Predicting failure in crystalline materials using machine learning techniques

Single crystal materials are extensively used in components ranging from the smallest (e.g. micro-lens, MEMS devices) to the largest sizes (e.g. turbine blades). Learn more at Predicting failure in crystalline materials using machine learning techniques from the Loughborough University

Key facts of the  Predicting failure in crystalline materials using machine learning techniques Predicting failure in crystalline materials using machine learning techniques programme offered by  Loughborough University Loughborough University

What is imperative in such components is to ensure safety and reliability of performance during the lifetime of the product. This implies that engineers would need a robust, accurate and fast numerical model which is able to assess strain accumulation in parts when subjected to complex loading states. 

These ‘hotspots’ are a precursor to failure. Knowledge of its location and state will allow for precautionary measures to be incorporated during maintenance and inspection.

The project will deal with integrating crystal plasticity models and machine learning techniques to gain insight into the microstructural effects of single-crystal metals. Robust computational models already exist. These will be enhanced to ‘hand-shake’ with a machine learning framework. Necessary small-scale experiments will be conducted in a nano-micro indentation testing machine available.

Once developed the computational framework will be used to predict failure in complex geometries under complex loading states which are of industrial relevance.

• To obtain additional information about the program, we kindly suggest that you visit the programme website, where you can find further details and relevant resources.