Designing effective heating, ventilation and air conditioning systems for railway cabins requires detailed, spatially resolved information as thermal loads vary significantly across different regions of the human body. Conventional thermal comfort assessments typically rely on point-based flow field evaluations to describe the environmental conditions at selected locations [34]. These point-based methods provide limited information regarding the actual local heat exchange experienced by passengers, failing to capture complex physiological interactions. Current cabin evaluations often lack a consistent integration of high-fidelity boundary conditions, such as radiative heat transfer and solar loading, alongside a representative physical obstruction of the passenger. This thesis develops a comprehensive numerical simulation-based methodology using a computational manikin (CM) to provide high-resolution, local thermal-comfort assessments. The framework integrates a validated CM workflow and employs linearized equivalent heat-transfer coefficients with radiative heat-transfer modeling and physically consistent boundary conditions. The methodology is demonstrated through a railway cabin numerical simulation model where the CM’s impact on flow and heat transfer is compared against experimental data and varying boundary-condition fidelities. The study extends the workflow by utilizing a reduced model paired with a neural network to map simulation parameters to boundary surface fields. Findings show that the manikin and proposed workflow significantly alters the local environment and comfort predictions, and the surrogate model successfully reduces evaluation costs through modal compression. However, the surrogate’s accuracy remains limited by training data coverage and regression sensitivity, highlighting a trade-off between speed and precision. Overall, the methodology improves spatial resolution for cabin design while providing a path toward faster, parametric comfort studies.      «

Designing effective heating, ventilation and air conditioning systems for railway cabins requires detailed, spatially resolved information as thermal loads vary significantly across different regions of the human body. Conventional thermal comfort assessments typically rely on point-based flow field evaluations to describe the environmental conditions at selected locations [34]. These point-based methods provide limited information regarding the actual local heat exchange experienced by passenge...     »