EFFECTS OF COOLING RATE ON THE VOLUME FRACTION OF RETAINED AUSTENITE IN FORGINGS FROM HIGH-STRENGTH Mn-Si STEELS

Various ways are sought today to increase mechanical properties of steels while maintaining their good strength and ductility. Besides effective alloying strategies, one method involves preserving a certain amount of retained austenite in a martensitic matrix. The steel which was chosen as an experimental material for this investigation contained 2.5% manganese, 2.09% silicon and 1.34% chromium, with additions of nickel and molybdenum. An actual closed-die forged part was made of this steel. This forged part was fitted with thermocouples attached to its surface and placed in its interior and then treated using the Q&P process. Q&P process is characterized by rapid cooling from a soaking temperature to a quenching temperature, which is between the Ms and the Mf, and subsequent reheating to and holding at a partitioning temperature where retained austenite becomes stable. The quenchant was hot water. Cooling took place in a furnace. Heat treatment profiles were constructed from the thermocouple data and the process was then replicated in a thermomechanical simulator. The specimens obtained in this manner were examined using metallographic techniques. The effects of cooling rate on mechanical properties and the amount of retained austenite were assessed. The resultant ultimate strength was around 2100 MPa. Elongation and the amount of retained austenite were 15% and 17%, respectively. Microstructures and mechanical properties of the specimens were then compared to the real-world forged part in order to establish whether physical simulation could be employed for laboratory-based optimization of heat treatment of forgings.

A special composition was designed for a 0.42 % carbon steel to depress the Ms and Mf, using iterative optimization in the JMatPro program. The Mf was below 100°C, thanks to which boiling water could be used for quenching (Tab. 1). The reduction in Ms and Mf was due mainly to a higher manganese level, i.e. 2.5 % [10]. Other alloying elements included silicon, chromium, and molybdenum [11]. The purpose of silicon was to prevent carbides from forming, to facilitate the supersaturation of martensite with carbon and to provide solid solution strengthening. Chromium improves hardenabil-ity and strengthens the solid solution.
Molybdenum was added with a view to depress the Ms and Mf and improve the stability of austenite. Nickel was added in a small amount. It makes austenite more stable during cooling, improves hardenability and provides solid solution strengthening. Niobium belongs to the most common microalloying elements. Even a minute amount substantially alters mechanical properties of steel [12,13]. Niobium has a strong affinity for carbon and nitrogen. It combines with them to form carbonitrides which dictate mechanical properties of the material [14].

Data acquisition and development of physical simulation regimes
First, a closed-die forged part was made of the experimental steel (Fig. 1). The data for developing physical simula-tion regimes to be conducted in a thermomechanical simulator were gathered in the course of heat treatment of the forged part [15][16][17][18][19][20]. For this purpose, the part was fitted with thermocouples, some attached to its surface (the fastest-cooling part of the forging) and others placed in its interior (the slowest-cooling location). Specifically, one thermocouple was attached to the surface (no. 1) and two thermocouples were placed in the part's interior (no. 2 and 3), (Fig. 1). The forged part was then Q&P processed. It was heated in an air furnace at 880°C to a fully-austenitic condition. Since the special alloying of the steel de-pressed the M f to 78°C in Table 1, it was possible to use boiling water at 100°C as a quenchant. Boiling water makes a better quenchant than oil or salt baths in terms of safety, the bath quality and degradation, as well as environmental aspects. Once the surface temperature reached approx.
100°C, the part was removed from water and transferred for partitioning for 1 hour in a furnace at 200°C (Fig. 2). The thermocouple data from point 1 on the surface indicated that the quenching temperature in that location was 100°C. At points 2 and 3, approx. 10 mm below the surface, the quenching temperature was higher, about 230°C. Several different cooling profiles were thus obtained for several locations across the forged part. The part was then heat-treated again, this time using a different regime. The austenitizing temperature was identical but the cooling step took place in air, until the surface temperature reached 240°C. The purpose was to explore the impact of the cooling rate on the final amount of retained austenite. Austenitizing was followed by partitioning in a furnace at 200°C (Fig. 3). In this case, the differences between the measured locations were less distinct than after the water-quenching regime.
Both regimes provided real-world data for developing regimes for a thermomechanical simulator. Chemnitz Univer-sity of Technology, which collaborated on this investigation, carried out numerical modelling of the heat treatment using FE software Simufact.forming 14.0.
Processes with several cooling rates were modelled [6]. Under laboratory conditions, the impact of changes in processing parameters, such as cooling rate, quenching tem-perature and partitioning temperature on the microstructure and mechanical properties can be determined.  Retained austenite in advanced high-strength martensitic steels contributes to their toughness. In order to stabilize retained austenite by Q&P processing, the right quenching temperature must be used along with an appropriate cooling rate. In these experiments, four regimes involving different cooling rates were performed on specimens of the experi-mental steel. The data for designing these regimes were those obtained from heat treatment of the closed-die forged part.
The data consisted of cooling curves for quenching in boiling water and for slow cooling in air of the surface and the interior of the forged part (Fig. 4).
The first regime was a simulation of quenching of location 1 on the surface of the forged part in The first regime was a simulation of quenching of location 1 on the surface of the forged part in boiling water. In this regime, cooling from the soaking temperature to a quenching temperature   (Fig. 1). The microstructures and properties of the specimens were then examined and measured using light and scanning electron microscopes and mechanical testing machines, respectively. The amount of retained austenite was determined using X-ray diffraction in Table 2.

Results and discussion
The specimens processed in the thermomechanical simulator were examined using a scanning electron microscope (Tescan Vega 3). All the microstructures consisted of a majority of martensite, a small amount of bainite and various volume fractions of retained austenite (Fig. 5a -7a).   (Fig. 5b). Retained austenite was present as globular grains and as particles between martensite needles. Regime 2 involved slower cooling, at 5.7°C/s. It was a simulation of the forged part's interior during quenching in boiling water.
After this regime, the ultimate strength was about 100 MPa higher than in the previous case. As the amount of martensite was larger than in the previous case, hardness was higher by approximately 30 HV10. Elongation dropped substantially to 8%. This can be explained by a lower fraction of RA in Table 2 and by coarser grains. The reduction in the amount of retained austenite (10%) was confirmed by special etching and metallographic observation. There was less RA than after regime 1. Most of it was in the form of globular grains. Some was found between martensitic needles. Two subsequent regimes were similar to each other. They were simulations of air cooling of the surface and interior of the forged part. Their cooling rates were low: 3.5°C/s and 2.9°C/s, respectively. As a consequence, the resulting elongations were even lower than in the previous caseas low as 3%. Related to this was the lower amount of retained austenite: a mere 8%. A larger amount of martensite led to a higher hardness: approx. 690 HV 10 in Table 2, (Fig. 6a). A small amount of retained austenite was in a globular form (Fig. 6b).
Mechanical properties of the specimens more or less corresponded to those of the real forged part. The largest amount of retained austenite, 10%, was found in the surface of the forged part upon quenching in boiling water. Re-tained austenite was present as globular grains and as particles between martensite needles (Fig. 7b). The ultimate strength of the forged part, 2131 MPa, and its elongation of 12% are nearly identical to those of the physical simulation specimens in Table 2.