Answers are grounded in this paper and cite their sources.
Comprehensive review highlights artificial intelligence techniques for optimizing thermal-fluid systems, indicating future directions for robust applications.
Artificial intelligence (AI) has become an active tool in thermal-fluid research, with applications in data-driven modeling, surrogate prediction, design optimization, flow-field reconstruction, and intelligent control. This review examines 160 related publications and groups them into four themes: heat-exchanger design, operation, and maintenance; thermal properties and flow characteristics of advanced working fluids; complex fluid dynamics and multiphase flow; and system-level energy management and control. These studies show that AI can reduce repeated CFD calculations, support fast parameter screening, identify design trade-offs, and improve prediction in selected thermal-fluid problems. Many reported gains are still tied to limited datasets, CFD-generated data, or narrow operating ranges, leaving their extrapolation to new fluids, geometries, and real operating conditions uncertain. More reliable use of AI in this field will require physics-informed and gray-box models, uncertainty quantification, reproducible benchmarks, and validation against experiments or high-fidelity simulations. AI mainly supports modeling, optimization, monitoring, and decision-making; physical models and experiments are still needed to explain mechanisms, test extrapolation, and justify design decisions.
No takes yet. Share an insight, caveat, or question.
Zhang et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: