A STUDY ON SMALL SCALE MECHANICAL BEHAVIOR FOR DEFORMATION ANALYSIS OF DUPLEX STAINLESS STEEL

Abstract

Stainless steels are an important class of alloys and the use of stainless steels is indispensable. Above all, duplex stainless steels (DSS) are generally known as material combining good strength with corrosion properties. However, mechanical behavior of DSS is complex and inhomogeneous, since respective phases in DSS have different response to applied stress or strain. Therefore, in recent years, great effort has been made to understand complex deformation behavior of DSS. There are three conventional deformation analysis methods for DSS

(1) neutron diffraction analysis, (2) EBSD (texture) base analysis and (3) digital image base analysis. In this paper, these deformation analysis methods for DSS are reviewed and some demerits of the methods are reported. Then advanced analysis methods are developed and suggested in order to analyze complex deformation behavior of DSS, precisely. Neutron diffraction is one of the conventional methods to analyze deformation behavior of DSS. Especially, the plastic yielding behavior of each phase in DSS was often focused by time-of-flight (TOF) analysis using in-situ neutron diffraction. However, it is difficult to separate the individual contribution of each phase from macroscopic applied stress or strain. And applied stress cant be measured precisely by this method since there is a limitation in the step for the static load control during in-situ neutron diffraction. Besides, neutron diffraction has bad accessibility because the size of equipment is quite huge and there are few places that can use it. For these reasons, as more accessible technique, nanoindentation was suggested to measure plastic deformation behavior of each phase in DSS. Nanoindentation was chosen since separated intrinsic mechanical behavior of each phase can be measured, and applied stress can be obtained as continuous value by this technique. In this research, two specimens of Fe-24.67Cr-7.04Mn-3.98Ni-3.88Mo-0.49Si-0.45N-0.022C (DSH) and Fe-17.2Cr-5.9Mn-5.01W-2.54Mo-0.31Si-0.43N-0.012C (DSL) were used. Nanoindentation tests were performed to measure intrinsic deformation behavior of individual α and γ grains guided by electron backscattered diffraction (EBSD). Generally, maximum shear stress underneath indenter tip when first pop-in occurs is considered as elasto-plastic transition stress by dislocation nucleation. Thus maximum shear stress when first pop-in occurs (τm) is measured to define elasto-plastic transition stress of each phase in DSH and DSL. In order to correlate the small-scale nanoindentation behavior to the macro-scale tensile behavior, an angular-dispersive in-situ neutron diffraction test was performed using a residual stress analysis diffractometer equipped with a deformation device enabling tensile deformation. And the correlation was investigated by comparing the indentation load-depth (L-D) curves for each phase with the lattice strains of various lattice planes in α and γ obtained by in-situ neutron diffraction. In nanoindentation, γ had approximately 20% lower elasto-plastic transition stress than α in DSH, while both α and γ had similar elasto-plastic transition stress in DSL. And this tendency of elasto-plastic transition corresponded correctly to in-situ neutron diffraction data. Thus we concluded that tendency of maximum shear stress when first pop-in occurs can represent that of yield stress. Furthermore, the dislocation nucleation energy (DNE) was calculated based on nanoindentation results, since it is generally known that the pop-in is closely related with dislocation nucleation. As expected, these results correctly correspond with results from nanoindentation. Finally we could conclude that the difference of elasto-plastic transition stresses in both DSS can be explained by DNE and the SFE in austenite could be evaluated by nanoindentation based DNE calculation. And secondly, EBSD supported DIC (digital image correlation) analysis was developed to improve both texture and digital image base deformation analysis methods for DSS. In texture base deformation analysis for DSS, the phase which has intensely distorted texture regards as severely deformed phase. However, correlation between strain and misorientation is not investigated clearly. Thus it is ambiguous to define deformation by texture information and, of course, accurate strain cannot be measured by EBSD base deformation analysis. On the contrary, precise quantitative strain can be directly measured by digital image base deformation analysis method called DIC. But with this method, phase/grain boundaries cannot be identified clearly because they are observed by etched surface morphology. However, in order to analyze complex deformation of DSS, accurate quantitative local strain and precise phase/grain information are both necessary. Thus EBSD supported DIC analysis method was developed in order to measure accurate quantitative strain with regarding precise microstructure and texture. In addition, relation between local strain and misorientation was investigated to verify texture base deformation analysis. In order to develop EBSD supported DIC technique, full annealed pure copper sheet (purity 99.9%) was used in this research. For DIC analysis, grayscale random pattern is required on the surface of specimen. Speckle patterning method has been conventionally used for this, but it is shown that speckle patterning is inappropriate for EBSD since the speckles disturb contact of electron beam. In order to obtain well indexed precise EBSD image with grayscale patterned surface image for DIC, carbon coat patterning technique was developed and applied. After carbon coat patterning, selected area of Cu specimen was captured while tensile test by in-situ SEM equipped with a deformation module. The area was also scanned by EBSD. The sequential SEM images during deformation were analyzed by DIC and quantitative strain maps were drawn. Consequently, accurate quantitative strain, precise microstructure and texture of the selected area could be measured together. Then in order to investigate relation between local strain and misorientation, Kernel average misorientation (KAM) map and grain average misorientation (GAM) map were drawn and directly compared with local strain. As generally known, overall average state of misorientation correlated with macroscopic applied strain, statistically. However, it was shown that locally evaluated KAM/GAM values have quite weak correlation with applied local strain of the spot. One of the reasons for this tendency, we found that crystal distortion and increasing misorientation by deformation are influenced by orientational stability for deformation mode. That is, locally increased KAM/GAM values can be varied by initial orientation of observed area, not only applied strain. Thus we conclude that orientational stability for applied deformation mode should be regarded for reliable strain analysis by EBSD.