| Grain boundary,as an important part of polycrystalline materials,exerts a significant influence on the materials performance,such as intergranular corrosion,precipitation,fracture and creep,which are closely related to the grain boundary structure.Many studies have shown that the grain boundary properties of materials can be enhanced through suitable thermomechanical processing to control grain boundary character distribution(GBCD),especially increase the fraction of low-? coincidence site lattice(CSL)boundaries and disrupt the connectivity of random boundary network.Until now,researches on the effect of thermomechanical processing parameters on GBCD mainly focus on strain level,annealing temperature,annealing time and number of process passes.Grain size,as an important microstructural feature in polycrystalline materials,has a crucial impact on grain growth behavior during annealing.Similarly,different strain states and stress levels in the material introduced by different deformation modes and paths also have remarkable effects on grain growth behavior during annealing.All of these(grain size,deformation mode and deformation path)will inevitably affect the evolution of GBCD.Thus,in this study,the effects of initial grain size,deformation mode and deformation path on the evolution of GBCD in 304 austenitic stainless steel were systematically and deeply investigated.Meanwhile,thermomechanical processing parameters were optimized by the combination of artificial neural network and genetic algorithm.The main research work and results are as follows:(1)Effect of initial grain size on GBCD in 304 austenitic stainless steel was studied without deformation and with deformation,respectively.Without deformation,abnormal grain growth occurred and the fraction of low-?CSL boundaries increased in specimens with small grain sizes(10.6?m and 34.9?m)after annealing.However,too small grain size(10.6?m)was bad for achieving a discontinuous random boundary network.During annealing,normal grain growth occurred in the specimen with large grain size(48.7?m),in which the fraction of low-?CSL boundaries changed insignificantly and the connectivity of random boundary network was not interrupted.With deformation,the fractions of low-?CSL boundaries in specimens with three kinds of grain sizes increased after annealing.During annealing,normal grain growth occurred in the specimen with smaller grain size(10.6?m),while abnormal grain growth occurred in specimens with large grain sizes(34.9?m and 48.7?m).Compared to normal grain growth,abnormal grain growth was more conducive to forming low-?CSL boundaries and disrupting the connectivity of random boundary network.(2)Under the same equivalent strain,effects of deformation mode and strain level on GBCD in 304 austenitic stainless steel were studied by using rolling deformation and tensile deformation.Regardless of deformation mode,the fraction of low-?CSL boundaries decreased with the increase of strain level after annealing.At the equivalent strain of 0.06,strain induced grain boundary migration occurred and the fraction of low-?CSL boundaries increased significantly during annealing.However,the rolling deformation was more beneficial to increase the fraction of low-?SL boundaries,especially ?3n(n=1,2,3)boundaries and disrupt the continuous random boundary network,which was due to the formation of texture along the direction<110>parallel to the normal direction(ND)orientation.For the equivalent strain of 0.12 or 0.26,full recrystallization occurred during annealing.In addition,compared to the rolling deformation,the tensile deformation can optimize the GBCD more effectively as specimen deformed by rolling has a larger stored energy,which was detrimental to the formation of low-?CSL boundaries.(3)Based on the same equivalent strain,effects of deformation path and strain level on the GBCD in 304 austenitic stainless steel were studied through the unidirectional tensile deformation and the cross tensile deformation.The result indicated that the fraction of low-?CSL boundaries decreased with increasing strain level after annealing.But the cross tensile deformation was more favorable to form low-ECSL boundaries and disconnect the random boundary network as compared to the unidirectional tensile deformation for the same equivalent strain and annealing process,which resulted from the differences in stored energy and texture.For the equivalent strain of 0.07 or 0.15,the cross tensioned specimen has a lower stored energy,which facilitated the formation of low-?CSL boundaries,especially ?3n(n=1,2,3)boundaries.For the equivalent strain of 0.24,stored energy in the cross tensioned specimen was similar to that found in the unidirectional tensioned specimen.However,the formation of the{110}<111>P texture in the cross tensioned specimen promoted the formation of low-?CSL boundaries during annealing.(4)Thermomechanical processing parameters were optimized through the combination of artificial neural network and genetic algorithm,and the optimal thermomechanical processing parameters under rolling deformation in 304 austenitic stainless steel can be achieved when strain level,annealing temperature and annealing time were 5%,1022? and 23min,respectively.The predicted fraction of low-?CSL boundaries through the optimal thermomechanical processing parameters was 78.8%,which coincided well with the experimental result(77.9%).(5)Effect of GBCD on intergranular corrosion was studied by the ferric sulfate-sulfuric acid test.The result showed that the corrosion rate in the BM was much higher than that of the grain boundary engineered material(GBEM)during intergranular corrosion testing,and the corrosion rate of the GBEM was only 50%of the BM after immersed in the boiled ferric sulfate-sulfuric acid solution for 96h.Almost all grasurface dropped seriously in the BM,while grain boundaries in the GBEM were corroded in boundaries were attacked and grains on the slightly and the surface of the GBEM was still integrated.The formation of large size ?3n(n=1,2,3)boundaries clusters and the connectivity of random boundaries interrupted by low-?CSL boundaries were the main reasons for the significant improvement of intergranular corrosion resistance. |
PDF Full Text Request