| Electron backscatter diffraction(EBSD)can be used to characterize crystallographic information at microscopic scale in materials,which plays an important role in material designing.Combined with the integrated digital image correlation algorithm,an advanced image processing algorithm,high resolution(HR-)EBSD can measure precisely the relative strain and stress field to the reference point on the surface of materials by tracking small shifts and rotations in diffracted patterns.During calculation,calibration of pattern center(PC)and selection of reference patterns are two major challenges,and will greatly affect the precision.Firstly in this thesis,criteria such as image residuals and similitude between experimental patterns and virtual patterns were used to reduce the errors in PC values,which eventually leads to a reduction in phantom stresses.Meanwhile,in this paper,locations of all scanned points were determined through the calibration method based on virtual patterns,and the scanned surface was hereby reconstructed in three-dimensional space.In this case,the impact of errors in the experimental setup on the calculation was avoided.Besides,during strain and stress measurement,deformation in experimental reference patterns can induce deviations of the calculation.To solve this problem,this research proposed to optimize the deformation gradient tensor in the reference pattern to make the calculated stress field approach the plane stress hypothesis.Through this method,the values of two out-of-plane stress components decreased,the dependence on the selection of reference patterns was weakened and the continuity of the stress field across grain boundaries was improved.Based on the methods proposed above,including 2 PC calibration methods and a method for absolute stress measurement,this thesis improved HR-EBSD technology and increased the precision of local strain and stress measurement. |