piece was immersed in the lead bath together with anotner thermopyle. ‘hen the couvle insice the test piece read 1s00°F., and the counle in the lead pot read 1£00°F., the tect niece was removed and quenched to 700°F. in the ouenching medium under consideration. At the start the mecium was maintained at room temmerature. The time that it took the test piece to rall from a temversture of 1Z00°F. to 700°F. was noted with the aid of a stop watch. The cooling or the test piece in the e medium naturally raised the temperature of the mediun. The test piece was then replacea in the leec, heated to 1Z00°F., cuenched in the medium et this known hish- er temperature anc the time av vain taken vith the ston watch. These onerations were continued until the quenching media had attained such 2 temnerature that excessive smokin;; or boiling prevented its further use. Durin~ the above operations care was taken that the test niece wes immersed the same emount each time both in the leaa bath anc in the coolins media. The test piece curing ell the quenching rested on the bottom of the container of the medfum and the medium was not etirrec excest when takin: its temn- ersture just before immersing the tect viece. -4— RESULTING CURVIS AND THUIR ANALYSIS. In plotting the followin= curves, the results obtained by our exneriments, time in seconds, Yor a fall from 1500°F. to 700°F., were vlotted ageing the temperature of the cuenchins medium. An analysis of the curves is interestin C3 fairly constant ™ The M.A.C. well water, Fis. II, hag cuenchins rate up to 1£O°F. At 150°F. the slove of the curve is quite merkeac. Uxcecsive toilin. vrohtibits further use of the medium over 180°F. A saturated salt solution (NaCl) gives & ouicker rate of cooling than water and a feirly constant rete up to 125°F. From 125°F. to 190°F. where excessive boilin: occurred, the slove is very decidec. The soda solution, Fig. I’, Nakccs 7 HeoO, has an extremely cuick rate of quenching and has the veculiar characteristic that it cools much faster at moderate temperature (385°-110°) than at either lower or higher temperatures. Foaming and boilins was excessive at 1£0°F. This medium would - orobably ceuse crackin; of steel “here the plecese wers of considerebls size, vut for small pieces ..nere extreme harcness was desired, should prove satisfactory. The sveed of cooling is much ¢freater than even the salt (NaCl) solution. - Merosene oil, Fis. V., at low temmer- atures hes a cuick cooling rate, vor en oil, but de— creases ét a nearly uniform rate tnroughout its use- ful ranzge. At 250°F. end above unnleasant gasses are Siven off ana the use of this oil should be confined to temveratures below 250 F. A temperature above 400°F. was reached before the oil caucht fire. "Renown" engine oil, Fis. VI., ig a fairly thick oil similar to gas ensine cylinder oil. Its slope is less than kerosene oil but is much slower. “here the piece recuired 100 seconcs to cool in this oil at 70°F. it cooled in 70 seconds in the kerosene. The rate of cooling was nearly constant up to 360 Fe but smoking was excessive above 300°F. and this medium should not be used on this account above that temperatures. The lard oil used, Fig. VII., had been in use for some time in the M.A.C. forge shops. This oil is particularly interesting in that throuznout the entire range of temperature used, the rate of cooling was constant. The rate was slower then any of the other oils, being about 10% seconds for all temv- eratures. Even at the extreme temperature of 420°F. ~6- [0ut }~- ™! go the rate of cooling varied but little. 300°F. the practical limit for this oil as above that a gas is liberated which is very annoyins. Of the extremely heavy, viscous oils it mi-ht be well to say that those of the steam cylinder Oil type are unsuiteovle as coolin: media.’ Heavy steam engine cylinder oil under test was unable to carry the heat away from the test piece rapidly enough and the resulting highly heated sasses which were liberated caucsht fire. Tnere is a possibility however that this types of oil if used at a temperature somewhat above room temperature might then be used, but we believe that the viscosity is too great for a good cooling medium. ~7~ CCMPARIZON OF ABGORBING POWER OF COOLING VaDIA AT VARIOUS TIMPERATUNES WITH ILLUGTAATIV: CURVE. In determinins the heat flow per ec. in. of surface from the test niece, ~e referred to tent's rTenabook for the specific heat of steel. Observations by Dr. I. P. Obenhoffer, cuotead in the handbook, gave an averace of .155 es the snecific heat between the points 700°". and 1200°F. The weight of the test niece (without handle) was co10 grams. Thererore the total heat given uv in cooling from 1200°F. to 7O0°F. eauals (1.00-700)x.310 x .155 equals 156000 calories. The sq. in. of surface of the test viece (without hanale) equals the circumference times the length vlus the area of each end. aquels (Umr x = 1/3)+ cier® Squels ( tex © 30/16 x 41/3) + We seals 45,0 sc. in. Therefore the number of calories fSiven orf per sc. in. of surface ecusls 156000 ecuals S6c0 cal. In nlotting the accompanying curve, the calories absorbed ver sc. in. of steel surface ver second are vlotted against the temperature of the —3- quenching bath decree F. This gives a series of curves which, after one knows how many calories ver second ouzht to be tazen away for a siven hardness, should be valuable. As en examole, supnosce we have the quenching mediums at hanc that are listed on the curve ana they are at varioue temrsratures becaure oF recent use. Suppose that we have a niece to be harcenea anc for the desree of hardness desired 45 calories per second should be taxen avay. A glance then at the curve shows that slow, while tJ oO Ss oO 3 es a ) ) oO te }~ m9 5 Ou om 69 ry Cu oO j- o by Q ct oO oO in a soca colution is eltosether toc fast. The choice lies between well water, ealt water and -.crosene oil. salt water and well water, if cool, arc too fast and the voint at which they carry avay 45 calorics ver second is et a temperature which is almost prohibivive. Therefore, if we have irerosene oll at amy tempersture betveen SOF. and 1l30°F. it will answer the purpose. A voint ‘which has besn proven a sreat number of times but which is neverthelces interesting to note, is the fact of ths tendency of the test piece to become round after reveatec heatings anc coolings. In all the test nieces was cuenehed some 60 times. he @ iazeter of 8) fs Upon examination, efter these tests, thi se the test piece at the center was found to be 1/8 in. larger. The disceter of the piece at the 56/16 in. lees than originally. The hole in the niece where the thermal couple was inserted had closed up so that the thermal couple coulc not be withdrawn, althoush the hole was ruch too larse originally. ze Te AL Sh A BI te Viti i GO the 750 et Brn sents Bee. eets ios dancreseccetuyhene erray Sauk nee eae 14 ee Kerosene Oil. Saoneaaees % 4 e abeh btrhesei: Sabss$ Sascasr 2 6 th a3 es » Tit a - anak cs TST ht 8 + gee 7 . a ‘ &> at bes bes oe eeded-> t es are Pere ger. Ht 4 eve ss mee : i“ ao e4Fre Feee Satthaw? she b Ob Er © JHeabee ee soebet aan EPC LASE OY hoes : ome os ae de bos ced tobe fl PY a re VEY SD) ae at “ Lee = On A) ee oe eo 3SS° Boiled So Diu” BICARBONATE a el g — Ao 7 Zao° VA ome 2) TEMP MEDIUM TIME of COOLING — (Aa Sd OA eae c= i ae ee) Th ae a es ae i oe /os°* oe a ee / oe 72 ee oe 2rye ieee fae , 1 Ee 8 Pie W127 Paw os heey >| a eee, Fe Fa acini y ee ees} en 3 Ro oot ee se = ge fas aD Ya ce ene a 43 al a acm at KEROSENE TEMP.MEDIVUM all ot 1,30 a ae Pcie 7a he 45" esl Z8o° ~ Bi?’ 420 BOILING SALT SOLUTION he a Pes ist “3 eo ere . 1/20 Ye ee 2 hry = fo os ; ie i . er. foe Wie: rs | pe oe | er Se - Foy >) =, vs >, oe a —*. ~~ . oe ine — a >=. ~ = in ie te, c , 7. i ¥ . nS TES PE a * Sonali satay ne ae a nae d 3 TAP WATE Rr Pag — —_ es on ae . 95° ie a 200° Boiling LARD Witkes (om FEO" yc ee 200 pe Be | aye: 2978 320 . 7, 370 385 Cie Mes H10 ae) 440 450 470 BURNED i ay ‘*s 7 eee i sae om at Y . se f a * me fee Fe os + fh id Pa fee - Dn a es SEC - 26 ae , pe ae sey a Eee oe tite 7 kOe ra = cn " “> - oe | “A : r ti < ih Ae a Cr ae = ys . oy Mere 1] - - . a2 4: ~ ‘¢ ‘e + . : , , ef « 4 on » . , ‘ . ‘ ‘ , * . , ‘ ” °