LAUSR

Abstract Human thermoregulation models, particularly the Fiala model, are well accepted for the prediction of thermoregulatory responses and the assessment of clothing and building systems with regard to human thermal comfort. However, the convective heat transfer coefficients (hc), used for the heat transfer calculations, were obtained from measurements with a thermal manikin with poor body resolution from decades ago, and they were not distinguished between body postures. In this study, the overall and local hc of the human body in both standing and seated postures were evaluated by CFD simulations and implemented in the Fiala model to investigate the resultant influence on the predicted thermoregulatory responses by comparing the predictions with measurements of 14 human exposures. It was found that the original hc used in the Fiala model was similar to the simulated hc of the standing body, but higher than that of the seated body at most body parts. In 64% of the investigated human exposures, the root-mean-square-deviation of the skin temperature predicted by the Fiala model with the simulated hc was lower than that achieved with the original hc, indicating an improvement of the accuracy of the Fiala model. Additionally, the higher hc of the standing body resulted in a lower mean skin temperature by up to 1.5 °C when compared to the seated body in the environment of 10 °C and 2.5 m/s. This emphasises the necessity of ensuring the accuracy of the hc in the thermoregulation model in order to improve its validity for specific investigated conditions.