Thermo-Mechanical Material Screening and Constraint-Based Optimization of an Automotive Clutch Using Response Surface Methodology
Abstract This study presents an integrated computational framework combining multiscale material homogenization, transient thermo-mechanical analysis, and constraint-based optimization for automotive clutch design. Using ANSYS Material Designer, polymer matrix composites containing 15 vol.% Al₂O₃ were initially evaluated against a commercial friction material (COM_FM). Preliminary finite element analysis (FEA) indicated that increased thermal conductivity alone was insufficient to limit localized temperature rise. A composite containing 25 vol.% Al₂O₃ (PMC-M1-25) was therefore selected for further investigation. A 30-point design space was constructed using Latin hypercube sampling (LHS) with an optimal space-filling (OSF) configuration, considering clamping pressure (1.0–3.0 MPa), engine torque (150–350 N.m), and friction coefficient (0.25–0.40). Response surface methodology (RSM) was used to model global component deformation and peak interface temperature. The maximum sampled temperature reached 456.83°C, highlighting the importance of evaluating thermally unfavorable operating conditions. Sensitivity analysis and response surface models were used to examine the effects of the operating parameters on the thermo-mechanical responses. Constraint-based optimization was then applied to identify operating conditions satisfying the prescribed thermal limit, followed by direct FEA verification of the selected condition. The proposed framework links composite material homogenization, thermo-mechanical response prediction, and constraint-based optimization to support the numerical assessment of automotive clutch operating conditions.
Authors
- Thanh Dinh Van
- Trinh Phi Hoang
Institutions
- East Asia University of Technology (VN)
Publication Details
- Journal
- Research Square
- Published
- 2026-11-16
- DOI
- https://doi.org/10.21203/rs.3.rs-11421036/v1
- Primary Topic
- Brake Systems and Friction Analysis
- Type
- preprint