Hybrid high-performance computing enhanced machine learning framework for nano-thermal conductivity in MWNT-oil-based solar cooking systems

Abstract Solar cooking is a very pertinent alternative for energy-challenged regions, but conventional systems are miserably lacking in heat retention and efficiency, as well as adaptability to changing solar conditions. The presented theoretical study proposes an HPC hybrid model designed to optimize nano-thermal behavior in oil-based solar cooking systems with MWNTs. The proposed scheme is an amalgamation of five advanced modules: adaptive multiphase heat transfer modeling (AMPHTM), topological data analysis (TDA), graph-theoretic heat flow optimization (GTHFO), fractal-based multi-scale thermal transport modeling (FB-MTTM), and thermo-optical spectral mapping (TOSM-HA). These modules together provide real-time corrections for MWNT dispersion, topological clustering, and spectral mismatch while enhancing thermal transport at multiple scales. Modeling and simulation predicted enhanced effective thermal conductivity under dynamic solar conditions, ranging from 0.49 to 1.27 W/mK in the 0.01%–0.1% MWNT volume fraction. Reduction of heat loss by 45% and improvement of cooking efficiency by 25% to 30% in 30 min compared to baseline methods. The topological instability in MWNT dispersion was diminished using a clustering index of 0.027, and spectral absorption in the near-infrared region saw a 2.85-fold enhancement compared with base fluids. The very multi-paradigms adaptive thermosystem presents new horizons in precision thermal control for solar cooking and puts into perspective a real-time field-scalable envision for nano-thermal optimization in sustainable energy technologies in process.