Optimization of cooking vessel geometry for enhanced flame impingement heat transfer using a hybrid CFD and artificial intelligence approach
Issued Date
2026-09-01
Resource Type
ISSN
07351933
Scopus ID
2-s2.0-105042242400
Journal Title
International Communications in Heat and Mass Transfer
Volume
178
Rights Holder(s)
SCOPUS
Bibliographic Citation
International Communications in Heat and Mass Transfer Vol.178 (2026)
Suggested Citation
Kamma P., Loksupapaiboon K., Phromjan J., Boekfah A., Suvanjumrat C. Optimization of cooking vessel geometry for enhanced flame impingement heat transfer using a hybrid CFD and artificial intelligence approach. International Communications in Heat and Mass Transfer Vol.178 (2026). doi:10.1016/j.icheatmasstransfer.2026.111831 Retrieved from: https://repository.li.mahidol.ac.th/handle/123456789/117522
Title
Optimization of cooking vessel geometry for enhanced flame impingement heat transfer using a hybrid CFD and artificial intelligence approach
Corresponding Author(s)
Other Contributor(s)
Abstract
Flame–surface interaction governs the thermal performance of combustion-based heating systems; however, the role of surface topology in premixed flame-impingement heat transfer remains inadequately quantified. This study presents an integrated computational fluid dynamics (CFD) and artificial intelligence optimization framework to systematically evaluate premixed impinging flame jets interacting with cooking vessels of varying bottom geometries. Ten surface configurations—including flat, concave, wavy, rectangular, cavity, bulged, and triangular designs—were examined to elucidate the influence of geometric modification on flow topology, wall-jet development, and near-wall heat transfer behavior. The CFD results reveal that surface shaping strongly redistributes stagnation-zone structure and wall-jet momentum, leading to substantial variations in local thermal fields and overall thermal efficiency. Among the investigated configurations, the bulged-bottom vessel achieved the best performance, enhancing mass-specific thermal efficiency by approximately 50% relative to the conventional flat-bottom geometry. To further improve this design, a surrogate model based on an artificial neural network (ANN) was constructed using high-fidelity CFD datasets. The optimized multilayer perceptron architecture (5–10–15–2) demonstrated strong predictive accuracy (R<sup>2</sup> = 0.9256, MSE = 0.01995). The trained ANN was subsequently coupled with a genetic algorithm (GA) to perform global optimization of key geometric and operating parameters. The ANN–GA optimized configuration yielded a predicted thermal efficiency of 0.920, which was confirmed by additional CFD simulations with a deviation below 2.6%. Overall, the proposed CFD–ANN–GA framework provides a robust and computationally efficient strategy for optimizing flame-driven heat transfer systems, offering mechanistic insight and practical design guidance for high-efficiency cooking vessels and related thermal devices.