PhD student W/M

Thesis Subject: Microstructure and texture analysis of x-ray diffraction data using Machine Learning

You will join a collaboration between the ID03 and Algorithms & Scientific Data Analysis teams at the ESRF to study the microstructure and texture of engineering materials.

Upon plastic deformation, the crystal lattice accumulates defects such as dislocations. With increasing deformation these dislocations self-organize into 3D structures such as grain and subgrain boundaries. Eventually this process leads to work-hardening. Beamline ID03 uses Dark Field X-ray Microscopy (DFXM) to study this microstructure in-situ and in 3D.

You will join the Algorithms & Scientific Data Analysis (ADA) group and develop machine-learning techniques to identify and characterize large-scale dislocation structures such as low-angle grain boundaries. Your objectives will be to:

• Segment 3D orientation maps acquired with DFXM into cells and subgrains.
• Identify and classify the boundaries between these cells according to their orientation with respect to the neighbouring (sub)grains, orientation change across the boundary, cell size, etc.
• Set up and evaluate the performance of various machine learning models

The project will bring much-needed automation of the data analysis workflows for DFXM, enable non-expert users to make better use of their data, and make the techniques more accessible to an even larger scientific community, and especially industry. All work will be carried out in close collaboration with the ESRF Dark-Field X-ray Microscopy beamline ID03.

Further information may be obtained from Carsten Detlefs (tel.: +33 (0)4 76 88 2556, email: detlefs@esrf.fr or Vincent Favre-Nicolin (tel.: +33 (0)4 76 88 2027, email: favre@esrf.fr)