" The scientific work funded from resources of Co mmittee of Scientific Investigations in years 2003 ÷2005 as investigative project" USE OF AQUEOUS POLYMER QUENCHANTS FOR HARDENING OF CARBONITRIDED PARTS gorzata Przy , Wojciech G stwa and George E. Totten (1) Pozna University of Technology; Pozna University of Technology; Pozna ; Poland: (2)G.E.Totten & Associates, LLC, Seattle, WA USA; ABSTRACT Water, 10% aqueous polymer solutions and a conventional quench oil were evaluated to harden carbonitrided parts. The resulting as-quenched and quenched and tempered microstructures were compared. Ty pically, the gas carbonitriding process produced (carbonitrides) – martensite – retained austenite in the case and martensite – solution of a polyalkylenr glycol quenchant was used. KEY WORDS Carbonitriding, quenching medium, polymer quenching medium, structure, properties " The scientific work funded from resources of Co mmittee of Scientific Investigations in years 2003 ÷2005 as investigative project" INTRODUCTION Properties such as: hardness, impact strength, tensile strength and fatigue bending strength of carbonitrided parts of machines depend on proper selection and maintenance of two main operations: austenitizing and quenching. The quenching process conditions affect end-use properties such as: type and value of internal stresses, mechanical properties and the extent of hardening deformations [1], [2]. The link between cooling process and properties of hardened carbonitrided components is structure, especially retained austenite and carbides content. Table 1 summarizes the influence of these and other Structural factors Hardness Fatigue strength fatigue Cracking resistance Wear resistance Residual austenite As it increases hardness diminishes Preferred limits 25- It favours up to =70% influence No influence Carbonitrides Low sensitivity Influence Little influence The lack of influence Increases resistance. The scope of the use of aqueous polymer quenchants to quench a wide " The scientific work funded from resources of Co mmittee of Scientific Investigations in years 2003 ÷2005 as investigative project" transformational stresses leading to cracking and poor dimensional control. Additional deficiencies of water are non-uniform wetting and corresponding high thermal gradients and susceptibility to contamination. From a literature review [1], [39] - [51] it is clear that oil has become the most commonly used quenchant because it provides optimum hardness with maximum dimensional control in spite of its deficiencies. There are many different quench oils available however, a quench oil is typically classified as either a conventional (slow) oil or an accelerated (fast) oil. In addition there are high-temperature (hot or martempering) oils. As noted above, mineral oils exhibit many deficiencies due to the non-biodegradability and toxicity of the Taime; t [s] Temperature; T [°C] Fig. 1. Cooling curves received for different positions within an instrumented cylindrical probe: a.) The solid line is for water; b.) The dashed line is for a 5% aqueous solution of a PAG polymer quenchant [8]. The influence of the physical properties of the quenchant and the probe on " The scientific work funded from resources of Co mmittee of Scientific Investigations in years 2003 ÷2005 as investigative project" summarized in Table 1. In addition to excellent uniformity of surface heat transfer, other advantages of aqueous PAG quenchants specifically and polymer quenchants in general include: high degree of against fire protection; biodegradable and non-toxic particularly for algae and fish [1]; no soot and smoke during quenching; fire resistance; broad processing latitude. TABLE 1 . Effect of Fluid and Metal Property Variation on Quench Severity Effect on Property Variation ( ↑ = Increasing, = Decreasing) Fluid Property Type of Quenchant Addition of Additives Increasing Agitation (v) Increasing Bath Temperature (T Metal Property Increasing Thermal Diffusivity (a) Increasing Cross- Section Size Increasing Surface Roughness Increasing Surface Oxidation 1.Parameters: t : time when wetting starts; t : time when wetting is finished; time interval of wetting [s]; : heat transfer coefficient. This paper discusses the use of an aqueous polymer quenchant to harden carbonitrided steel. The results are contrasted to those obtained with water and oil. " The scientific work funded from resources of Co mmittee of Scientific Investigations in years 2003 ÷2005 as investigative project" " The scientific work funded from resources of Co mmittee of Scientific Investigations in years 2003 ÷2005 as investigative project" Figure 2 – Schematic illustration of the Tensi Agitation System (all dimensions are in mm). Figure 3 – Illustration of flow uniformity in the Tensi Agitation System. The flow strings show that the flow in the quench zone is uniform without bubbles or twist.. The flow rate in the figure on the left is: 0.25 m/s and 0.6 m/s in the figure on the right. This agitation device was used for all quench media reported in this study. vaporizable liquid quenchants (water, oil and a 10% aqueous polymer solution) are used which typically exhibit boiling temperatures between 100 and 300°Cat " The scientific work funded from resources of Co mmittee of Scientific Investigations in years 2003 ÷2005 as investigative project" surface temperature is less than the boiling point of the quenching fluid at which point convective cooling begins. The nonsteady state behavior is indicated in Figures 4 and 5 where it is seen that all three cooling mechanisms occur on the surface simultaneously: film boiling, nucleate boiling and convective cooling. This is significant because the heat transfer coefficients ( ) for these cooling processes vary by an order of magnitude (or greater) thus producing substantial thermal gradients during the cooling process which lead to increased residual stresses. (Typical heat transfer coefficients for these cooling processes are: 100 – 250 W/(m 10 – 20 kW/(m K), and CONV 700 W/(m [63]. Figure 4 – Wetting process of a cylindrical CrNi-steel specimen (25 mm dia. x 100 mm) quenched from 850°C into distilled water at 30ºC with an agitation rate of 0.3 m/s. Figure 5 – Wetting process of a cylindrical CrNi-steel specimen (25 mm dia. x 100 mm) quenched from 860°C " The scientific work funded from resources of Co mmittee of Scientific Investigations in years 2003 ÷2005 as investigative project" Figure 7 – Wetting process of a cylindrical silver specimen (12.55 mm dia. x 45 mm) quenched from 850°C into at 10% aqueous solution of a polyalkylene glycol solution at 25ºC without agitation. (An agitated specimen would exhibit substantially more uniform film formation). Many aqueous polymer solutions, such as the 10% solution of the polyalkylene glycol quenchant used for this work exhibit a pseudo-Newtonian cooling process. This means that the cooling mechanism is essentially the same all over the surface of the cooling part at any point in time. This is illustrated in Figure 7 [64]. In this case, the hot metal surface is covered by a uniform " The scientific work funded from resources of Co mmittee of Scientific Investigations in years 2003 ÷2005 as investigative project" nucleate boiling process ends, the polymer is redissolved into solution and the part is cooled by convection. It is important to note that quenching into an aqueous polymer quenchant solution results in a fundamentally different and more uniform cooling process than when quenching into either water or oil. Figure 8 – Surface temperature decrease of a cylindrical steel probe (25 mm dia x 100 mm) at various distances from the lower end. The effect of non-newtonian versus Newtonian wetting on thermal gradients and residual stress formation were described by Tensi [64] and Narazaki, et. al. [65]. The influence of the a non-newtonian wetting process on the temperature distribution within the quenched specimen is shown in Figure 8 where the temperature measured near the surface of a submerged cylindrical specimen at different locations from the lower end are shown [65]. The wetting front requires about 18 s to arrive at a height of 80 mm. If there is an explosion-like wetting of " The scientific work funded from resources of Co mmittee of Scientific Investigations in years 2003 ÷2005 as investigative project" Figure 9 – Time-dependent temperature distribution during cooling of cylindrical steel probes (25 mm dia. x 100 mm) for a.) slow non-Newtonian cooling and b.) explosive Newtonian cooling. The temperature distribution within the cylindrical specimen during quenching (indicated by isotherms) exhibiting non-newtonian wetting is shown in Figure 9a, where there is a great difference relative to that of the Newtonian wetting process shown in Figure 9b [64] [65]. Thus substantially greater residual stress distribution would be expected in the non-Newtonian cooling process indicated by Figure 9a. Hardness was measured using Rockwell hardness tests according to PN-78/H-04355 as well as Vickers hardness tests according to PN-78/H-04360 and PN-79/H-04361 using a load of 0.981 [N] (0.1[kG]) with a Zwick hardness Impact strength of hardened carbonitrided test specimens was " The scientific work funded from resources of Co mmittee of Scientific Investigations in years 2003 ÷2005 as investigative project" 00,20,40,60,811,21,4 Distance from surface; x[mm] The carbon and nitrogen content; %C and %N 16HG (16CrMn5) steel Carbonitriding 850°C/5h Fig. 10. The change of carbon and nitrogen content as a function of distance from surface in the diffusion layer produced in 16HG steel after carbonitriding at 850 °C and 5 h. The carbon and nitrogen potential of carbonitriding atmosphere was 0.9%C and 0,4%N. The amount of retained austenite and carbide structure is shown in Figure 11. These data show that at 1.15 – 1.25%C and 0.3 – 0.35 %N the greatest amounts of retained austenite (~48%) was formed in the hardened carbonitrided layer quenched in oil. The least amount of retained austenite was obtained the carbonitrided layers quenched in water and the 10% aqueous PAG polymer solution. The carbide (carbonitride) content in hardened carbonitrided layer was not dependent on the quenching medium and was about 4%. " The scientific work funded from resources of Co mmittee of Scientific Investigations in years 2003 ÷2005 as investigative project" The destilleted waterThe 10% water polymer solution The OH70 oil The kaind of medium quenching %Ra %Carbides 16HG (16CrMn5) steel Carbonitriding 850°C/5h Fig. 11. The influence of kind of cooling medium onto measured maximum 20 40 50 60 70 90 100 HV0.1 16HG (16CrMn5) steel Carbonitriding 850°C/5h Fig. 12. The influence of hardening medium on hardness HRc and HV0,1 of the hardened carbonitrided layers " The scientific work funded from resources of Co mmittee of Scientific Investigations in years 2003 ÷2005 as investigative project" analogous to that obtained for distilled water and OH70 oil. (Note BREOX Quenchant A would be expected to yield the same results.) The destilleted waterThe 10% water polymer solution The OH70 oil The kaind of medium quenching The abrasive wear resistance (Iz*10) [mg/cm *h]; The impact strength (KV15/5) [J]; The thickness of the hardened carbonitride layer (Tl*10) [mm] The abrasive wear resistance (Iz) The impact strength (KV15/5) The thickness of the hardened carbonitride layer (Tl) 16HG (16CrMn5) steel Carbonitriding 850°C/5h Fig. 13. The influence of quenching medium on abrasive wear resistance (Iz) and impact strength (KV15/5) as a function of the thickness of the " The scientific work funded from resources of Co mmittee of Scientific Investigations in years 2003 ÷2005 as investigative project" Although lower residual stresses for Newtonian cooling have been predicted [64], this data is the first to the author’s knowledge where the resulting performance losses due to increased residual stresses of non-Newtonian cooling relative to Newtonian cooling have experimentally shown. CONCLUSIONS 1. The ability to optimize martensite formation is dependent on cooling rate BOTH within the and at the surface obtainment in her of martensite structure. The ability to achieve the optimum cooling rates is difficult to achieve with both water and oil quenchants. 2. The quench severity of aqueous polymer quenchants based on PAG copolymers can be varied by varying the molecular weight and composition of the PAG copolymer, polymer concentration, bath temperature and agitation. 3. PAG based-aqueous polymer quenchants produced excellent microstructure " The scientific work funded from resources of Co mmittee of Scientific Investigations in years 2003 ÷2005 as investigative project" wytworzonych na stalach konstrukcyjnych; In ynieria Materia owa, nr 6 (119), listopad-grudzie 2000, s. 407÷410 [11] G.E.Toten, Y.Sun, G.M.Webster, L.M.Jarvis, C.E. Bates: Quenchants Selection; Heat Treating, 1998, s. 183 ÷ 191 [12] G.E.Toten, M.E.Dakins, J.M.Jarvis: How H-factors can be used to characterize polymers; J. Heat Treating, Dec. 1989, s. 28 ÷ 29 [13] M.Przy cka, W.G stwa, G.E.Totten and G.M.Webster, ; "Polymer Quenching Media Selection"; in Heat Treating-Proceedings of the 21st Conference, Eds. S.Shrivastava and F.Specht, 5-8 November 2001, ASM International, Materials Park, OH, p.201-206; [14] M.Przy cka, W.G stwa, G.E.Totten and G.M.Webster ; "Polymer Quenchants and diffusion Layer Properties"; in Heat Treating-Proceedings of the 21st Conference, Eds. S.Shri vastava and F.Specht, 5-8 November 2001, ASM International, Materials Park, OH, p.207-213; [15] Totten, G.E., Blackwood, R.R. and Jarvis, L.M., "Analyzing Polymer Quenchants During Use", Heat Treat. , 1991, March, p. 16-18. [16] Mues, E., Orszak, K.B., and Totten, G.E., "Selection and Use of Polymer " The scientific work funded from resources of Co mmittee of Scientific Investigations in years 2003 ÷2005 as investigative project" [24] Tensi H.M.: Labormethode zur Prüfung der Abschreckleistung von Polymerlösungen. Referat wyg oszony na posiedzeniu Mi dzynarodowego Komitetu Ch odziw Hartowniczych. Wiede 30 maja 1983. [25] Luty W.: Die Reproduzierbarkeit der Abschreckkurven bei der Untersuchung von Polymerlösungen und Ölen. Hätrerei Technische Mitteilungen 1983. Bd. 38. Nr. 6, s. 263-267. [26] Foreman RW (1984) Polymer Quenching Update. Industrial Heating, January, pp. 22-29. [27] Hilder NA (1987) The Behaviour of Polymer Quenchants. Heat Treatment " The scientific work funded from resources of Co mmittee of Scientific Investigations in years 2003 ÷2005 as investigative project" [39] Mansion J., Bournicon C., Sayettat C.: Les huiles de trempe. Senlis, France, Centre Technique des Industries Mecaniques, 1978. [40] Beck G. I in.: Planungs- und Auswalhprinzipien für ein Abschrecköl. Härterei Technische Mitteilungen 1975 Bd. 30. Nr. 6, s.346 – 358. [41] Furman G.R.: Quenching – Metallurgical Aspects. Lubrication 1971. Vol. 57. No. 2, s. 13 – 36. " The scientific work funded from resources of Co mmittee of Scientific Investigations in years 2003 ÷2005 as investigative project" (Tensi Method)”, November 10, 1999, American Society for Testing and Materials, Conshohocken, PA. [62] H.M. Tensi, G.E. Totten and G.M. Webster, "Proposal to Monitor Agitation of Production Quench Tanks", in 17th Heat Treating Society Conference Proceedings Including the 1st International Induction Heat Treating Symposium Technology,Properties, Performance and Testing , G.E. Totten, S.R. Westbrook, R.J. Shah Eds , 2003, ASTM International, West Conshocken, PA., p. 587-634. [64] H.M. Tensi, A. Stich, and G.E. Totten, "Fundamentals of Quenching", Eds. G.E. Totten, T. Inoue, and M.A.H. Howes 2002, ASM International, Materials Park, OH, p. 248-295.