L2M – Accelerated Data-Driven Approach to Characterize Mechanical Properties of Marine Materials

Ocean industries, such as shipbuilding and offshore energy, face challenges in selecting materials that can endure corrosive environments, high pressures, and mechanical stresses. While materials like composites and aluminum offer promise, their adoption is slowed by their inherently complex and nonlinear mechanical behavior, which is sensitive to environmental conditions, as well as by the time-consuming, costly testing methods required to understand them. This project provides an innovative solution using data-driven algorithms and high-throughput characterization techniques to accelerate predictions of the mechanical performance of marine materials. The project proposes two solutions: first, leveraging historical data and open-source databases to develop machine learning algorithms for accelerated predictions in new scenarios without the need for further experiments, particularly when time is critical. Second, it integrates a real-time, non-destructive, high-throughput material characterization technique with the manufacturing stage to generate reliable data for ML model training. This approach enables accurate, accelerated property predictions under consistent conditions, significantly reducing the need for experimentation and providing environmental benefits. By speeding up material testing and lowering costs, this project aligns with the internship’s goal of promoting innovative, sustainable approaches in marine industries.

Faculty Supervisor:

Reza Rizvi

Student:

Partner:

Springboard Atlantic Inc.

Discipline:

Engineering

Sector:

Water; Artificial Intelligence; Sustainability & the Environment

University:

York University

Program: