Although the austenitic stainless steels 304 and 321 are often treated nominally as equivalents in their hot rolling characteristics, the question remains whether any subtle differences between the two allow further optimisation of their respective hot rolling schedules. The hot workability of these two types of austenitic stainless steels was compared through single-hit Gleeble simulated thermomechanical processing between 800℃ and 􀀄􀀅00℃ while the strain rate was varied between 0.00􀀄s􀀈􀀉 and 5s􀀈􀀉. It was found that the constants for the hyperbolic sinh equation for hot working of AISI 321 steel are Q = 465 kJ/mol, 􀀖􀀗 = 􀀘.􀀙6 􀀚 􀀄0􀀉􀀛 􀀜􀀝􀀞􀀈􀀉􀀟􀀈􀀉, 􀀠 = 0.00􀀘 􀀜􀀝􀀞􀀈􀀉 and 􀀡 = 6.􀀄 while for 304 steel the constants are Q = 446 kJ/mol, 􀀖􀀗 = 􀀅.􀀄4 􀀚 􀀄0􀀉􀀛 􀀜􀀝􀀞􀀈􀀉􀀟􀀈􀀉, 􀀠 = 0.008 􀀜􀀝􀀞􀀈􀀉and 􀀡 = 6.􀀄. It is shown that the occurrence of dynamic recrystallisation starts when the Zener-Hollomon parameter 􀀢 􀀣 6.4 􀀚 􀀄0􀀉􀀛s􀀈􀀉 for both steels but that the differences in the values of Q and A3 (the structure factor) between the two steels does lead to consistently lower steady state stresses for the steel 321 than is found in the steel 304 at the same Z values. This may, therefore, offer some scope for further optimisation of the hot rolling schedules and in particular in the mill loads of these two respective steels. A modelled constitutive equation derived from hot working tests to predict hot rolling mill loads is proposed and validated against industrial hot rolling data for AISI 321 stainless steel. Good correlation is found between the predicted Mean Flow Stress, the Zener-Hollomon Z parameter and actual industrial mill load values from mill logs if allowances are made for differences in Von Mises plane strain conversion, friction and front or back end tension. The multipass hot working behaviour of this steel was simulated through Gleeble thermomechanical compression testing with the deformation temperature varying between 1200℃ down to 800℃ and the strain rate between 0.001s-1 and 5s-1. At strain rates greater than 0.05s-1, dynamic recovery as a softening mechanism was dominant, increasing the dynamic recrystallisation to dynamic recovery transition temperature DRTT to higher temperatures. This implies that through extrapolation to typical industrial strain rates of about 60s-1,most likely no dynamic recrystallisation in plant hot rolling occurs in this steel but only dynamic recovery. Grain refinement by DRX is, therefore, unlikely in this steel under plant hot rolling conditions. Finally, mill load modelling using the hot working constitutive constants of the near-equivalent 304 instead of those specifically determined for 321, introduces measurable differences in the predicted mill loads. The use of alloy-specific hot working constants even for near-equivalent steels is, therefore, recommended. / Thesis (PhD)--University of Pretoria, 2014. / lk2014 / Materials Science and Metallurgical Engineering / PhD / unrestricted

The current study focus on investigating a relationship between processing variables and microstructure evolution mechanism in 304L stainless steel subjected to hot torsion. The Gleeble 3800 with Mobile Torsion Unit (MTU) is utilized in the current study to conduct hot torsion test of 304L stainless steel. Samples are rotated at 1100℃ in the shear strain rate range of 0.02s-1 to 4.70s-1 and the shear strain range of 0.5 to 4. Orientation imaging microscopy (OIM) technique is used to collect and analyze the microstructure. At low strains (≤1) and strain rate (0.02s-1), average grain size remains relatively constant, but the lengths of DSs and LABs increase within grains. These are characteristics of the dynamic recovery (DRV). With increasing strain and strain rate, the lengths of DSs, LABs and HABs increase, accompanied by the decrease of average grain size. Subgrains with HAB segments are observed. These are characteristics of continuous dynamic recrystallization (CDRX). At strain rates greater than or equal to 0.94s-1, the fraction of deformation texture is about 3 times higher than that of rotated cube texture. The average grain size increases relative to that at a strain rate of 0.20s-1, accompanied by the increase of twin length per area. This indicates that grain growth take place after CDRX. Sigma phase is not observed in the current study due to the lack of static recrystallization (SRX) and the higher cooling rate.