Date of Award

2026

Document Type

Open Access Master's Thesis

Degree Name

Master of Science in Mechanical Engineering (MS)

Administrative Home Department

Department of Mechanical and Aerospace Engineering

Advisor 1

Vinh Nguyen

Committee Member 1

Geordan Gutow

Committee Member 2

Cameron Hadden

Abstract

Composite materials have become a critical component of modern manufacturing, especially in the automotive and aerospace industries. The curing process for these composites has been modeled using a variety of partial differential equations representing the heat transfer and composite curing kinetics. Optimizing the applied temperature profile is critical for maximizing the efficiency and capacity of composite part manufacturers. Constraints must be placed on the inputs and outputs of the model, including but not limited to, the applied temperature profile, part temperature, and final degree of cure. Conflicting sets of constraints are easy to unknowingly impose due to the highly coupled and non-linear partial differential equations characterizing the optimization problem. This work presents an approach to managing conflicting constraints in the optimization process through a new method of co-training Physics Informed Neural Network (PINN) models. The proposed training algorithms, Loss Categorization and Trenching Algorithm (LoCaTA) and Automatic Loss-Based Constraint Relaxation (ALBaCoRe) allow for the detection of conflicting constraint sets through the categorization and modification of traditional loss terms. If conflicts are detected, ALBaCoRe will automatically adjust the constraints to a non-conflicting state. Both algorithms are tested using sets of constraints with known degrees of conflict and compared to finite element (FE) models to evaluate accuracy. LoCaTA was shown to accurately distinguish between conflicting and non-conflicting sets. For the conflicting sets, ALBaCoRe was able to relax conflicts to a new set of non-conflicting constraints.

Share

COinS