Inverse computational design of friction

This project will focus on developing a inverse computational design pipeline for fabricating 3D surfaces that meet specific requirements. This pipeline will leverage novel friction models being developed by other students and researchers in the group. The student will explore generative models for synthesizing micro-geometry of 3D surfaces that meet user-defined specifications. For instance, aggregate properties of the surface such as roughness, microfacet distributions, and friction. A simulation platform will be used to optimize for surface micro-geometry such that specific frictional behavior is realized. An consideration will be to investigate differentiable simulations, such that changes in frictional behaviour based on the input parameter space of the friction model have a differential. Fabrication of the surfaces resulting from this computational design framework will be undertaken in collaboration with junior students in the group. 
 

Required knowledge

Knowledge of video games, physics simulation, and/or 3D graphics is beneficial.
The project will also build on concepts in machine learning, such as supervised learning generative models. Experience with numerical optimization is also helpful. All candidates should have strong programming skills, particularly in C++ or Python.
 

Desired program of studies

Doctorate

Research domains

Innovative Materials and Advanced Manufacturing, Software Systems, Multimedia and Cybersecurity

Financing

Accepted candidates will be funded through a stipend provided. Funding is competitive with other Quebec universities.

Additional information

Keywords: geometry, generative models, 3D printing, computational design, physics-based animation, friction

Starting : Autumn 2022 / Winter 2023