Coupling CFD and Quantum Machine Learning Algorithms for Simulating PMMA Exposed to Pool Fire

A pool fire is a turbulent diffusion fire that occurs when a flammable liquid, such as gasoline or oil, spills, forms a pool on a surface (or floats on water), and ignites. It is characterized by low-momentum, buoyancy-driven flames, significant thermal radiation, and a heat feedback mechanism that sustains fuel evaporation. The numerical simulation of PMMA pyrolysis using Computational Fluid Dynamics (CFD) and Quantum Machine Learning (Time Series Forecasting by using Pytorch and Pennylane libraries) represents a cutting-edge approach combining high-fidelity physical modeling with hybrid quantum machine learning (QML) for material optimization and durability enhancement. While CFD models predict transient thermal boundary conditions and heat transfer reduction, QML and classical machine learning (ML) accelerate design optimization. The PMMA surface recession obtained by using FDS software version 5 has been successfully validated against the analytical method. It has been shown that the calculated recession velocity of the PMMA obtained by FDS is 0.074 mm/s. The calculated recession velocity of the PMMA obtained by using the pyrolysis kinetic model is 0.079 mm/s. Thus, the error is about 6.12%. The coupled hybrid classical and quantum time forecasting algorithm predicts very well the temperature reading of TC1 thermo-couple obtained by FDS simulation. This theoretical work uses mass loss rate derived from non-isothermal pyrolysis kinetics. In the framework of this research, a coupled classical–quantum algorithm has been developed in order to solve the one-dimensional transient heat conduction in poly (methyl methacrylate) (PMMA), whose surface pyrolyzes away at a constant recession velocity, with a constant temperature imposed on both faces. The analytical and numerical obtained for the temperature profile are similar.