The Evolution And Optimization Mechanism Of Annealing-twinning Induced Grain Boundary Character Distribution Of 304 Stainless Steel

The effects of strain-annealing treatments and initial microstructures on the grain boundary character distribution(GBCD) in 304 stainless steel were studied by electron backscatter diffraction(EBSD),scanning electron microscopy(SEM) and optical microscopy(OM).The single-section trace analysis method based on EBSD was employed to distinguish statistically the incoherent and coherent∑3 boundaries in the samples as processed.The mechanism of GBCD optimization was clarified based on the in-situ observation on the GBCD evolution at the different stages of strain-annealing treatments.Finally,the corrosion characteristics of various grain boundaries were assessed and compared by intergranular corrosion testing.The processing varieties of cold-rolling and annealing have a significant influence on the fraction of special grain boundaries(f_SBs) and the GBCD.It is found that low strain(6～10%) followed by annealing can produce more∑3ⁿ(n=0,1,2,3...) grain boundaries(more than 70%) than for intermediate strain(20～50%) followed by annealing does.The trend is also true for the ratio of fraction summation of∑9+∑27 boundaries(f_{∑9+∑27}) to the fraction of∑3 boundaries (f_∑3),i.e f_{(∑9+∑27/∑3)}.The averaged f_{(∑9+∑27/∑3)} is 0.13 in the case of low strain(6%) followed by annealing and only 0.05 in the case of intermediate strain(20～50%) followed by annealing.As to the dependence of the f_SBs and f_{(∑9+∑27/∑3)} on the annealing time,it shows some variations for different grain boundary engineering(GBE) processes.In the GBE1 process which involves low strain(6%) and subsequent long-time(24～96 h) annealing at relatively low temperature(900℃),the f_SBs and f_{(∑9+∑27/∑3)} are increasing with the annealing time prolonged. Upon annealing for 96 h,their value hit the record of 80%and 0.2,respectively.However,in the GBE2 process which involves low strain(6%) and subsequent short-time annealing at high temperature(1050℃),the f_SBs and f_{(∑9+∑27/∑3)}were increasing firstly and then decreasing with the annealing time.The special grain boundary cluster with large Size(more than 200μm) appeared as the typical microstructure feature in the optimized GBCD.The larger cluster(nearly 1mm in dimension) has developed with connectivity of HABs network intorrupted substantially in the 304 stainless steel specimens treated by GBE1 method.In such cluster,densely-populated∑3ⁿ(n=1,2,3...) grain boundaries emerge.As the majority of∑3ⁿ grain boundaries,∑3 boundaries have been measured by the single-section trace analysis method based on EBSD.It is found that the major part of∑3 boundaries is incoherent.The two-step annealing treatment applied to the 304 stainless steel specimens previously treated by 6%cold-rolling is evidenced to be effective for GBCD evolution.The first annealing at relatively low temperature(900℃) for short time(1 h) contributes a lot to the final optimization of GBCD.However,multi-cycle treatment didn't show different effects on the GBCD apart from singe cycle treatment except it brings some change to the residual sub-structure population.The effects of carbide and grain size in the starting state on the GBCD are also investigated by SEM and EBSD technology.The results show that presence of carbide in the initial specimen plays a negative role in the GBCD optimization.When the initial random orientation distribution was obtained,it is suggested that the finer microstructure is better for GBCD optimization.Based on the in-situ microstructure observations,the GBCD evolution process is revealed and the mechanism of GBCD optimization clarified as well.The strain-induced grain boundary migration occurs during annealing in GBE1 treated 304 stainless steel.Accompanying with this process,the∑3ⁿ grain boundary clusters come into being,within which∑3ⁿ grain boundaries are multiplied significantly through the migration and interaction of incoherent∑3 grain boundaries. Meanwhile,the preferentially migrated grain boundaries can modify the neighboring grain boundaries and promote∑3ⁿ(n=0,1,2,3) and non-∑3ⁿ low∑CSL boundaries,which in return break the connectivity of the HABs network effectively.Further discussion points out that the formation of a large number of incoherent∑3 grain boundaries and strain-induced grain boundary migration might be the mechanism of GBCD optimization based on the twin-induced GBE.Finally,the corrosion behaviors of random and special grain boundaries are compared in a sensitized type 304 stainless steel previously treated by particular GBE process.The results show that∑1 and∑3(incoherent and coherent) grain boundaries exhibit better intergranular corrosion resistance.Meanwhile,nearly half of∑9 grain boundaries and a small number of random HABs possess good intergranular corrosion resistance,which indicates that the specialness of grain boundary is not only related to the misorientation of the neighboring grains but also related to the grain boundary plane location.The main achievements include:(1) The incoherent and coherent∑3 boundaries are distinguished statistically by the single-section trace analysis method based on EBSD.It is found that the major part of ∑3 boundaries is incoherent in the optimized GBCD.The high order∑3ⁿ boundaries are produced via the migration and interaction of incoherent∑3 boundaries.(2) Based on the in-situ microstructure observations,the evolution process of GBCD are elucidated in detail.It is pointed out that the formation of incoherent∑3 boundaries and the strain-induced preferential migration of particular grain boundaries are responsible for the GBCD optimization based on twin-induced GBE.