High Compressive Residual Stresses in Through Hardened Steel Parts as a Function of Biot Number N..KOBASKO 1, M.A.ARONOV 1, KATSUM CHTAN 2, MAYU HASEGAWA 2, KENRO NOGUCH 2 1 Q Technologies, nc., Akron, USA 2 demitsu Kosan Co., Ltd., chihara, Chiba, JAPAN Abstract:- n the paper, a correlation between residual surface stresses and the Biot number is discussed. Based on the results obtained by the authors and other researches, it is shown that during intensive quenching high compressive residual stresses can be formed in steel parts regardless of whether the material experiences phase transformations during quenching or not. t is underlined that intensive quenching occurs when the film boiling process is absent completely. The film boiling process can be eliminated by adding special additives to water and water solutions resulting in the development of high compressive residual stresses in the surface layer of steel parts. Q Technologies nc. of Akron, USA, demitsu Kosan Co., Ltd., of Chiba, Japan and ntensive Technologies Ltd, of Kyiv, Ukraine have initiated a joint project on developing of special additives which can eliminate a local and full film boiling process during quenching. Key Words:- Compressive residual stresses, Biot number, Local and full film boiling, Additives, ntensive quenching. 1. ntroduction Residual surface compressive stresses are observed usually after quenching of carburized parts or after induction case hardening of steel parts. When steel parts are heated through and quenched conventionally (for example, in oil), surface tensile or neutral stresses are generated. This was a conventional wisdom in the field of heat treatment. For intensive quenching methods, it is not necessarily true. During intensive quenching, high compressive residual stresses are formed in the surface layer of steel parts [1-5]. This phenomenon was discovered for the first time by authors [1] in 1983 and was discussed at the FHTSE Congress in 1985 [2]. n the present paper, data on residual surface compressive stresses obtained by different authors for through hardened parts are discussed and analyzed. Cylindrical samples made of AS 1045 and AS 52100 steels were used in these experimental studies. Computer simulations were also conducted for these samples [1-3]. CCT diagrams and data on mechanical properties vs. temperature for the above steels were available for computer modeling. The results of these studies were presented to the conference [2] and were published in the book [4]. 2. Residual surface stresses versus Biot number A correlation between residual surface stresses and a generalized Biot number was obtained by authors [1,2] and presented in Fig.1. Note that the hoop stress changes from tensile to compressive at the Bi V number of about 4.3. Note also that the Bi V number value is directly related to the part cooling rate. Fig. 1 Residual surface hoop stresses for a cylindrical specimen versus a generalized Biot number Bi V, SBN: 978-1-61804-065-7 35
An intensive quenching process can be implemented by two ways: a) using a nucleate boiling process while film boiling is absent completely, and b) using a direct convection mode of heat transfer (in this case, both, the film boiling process and the nucleate boiling process are eliminated). The criterion for eliminating nucleate boiling process is the following [4]: Bi= 2( 0 ) (1) + uh Where 0 = T 0 TS ; uh = TS Tm ; is calculated from the following equation: 0.3 1 2λ( 0 ) = (2) β R A generalized similarity equation (3) for a forced flow of the liquid in circular channels of any cross-section shape can be used for calculating of convective heat transfer coefficients (HTC) [17, 18]: Nu 0.8 0.43 0.25 = 0.021 Re Pr (Pr Pr ) ε (3) For quenching, the liquid (quenchant) properties used for calculating Nu, Re, Pr and Pr numbers are determined for the average quenchant temperature in the quench chamber. A characteristic size d (used for calculating Nu and Re numbers) is equal to: m m sf m e eq Fig, 2 Production high-velocity Q system An evaluation of the intensive quenching process has been started also at the Bremen University in Germany [5, 6, and 7]. A picture of the laboratory type quenching system used is shown in Fig. 3, while a system schematic presented in Fig. 4. 4S d eq = (4) u Since 1999, Q Technologies, nc. has designed and built several production high-velocity intensive quenching (Q) system for quenching of different steel parts (automotive shafts and gears, bearing products, coil springs, torsion bars, etc. A direct convection mode of heat transfer is used for quenching steel parts in all these systems. High compressive residual surface stresses developed in steel parts processed in the Q systems even in though hardened parts. As an example, Fig. 2 shows a picture of the Q system designed for processing parts of up to 200mm diameter and up to 500mm length and installed at Euclid Heat Treating Co. of Cleveland, Ohio, USA. Fig. 3 High-velocity quench system installed at Bremen University in Germany SBN: 978-1-61804-065-7 36
The authors are planning to continue this study to evaluate more accurately boundary condition during intensive quenching. t should be noted that residual stresses depend on the carbon content in steel, martensite start temperature and on intensity of cooling characterized by the dimensionless Biot number or generalized Biot number. Such dependence is shown in Fig. 1 above and Fig.5. Both figures show the same tendency of the residual stress behavior depending on Biot numbers [4-7]. 3. Transient nucleate boiling process produces intensive quenching Compressive residual stresses are formed during nucleate boiling process when film boiling is eliminated and during intensive direct convection cooling (see Fig. 6). Fig. 4 Schematic of high-velocity system for quenching cylindrical specimens installed at Bremen University in Germany The inner channel diameter of the above quench system (Fig.4) is 40 mm and the diameter of specimens used was 20 mm. The water flow velocity is up to 19 m/s. Prior to calculating residual stresses, a cooling capacity of water flows were investigated and CFD modeling was conducted [5, 6]. Results of residual stress calculations are presented in Fig. 5. Fig. 6 Scheme of boiling process during quenching when film boiling is absent and film boiling exits [4]. A duration of the transient nucleate boiling process can be calculated from the following equation: τ nb = 0.24k+ b ln K a (5) Here τ nb is a duration of the transient nucleate boiling process; k is 1,2,3 for plate, cylinder and sphere; b = 3.21; is calculated from Eq. (2); is calculated from the following equation: Fig. 5 Axial residual stresses versus Biot number received by computer simulation and experiments for AS 52100 steel [7]. 1 = [ αconv( + uh )] β 0.3 (6) SBN: 978-1-61804-065-7 37
β = 4.3 ; α conv is a convective HTC; uh is an underheat; K is a Kondratjev form factor; a is a thermal diffusivity. Below is an example of calculations of the nucleate boiling process duration for the cylindrical specimen used by demitsu Kosan Co, Ltd for evaluating intensive quenching in water at 20 o C flowing with 1.5 m/s. A diameter of the cylindrical specimen was 28mm; an average heat conductivity of steel was 23W/mK; convective HTC was 6,500 W/m 2 K; thermal diffusivity a was 5.46x10-6 m 2 /s. The sample was heated up to 850 o C and quenched in the flowing water. Taking into account that k=2 and b = 3.21, we obtain: = 1 4.3 ( 750 ) 2 23 0.014 0.3 = 19.08 Fig. 7 Cooling curves obtained in demitsu Kosan Co., Ltd. lab for cylindrical specimen 28 mm diameter and 112 mm long when quenching in water flow 1.5 m/s at 20 o C. = 1 3 4.3 0. [ 6,500 ( 80+ )] = 12. 59 K = 33.89 x 10-6 m 2 /s τ nb 19.08 33.89 = 0.48+ 3.21ln = 11.26sec 12.59 5.46 The results of experiments are presented in Fig. 7. As seen from the figure, a full film boiling process is absent and the surface temperature during the transient nucleate boiling process is maintained at the level of the boiling point within 11.7 seconds. These results coincide well with the simplified calculations according to which a duration of the transient nucleate boiling process is 11.26 seconds. During the nucleate boiling process the real HTC is within 50,000 W/m 2 K 200,000 W/m 2 K [4]. t means that during the transient nucleate boiling process a very high intensive quenching can be achieved if the local and film boiling processes are absent. Local film boiling and full film boiling can be eliminated by adding special additives to water and water solutions. Q Technologies nc of Akron USA, demitsu Kosan Co, Ltd of Chiba, Japan and ntensive Technologies Ltd of Kyiv, Ukraine initiated a joint study on developing such additives. Using experimental data presented in Fig. 7, the effective HTC, Kondratjev number Kn and real HTC were calculated. The results of calculations are presented in Fig. 8. Fig. 8 Effective HTC (a), Kondratjev number Kn (b), and real HTC versus time obtained in demitsu lab during quenching of cylindrical specimen of 28 mm diameter and 112 mm length in water flow of 2 m/s at 20 o C. SBN: 978-1-61804-065-7 38
As one can see from Fig. 8, the real HTC is greater than the effective HTC. t should be noted that the effective HTC cannot be used for accurate calculations of the part surface temperature during quenching. t can be used only for calculating of the part core temperature or a cooling time. 4. Discussion n this paper, it is shown that high compressive residual stresses are formed at the surface of steel parts during intensive quenching. nvestigations conducted in the USA, Germany and Brazil show that compressive residual stresses could be very high (up to 900MPa, see Table 3 below). Moreover, as shown in [8], compressive residual stresses could be developed even in stainless steels (with no martensitic transformation) if the cooling rate is high enough (see Table 3 below). During a transient nucleate boiling process, the HTC should be calculated as a ratio q/(t- T S ), because only the overheat T-T S effects the bubble growing process. n this case, the HTC is high enough to create compressive residual stresses in the surface layer of steel parts. t means that eliminating of the local film boiling and full film boiling processes during quenching can result in the development of high compressive residual surface stresses. That is why Q Technologies, nc., demitsu Kosan Co., Ltd. and ntensive Technologies Ltd. are working on a joint project on the development of special additives, which could effectively eliminate film boiling during quenching. Along with creating compressive residual stresses, the elimination of local film boiling decreases distortion of steel parts. n addition, the full elimination of film boiling increases the hardness of steel parts after quenching and improves mechanical properties of material. Table 2 Residual surface compressive hoop stresses for steel parts which were intensively quenched through (results of experiments [4]) Steel part 52100 Roller 3 (76 mm) 52100 Roller 1.8 (46 mm) 4140 Kingpin 1.8 (46 mm) S5 Punch 1.5 (38 mm) Residual hoop surface compressive stresses, MPa -840-900 -563-750 Table 3 Hoop and longitudinal stresses vs. size of tested samples in stainless steel after intensive quenching [8]. Dia, mm 25 37 50 63 σ -362-424 -470-444 hoop σ -332-480 -455-502 longt n all above experiments, compressive residual surface stresses were formed in the cylindrical specimens of different diameters when a high temperature gradient was obtained through the surface layer of the specimens [4-8]. A mechanism of the formation of compressive residual surface stresses is described in Ref. [4]. ntensive quenching methods, which explore the nucleate boiling process and direct convection, are patented and have several Know-How [9, 10]. 5. Summary 1. High compressive residual stresses at the surface of through hardened steel parts during intensive quenching are caused by a high temperature gradient in the part, which can be characterized by a conventional Biot number or a generalized Biot number. 2. f a local and full film boiling process is eliminated during quenching in water, an intensive quenching process can be achieved since the temperature gradient during a transient nucleate boiling process is very high. This is because the part surface temperature is maintained at the level of the boiling point almost from the very beginning of cooling. 3. For low and medium carbon steels, a nucleate boiling process can be used to produce high compressive residual stresses in the surface layer of parts. For parts made of high carbon steels, it is better to use a direct convection intensive quenching process for developing high compressive residual surface stresses. 4. Q Technologies nc, Akron, USA, demitsu Kosan Co., Ltd., of Chiba, Japan, and ntensive Technologies Ltd. of Kyiv, Ukraine have developed methods for critical heat flux densities evaluation to be used for designing special additives to water for the elimination of the local and full film boiling processes during quenching, The elimination of the film boiling process during quenching will result SBN: 978-1-61804-065-7 39
in high compressive residual stresses in the steel parts surface layer. 5. A transient nucleate boiling process produces intensive quenching if both the local film boiling process and the full film boiling process are absent. References: [1] Kobasko, N.., Morganyuk, V.S., Study of Thermal and Stress-Strain State at Heat Treatment of Machine Parts, Znanie, Kyiv, 1983, 16 p [2] Kobasko, N.., and Morganyuk, V. S., Numerical Study of Phase Changes, Current and Residual Stresses at Quenching Parts of Complex Configuration, Proceedings of the 4th nternational Congress of Heat Treatment Materials, Berlin, Vol. 1, 1985, pp. 465 486. [3] Kobasko, N.., ntensive Steel Quenching Methods, n a Handbook: Theory and Technology of Quenching, B.Liscic, H.M.Tensi, and W.Luty (Eds.), Berlin, Springer Verlag, 1992, p 367 389 [4] Kobasko, N.., Aronov, M.A., Powell, J.A., and Totten, G.E., ntensive Quenching Systems: Engineering and Design, ASTM nternational, West Conshohocken, 2010, 252 pages [5] Rath, J., Luebben, Th., Hunkel, M., Hoffman, F., Zoch, H.-W., Basic researches about the generation of compressive stresses by high speed quenching, HTM J. Heat Treatm. Mat. 64, 6, 2009, pp. 338 350 (n German). [6] Rath, J., Luebben, Th., Hoffman, F., Zoch, H.- W., Generation of compressive residual stresses by high-speed water quenching, The 4 th nternational Conference on Thermal Process Modeling and Computer Simulation, Shanghai, China, June, 2010. [7] Hoffmann, F., Lübben, Th., Rath, J., Frerichs, F., Simsir, C., Use of Dimensionless Numbers in Heat Treatment, Proceedings of the 19 FHTSE Congress, Glasgow, UK, 17 20 Oct., 2011. [8] Penha, R.N., Canale, A.C., Canale, L.C.F., Evaluation of an intensive quenching system through the heat transfer characterization, Proceedings of the 19 FHTSE Congress, Glasgow, UK, 17 20 Oct., 2011. [9] Kobasko, N.., Ukraine Patent #56189 [10] Kobasko, N.., US Patent # 6,364,974B1 SBN: 978-1-61804-065-7 40