Predictive model of local and overall thermal sensations for non-uniform environments

This paper proposes a quantified model for local and overall thermal sensations in non-uniform thermal environments. Human subject tests were carried out in a controlled environmental chamber. In total, 179 tests were performed on 112 participants, including 52 neutral condition tests and 127 local ventilation tests. Seven body parts (head, chest, back, arm, hand, leg and foot) were independently cooled or heated by local ventilation, while the rest of the subjects' bodies were exposed to a warm or neutral environment. Simultaneously, participants' responses for both local and overall thermal sensations were recorded. Skin temperatures were monitored by thermal resistance sensors on 14 of the local test sites. The relation between local thermal sensations and skin temperatures were analyzed using correlation statistic analysis. The principal component regression approach was adopted to eliminate the significant multicollinearity of thermal sensation among each body part. A predictive model was proposed for overall thermal sensation that integrates a local thermal sensation psychological model and an overall thermal sensation weighting factor model. The results show that local thermal sensations can be predicted by skin temperatures to a high degree of accuracy for local ventilation conditions. The weighting factors of local thermal sensation on overall thermal sensation produce integrated values, and there is little difference among the values derived from the seven body parts. Since this model is based on skin temperature, its application is not limited by environmental parameters and clothing conditions, and as a result it will be useful for designing and evaluating non-uniform thermal environments.