About Us

Education

Research

Collaborate

People

Magazine

Contacts
IT

DEEP LEARNING-DRIVEN PREDICTION AND DESIGN FOR AEROSPACE MATERIALS AND STRUCTURES

25 January 2024 — 1 minutes read

PhDAER Seminar
Thursday, January 25, 2024 at 12 pm - Dipartimento di Scienze e Tecnologie Aerospaziali, Politecnico di Milano, Sala Consiglio DAER, Second Floor, Campus Bovisa, Via La Masa, 34, Milano (MI)

Over the last decade, applied Artificial Intelligence (AI) research tackling engineering problems has exponentially increased. Reasonable success has been observed in applications ranging from AI-assisted materials design to intelligent data-driven condition (damage) monitoring of structures and engineering components. One aspect that is often overlooked in developing AI-driven tools for engineering applications is the reliability of such AI tools which manifests in terms of their physical correctness, generalisation capability and interpretability. In this context, the talk will focus on interpretable, physics-guided AI frameworks using deep learning algorithms for prediction of mechanical behaviour of heterogenous material systems. Two examples are considered, one on AI-driven prediction and generative design of fiber composite materials and another on deep learning-based crystal plasticity constitutive models for accelerated FE simulations.

The talk is part of the UK-Italy Trustworthy AI Researcher Award from the Alan Turing Institute, the UK’s National Institute for Data Science and AI. The purpose of the award is to initiate and develop long-term collaborative research between the UK and Italy on trustworthy AI development and their applications in aerospace engineering.

Speaker

Dr Sathiskumar Anusuya Ponnusami is a Senior Lecturer (Associate Professor) in Aerospace Structural Mechanics within the Department of Engineering at City, University of London, UK.

His research interests encompass computational modelling and design of aerospace materials and structures with an emphasis on machine learning tools for mechanics.

He serves as an Associate Editor of the Mechanics of Materials journal (Elsevier) and as the Data Study Group PI at the Alan Turing Institute.

He conducted his postdoctoral fellowship in the Solid Mechanics Group at the University of Oxford, UK within the Rolls-Royce University Technology Centre and his PhD at TU Delft, Netherlands.

During his career, he was awarded several prizes including Royal Society’s Newton International Fellowship Grant, Airbus-UNESCO Fly Your Ideas winner 2015 and NLF Dutch Aerospace Award 2015.

His ongoing research is supported by UKRI-EPSRC, IMechE and the Royal Society.

His research outputs include 40 journal articles, 35 international conferences, 7 book chapters and a patent.

22.01.2024