The influence of process parameters on the hot deformation behaviour of 2304 LD stainless steel

Abstract

A series of single compression tests were performed in order to improve the understanding of the influence of processing parameters (temperature, strain and strain rate) on the hot deformation behaviour of 2304 LDSS. The hot compression tests were carried out in the temperature range of 850 � 1050 oC with strain rates of 0.1, 1, 5, 10 and 15 s-1, and true strain of 0.6. The strain rates were also extended to 30 and 50 s-1 and true strain to 0.8 in order to broaden the knowledge base, and to observe if effects such as flow stress steady state can be achieved. Ferrite-to-austenite transformation was observed to take place at higher strain rates with an increase in strain beyond the peak strain; especially at 10 s-1. At lower strain rates of 0.1 and 1 s-1, no increase in the austenite phase fraction was observed. The increase in austenite phase fraction was observed to take place simultaneously with softening in the flow stress. It was concluded that the phase change was not the cause of the observed flow softening, since flow softening was still observed even in the absence of a phase change (an increase austenite fraction). EBSD results confirmed the flow softening to be due to DRX. The ferrite-to-austenite transformation is suggested to be strain-induced. This strain-induced transformation is based on the fact that the deformation of a dual austenite-ferrite structure results in unequal strain distribution over the two phases and ferrite bearing more strain since it is the softer phase. Upon deformation beyond the peak strain where softening processes are active in both phases; the two phases undergo different softening kinetics due to their difference in SFE. The austenite phase undergoes DRX due to its low SFE and ferrite basically undergoes DRV due to a high SFE (with possibility of CDRX at high strains). This leads to more deformation energy being stored in the ferrite, and hence the driving force for the nearby austenite strain induced boundary migration (SIBM) into the ferrite in order to minimize the system energy and achieve equilibrium condition. The increase in austenite fraction in turn leads to a shift in the equilibrium phase fractions of ferrite and austenite. Microstructural analysis and processing maps revealed DRV as the operating power dissipation process below the peak strain, ensuring good workability. The flow behaviour in work hardening and dynamic recovery regime was successfully modelled through a physically-based Estrin-Mecking (EM) model. Likewise, the fractional softening regime was also modelled through the Avrami equation. Consequently, an EM model coupled with the Avrami equation was shown to accurately predict the flow behaviour of the 2304 LDSS. A model based on the change in Gibbs free energy resulting in the observed change in phase equilibrium was proposed. The model was developed through incorporating the results from microstructural and EBSD analysis to substantiate the observed dynamic transformation. The model seems to concur with the observation in this study that the ferrite-to-austenite (? ??) dynamic transformation (DT) taking place was not responsible for the flow softening. Rather it was the DRX in austenite and DRV in ferrite resulting in dislocation density differences which resulted in SIBM that is responsible for the observed ? ?? DT. No phase fraction change was observed below the peak strain, i.e. up to a strain of 0.3. The significant observation applicable to typical industrial steel processing is whereby the strain per pass is generally below 0.3. Hence, no phase fraction changes would be expected at any stage during a typical industrial finish rolling which would lead to better control of the final microstructures and the subsequent mechanical properties.