ENGINEERING RESEARCH INSTITUTE THE UNIVERSITY OF MICHIGAN ANN ARBOR Final Report A COMPARISON OF QUENCHING CHARACTERISTICS OF HIGH-TEMPERATURE AND SILICONE OILS C. A. Siebert E. B. Mikus Project 1685 ASSOCIATED SPRING CORPORATION PLYMOUTH, MICHIGAN September 1955

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN ABSTRACT A Type-430 stainless-steel bar was used to obtain cooling curves at the center of the bar upon quenching in a silicone oil and a high-temperature oil at bath temperatures from 2000 to 5000F. From the cooling curves the severity of quench for each oil for a given test condition was computed, using Russell's Tables. It has been shown that the high-temperature oil had superior quenching characteristics at all temperatures tested. The maximum H value for the high-temperature oil was 1.67 in.ol, whereas for the silicone oil the maximum value was 1.1 in.1; both occurred at 500~F. OBJECTIVE A comparative evaluation of the quenching characteristics of a silicone oil and a high-temperature oil. ____ ii

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN INTRODUCTION In this investigation two oils were compared over a range of temperature as to the severity of quench achieved by each when a bar of 1-1/8 -in. diameter was quenched in them. One of the quenching media was Park Chemical Thermo Quench Oil. The other was Dow-Corning Silicone 550 Oil. At room temperature the viscosities of these oils are different, the high-temperature oil being more viscous than the silicone oil, However, above 200~F, at practical oil-bath temperatures, both oils were very fluid. Both oils tended to fume at high temperatures, It was noted that the silicome oil began to fume at 350~F, and this became very prominent at 500~F. The fumes were odorless, nonirritating, and appeared as steam. The hightemperature oil began to fume at 4000F and at 5000F fumed to a greater extent than the silicone oil. There was some irritating odor to the fumes of the darker high-temperature oil. EXPERIMENTAL METHOD The technique used to evaluate the quenching characteristics of these two oils was to quench a bar of 1-1/8-in.-diameter steel in an agitated bath held at temperatures from 200~ to 500~Fo The center temperatures of this bar during quenching were recorded on a high-speed recorder with a chart speed of six in./min. From these data the severity of quench, H, was computed, using Russell's Tables.1 Material Tested.-In order to eliminate the effects of thermal arrests due to phase transformations on cooling, a stainless steel, Type 430, was used in this investigation. This steel has the beneficial property of a constant thermal conductivity over the temperatures used in this work.2 Method of Heating.-The test bar was heated in a vertical-tube furnace, 18 in. high, to a quenching temperature of 1600~F. The specimen was suspended from a stainless-steel tube of 3/8-ino diameter welded to the specimen. The tube in turn was fixed to a transite board so that the specimen was placed in the same position in heating and quenching for all runs. _________________________________ 1 __________________________________

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN The quench pot had. a capacity of five gal and was heated by an immersion heater. After the test specimen had been quenched in the oil bath, it was found that the bath had increased in temperature a maximum of 200F. Test Method.-To eliminate end effects, the test specimen was made six in. long and had a diameter of 1-1/8 in. The steel had a machined finish. From one end of the specimen a 1/8-in. hole was drilled three in. deep. A chromel-alumel thermocouple, 24 gage, was threaded into spaghetti tubing and inserted into the hole and butt welded to the center of the specimen. A stainless-steel tube of 3/8-in. diameter was slipped over the thermocouple and welded to the end of the specimen. The stainless-steel tubing in turn was fastened to a transite board so that the specimen during quenching had 3-1/2 in. of oil above and below the ends of the specimen. By this means, the specimen was placed in the same position during each heating and quenching cycle. Temperatures at the center of the specimen during quenching were recorded on a Brown Electronik Recorder with a chart speed of six in./mino The accuracy of this instrument was checked with a Brown potentiometer and was + 5~F. In making a test, the specimen, attached to the transite board and equipped with a thermocouple connected to the recorderwas suspended vertically into the vertical-tube furnace and allowed to come to a uniform temperature. This temperature varied slightly from run to run but was in the range from 1600~F + 10~F. The oil bath, maintained at the desired temperature, from 200~ to 500F, was adjacent to the vertical furnace so that the specimen could be removed from the high-temperature furnace and quenched vertically in the oil in less than one second. The oil bath was agitated by an electrically driven propeller. Very slight agitation was accomplished at the low temperatures due to the high viscosity of the oils. At higher oil-bath temperatures, however, a moderate degree of stirring was achievedo Some deviations were encountered in:-the cooling curves on duplicate runs, but these occurred mainly in the lower temperature range. The data used for calculating the severity of quench were from the upper portion of the cooling curves where good reproducibility was obtained on duplicate runs. ANALYSIS OF DATA The method used for computing H, the severity of quench, from the cooling curves is outlined by Austin.3 The tables taken from Russelll found in Austin's book3 were extrapolated and used in the form of a curve. These extrapolated data are shown in Fig. 1. In this figure, hL is plotted as a __________________________________ 2 __________________________________

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN function of T, the relative time ratio, for a center position of the specimen. The value of T is given by.=, art T L2 where a2 = thermal diffusivity (in.2/min), t = time to 1/2 temp (min), and L2 = radius of bar squared (in.2). The severity of quench, h, as defined by Russell, was used in Fig. 1 and is twice as large as Grossman's H. Both have units of in. —. In computing H, a value of thermal diffusivity must be obtained. For a Type-430 stainless steel a value of 0.007 to 0.0075 in.2/sec has been reported 2 In this investigation the thermal diffusivity was taken as 0.007 in.2/sec. The following calculation will serve to illustrate the method used to determine the severity of quench from cooling-curve data: The cooling curve shown in Fig. 4 will be used in this example. From this curve the half temperature is determined. (1605F - 350~F) + 350F = 977F 2 The time required to reach this temperature was 25 sec or 0.417 min. The relative time ratio, T, is a2t 0.007 in.2 60 sec 0417 min L2 sec min (0;5625)2 in.2 T = 0.554 Referring to Fig. 1, at a value of T = 0.554 the corresponding hL value is 1.04. The value of Russell's h then is hL = 1.04 h 1.0o4 = 1.85 in, (.5625) in. or Grossman's H is 5

- ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN 2H = h H = 1.85 =0.93 in.1. 2 The cooling curves obtained for the silicone oil are shown in Figs. 2-7 and those for the high-temperature oil in Figs. 8-13. The results of the calculations for the severity of quench are summarized in Tables I and II. A comparison of the quenching characteristics for the two oils is presented in Fig. 14. It can be seen that the high-temperature oil is superior in quenching characteristics to the silicone oil at all oil-bath temperatures tested. The difference appears to become greater as the temperatures of the oil baths increase. It has been shown4 that three definite stages of cooling exist when a metal surface is cooled in a liquid. The first stage is one of slow heat transfer due to the formation of vapor films that separate the liquid from the metal. This stage is prominent in liquids having low and sharp boiling points. The second stage is one of rapid heat transfer and is characterized by a bubble formation around the metal which continuously brings the liquid in contact with the metal as the bubble of vapor ascends through the liquid. This stage is the most effective. The third stage of cooling is by convection since the surface of the metal is below the boiling point of the liquid. A possible explanation for the lower severity of quench obtained using the silicone oil as compared to the high-temperature oil lies in the stages of cooling. The Dow-Corning 550 Silicone Oil does not boil but passes directly from the liquid to the gel state. The lower severity of quench in the silicone oil is probably due to the fact that the most effective stage in cooling, the boiling stage, is absent in this oil. In making a comparison between these two oils, the economic aspect must be considered. The cost of the silicone oil is about eighty times more than that of the high-temperature oil. This exceedingly high price will necessarily limit the use of silicone oil as a commercial quenching agent. ________________________________ 4 ________________________________

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN CONCLUSIONS A comparison of quenching characteristics of a silicone and a hightemperature oil at bath temperatures up to 500~F has shown: 1. Silicone oil begins to fume at 350~F; however, at 500~F it does not fume as much as the high-temperature oil. Both oils fume heavily at 500~F, but only the high-temperature oil exhibits an irritating odor. 2o The severity of quench obtained with the silicone oil was inferior to that of the high-temperature oil at oil-bath temperatures of 200~F and above. The difference became greater with higher bath temperatures. At 500~F the high-temperature oil gave an H value of 1.67 in.1' compared to 1.1 in.l1 for the silicone oil. 3. The high cost of the silicone oil, together with its inferior quenching properties, as compared to the high-temperature oil does not warrant its use as a commercial quenching medium. REFERENCES 1. Russell, T. F., "Some Mathematical Considerations on the Heating and Cooling of Steel," Iron and Steel Institute, Special Report No. 14, 1956, p. 149. 2. Sinnott, M. J., and J. C" Syhne, "An Investigation of the Quenching Characteristics of a Salt Bath," Trans., Amer. Soc. for Metals, vol. 44, 1952, p. 758. 5. Austin, Jo B., The Flow of Heat in Metals, Amer. Soc. for Metals, 1941, p. 120. 4. Scott, H., "The Problem of Quenching Media for the Hardening of Steel," Trans., Amer. Soc. for Metals, vol. 22, 1934, p. 577. --------------- 5

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN TEABLE I SUMMARY OF COOLING DATA FOR CENTER OF BAR SILICONE OIL Time to Run No. Bath Temp 1/2 Temp /2 Temp T L h H ~F ~F Min 1 200 911 0.484 0.642 0.81 1.45 0.73 2 200 902 0.483 0.641 0.81 1.45 0.73 3 300 950 0.450 0.598 0.91 1.62 0.81 4 350 977 0.417 0.554 1.04 1.85 0.93 5 400 1008 0.392 0.520 1.16 2,06 1.05 6 400 1000 0.412 0.547 1.07 1.90 0.95 7 450 1034 o0408 0.542 10.8 1.92 0.96 8 450 1026 0.402 0.534 1.11 1.98 0.99 9 500 1064 0.367 0.487 1.30 2.31 1.16 10 500 1058 0.400 0.531 1.12 1.99 1.00 11 500 1049 0.392 0.521 1.16 2.06 1.03 TABLE II SUMMARY OF COOLING DATA FOR CENTER OF BAR HIGH-TEMPERATURE OIL Time to Run No. Bath Temp 1/2 Temp 1/2 Temp T hL h H OF OF Min 12 200 903 0.433 0.575 0.97 1.72 o.86 15 200 907 0.417 0.554 1.03 1.83 0.92 14 305 958 0.400 0.531 1.12 1.99 1.00 15 305 958 0o383 0.509 1.21 2.15 1.o8 16 350 980 0.367 0.487 1.30 2.31 1.16 17 350 982 o.383 0.509 1.21 2.15 1.08 18 400 1008 0.333 0.442 1,53 2.72 1.36 19 400 1003 0.367 0.487 1.30 2.31 1.16 20 450 1030 0.317 0.421 1.68 3.00 1.50 21 450 1034 0.317 0.421 1.68 3.00 1.50 22 500 1058 0.300 0.598 1.88 3.34 1.67 ______________________ 6 ______________________

3.0 m Z m 2.5 ------------ _ 2.0 _ _ _ _ _ _ _ 1I.0 0.5 -m_ v0. 0 0 0.5 1.0 1.5 2.0 2.5 Fig. 1. Extrapolation of Russell's Tables for center bar position Z and unaccomplished temperature increment, U = 0.5.

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 1800 I - 1600 140 0 8 1 1200 1-I000 0 20 40 60 so ---- 120 ITIME - SCONDS ~~~~~I.-u~I 200 ------- 0 20 40 60 80 1OO 120 TIME- SECONDS Fig. 2. Cooling curves obtained, with silicone oil at 200~F. ______________________________ 8 ______________________________

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 1800,.-, - 1600 l — 1400 1200 '" - - L. 1000 20 4608102 i \. 1 -800 200 0 20 40 60 80 800 120 TIME - SECONDS Fig. 5. Cooling curve obtained with silicone oil at 300~F. 9

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 1800 1600 - -- 1400 1200 I 1 -000 40 0. \ 800 4 400 --- ' — 200.... --- 00 20 40 60 80 100 120 TIME - SECONDS Fig. 4. Cooling curve obtained with silicone oil at 3550~F. ~~____~______________ 10 --- —--------

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 1800 1600 1400 1200 * 1000i IS 800 2 600 5 400 200 0 20 40 60 80 100 120 TIME - SECONDS Fig. 5. Cooling curves obtained with silicone oil at 400~F. 11

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 1800 1600 -- 1400 "- 40 — 1200 looo 800 600 400 0 20 40 60 80 100 120 TIME - SECONDS eig, 6, ~oo~~ng owrvs ob~h@4o with leeft~ @ at ^Q?, _____________ -- -— 8.@ 'r^

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 1800 1600 - 1400 1200 F 800 ----— e 600 0 —. 400 200 0 20 40 60 80 100 120 TIME - SECONDS Fig. 7. Cooling curves obtained with silicone oil at 5000F. 13

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 1800, 1600 1400 - 1200 - 20-.- 1000 oo I600 200 0 20 40 60 80 100 120 TIME - SECONDS Fig. 8. Cooling curves obtained, with 'high-temperature oil at 200~F. 400 - '....... ',

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 1800,_. 1600 1400 1200 ', -- 800 40 60 80 100 12 6. 1005 200 0 220 40 60 80 100 120 Fig. 9. Cooling crves otined with high-temperature oil -------------- 15 ----

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 1800. 1600 1400 ' ' 1200 1000 S oo D 16 400 0 20 40 60 80 100 120 TIME SECONDS Fig. 10. Cooling curves obtained with high-temperature oil at 350~F.

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 1800,,, i 1600 1400 --- Sooo u.J 800 --- l ---,, Fig. 11. Cooling curves obtainedwith hightemperature oil9 600 - --- 400o-at 400-F17 200 0 0 20 40 60 80 100 120 TIME - SECONDS Fig. 11. Cooling curves obtained. with high-temperature oil at 400OF. 17

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 1800, 1600 ". 1400 -- 1200 U. I 10 0 I — 800 a. 200 ----- at 4500. 200.. 0 20 40 60 80 I00 120 TIME - SECONDS

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 1800 11 - 1600 ' 4608 ---1400 1200 --- - 1000 TIMF - q\riun^ 0. I I-.600 ' — 400 ' 0 20 40 60 80 100 120 TIME - SECONDS Fig. 15. Cooling curve obtained with high-temperature oil at 500~F. 19

1.8 H.T. m OIL Z 2,6________________ rn cn 1I6 o ee Ic 0 Cn~~~~~~~~~~~~~~~~~~~~~~~~~~~~~r I.4 m --- —-------- 01 0 OI____ _ _ ____ 0.2 0 w -t w~~~~~~~~~~ 00 u.. I SILICONE _ ^ 1. 0 i Oi L > A.O --- —— ^' - - — _____ ---- _ _ _ _ w ~0.6 -- ^0 Cn 0.6 --- —------------------------------------— _ 100 200 300 400 500 600 I TEMPERATURE F >F Fig. 14. Effect of oil-bath temperature on the severity of quench obtained with two oils..