Shear Coupling Grain Boundary Migration scripts

Misorientation maps

misorientation_maps.py computes minimum misorientation angles from two orientation maps of the same area and highlight area where grain boudaries have moved.

• Input: angles.txt contains the two sets of Euler angles phi_1a, phi_a,phi_2a,phi_1b, phi_b,phi_2b (flattened image)

• Outputs: rotation.txt: the misorientation axes and angles, color.txt: the associated grey levels according to the misorientation angles.

Frank Bilby coupling factors

beta-FB.py computes the coupling factor according to the Frank-Bilby equation for a given rotation axis for tilt grain boundary (cubic crystal).

Automatic image correction during creep

Use to provide automatic geometric transformation of AFM and EBSD maps before and after creep using fiducial markers (surface imperfections).

• A deep learning model (ResUNet network, Keras implementation) was trained to segment markers . res_u_net-markers.py: to train images and masks and res_u_net-markers-predict.py for inference.

• Markers centroids from segmentation maps were computed using watershed (OpenCV implementation) and geometric transformation was estimated between markers before and after creep. Need more than two markers (not the case for the images avant.bmp and apres.bmp): image_correction.py (input: two images before and after creep, output: the transformation matrix; the corrected image)

Digital Image Correlation

DIC.py helps in defining markers to follow at sample surface during grain boundary migration. Track these markers between two images by template matching (roi search). Input: two images and markers position in the first (array of (x,y,w,h) ROIs). Compute displacements and draw vectors at ROIs. Use openCV.

Contributions

Romain Gautier, Frédéric Mompiou