Skip to main content

Machine Learning for understanding strain pattern development during deformation of metals

Primary supervisor

Yasmeen George

Co-supervisors

  • Michael Preuss
  • Juan Nunez-Iglesias

For most engineering applications we use metals and alloys (mixture of metals) for components that need to carry significant loads. These materials have an elastic limit beyond which they start to deform irreversibly. Such irreversible deformation, called plasticity, is generally of very discrete nature and the development of such discrete strain patterns, particularly during the early stage of plasticity, is very poorly understood. Such discrete plasticity can lead to pre-mature failure of such materials and therefore it is of great importance to develop a better understanding of the discrete strain pattern development. Novel imaging techniques in connection with using digital image correlation techniques enable us now to record the development of strain patterns, see Figure 1, but the evolution of such strain patterns is highly complex and difficult to understand from a materials point of view. Specific questions are related to the time-resolved evolution and the spacing between discrete slip traces and how they extend across the material.

Figure. 1: Shear strain maps recorded during in-situ tensile loading experiment. The white boundaries highlight the boundaries between individual crystals that make up the material
Figure. 1: Shear strain maps recorded during in-situ tensile loading experiment. The white boundaries highlight the boundaries between individual crystals that make up the material.

Student cohort

Double Semester

Aim/outline

Experimental data are available from current research activities at the University of Manchester that bring together maps of orientations of such crystals with the deformation pattern generated during mechanical loading of such samples

In this project, we will use machine learning models and advanced image analysis techniques to

  1. Understand the materials deformation process (discrete strain pattern development) by analysing time-series (sequential) images data
  2. Understand the interplay between the crystal orientations and deformation patterns in a polycrystalline material during plastic deformation
  3. Predict materials plasticity deformation and failure

Required knowledge

Python programming, data/image analytics, data wrangling, signal processing, machine/deep learning