Study of the 3D microstructure of cementitious and alternative binders using image analysis
The objective of the project is to perform quantitative image analysis of the microstructure of cementitious binders and by alkaline activation. Sets of images of the materials were collected at the synchrotron in the static state and during in situ compression tests. The student will need to learn how to process a large volume of images to calculate material properties and to study figure propagation to compare materials. Computational approaches can be compared (physical approaches, supervised and non-supervised neural networks, and deep learning). Ambitious students will be able to try generative algorithms to augment the database. Programming code must be written to be shared. The results from the image analysis must contribute to the discussion on the durability tests.
Required knowledge
- Programming skills in Python, C++, Linux interface, and other software
- Ability to write a scientific article in English
- Ability to write according to curriculum rules
- Basic planning and time management
- Coordination with other students and experts