| Author: | Jiang, Ying |
| Title: | Development of a physiological-parameter-based thermal sensation model for outdoor microclimate assessment |
| Advisors: | Niu, Jianlei (BEEE) Xie, Yongxin (BEEE) |
| Degree: | Ph.D. |
| Year: | 2025 |
| Department: | Department of Building Environment and Energy Engineering |
| Pages: | xxvi, 281 pages : color illustrations |
| Language: | English |
| Abstract: | In the context of climate change, outdoor thermal comfort models serve as useful tools for evaluating thermal environment conditions and advanced active cooling strategies for outdoor users. Currently, no applicable and accurate model can provide a reliable prediction of thermal perception across diverse outdoor environments. There are several key limitations: 1) Most existing classical indices were developed based on indoor chamber studies, and many emerging outdoor thermal comfort models did not fully consider various outdoor thermal environments. 2) The developing outdoor thermal comfort models only considered overall thermal sensation prediction and did not provide the prediction of thermal sensation for local body parts, which are important in asymmetric outdoor conditions. 3) Few studies explored the combined psycho-physiological effect of thermal alliesthesia and adaptation phenomena on dynamic thermal perceptions in real-life environments. This study aims to build a thermal sensation model that correlates the subjective thermal sensation votes (TSV) with the human-body thermo-physiological responses in warm-biased outdoor conditions, in order to improve the assessment of outdoor thermal comfort in urban spaces during the design and renovation stages. The study includes the steady-state model, the step-up phase, and the step-down phase of the dynamic model, as well as an exploratory investigation into thermal alliesthesia and short-term adaptation phenomenon. The steady-state part involves the local TSV model for 11 body parts when the human body is in a steady state, as well as an overall TSV model. A series of semi-controlled experiments were conducted, with a total of 86 human subject tests under varying wind speeds and solar radiation levels using an adjustable windshield and shading device in real outdoor settings. Questionnaire survey data on local and overall TSV and simultaneously measured data of environmental parameters and physiological responses in terms of local skin temperatures of 18 body parts, core temperature, and sweat rate of 3 body parts, were collected. Results show a strong correlation between local sweat rate and local skin temperature, leading to the selection of local skin temperature as the sole variable for the local TSV model. Individual differences and thermal sensation overshoot outdoors were observed, and a new parameter accounting for the neutral skin temperature variation was introduced into the local TSV model to provide three prediction values, i.e., the mean, the 10th, and 90th percentiles. The overall TSV model, which calculates overall TSV as a weighted average of local TSVs, demonstrated a high fit with actual overall TSV (R² = 0.96). The step-up phase of dynamic model provides thermal sensation prediction when the human body experiences positive changes in skin temperatures. This model uses the dynamic-state dataset collected from the semi-controlled experiment when subjects transitioned from indoors to outdoors. To address naturally occurring thermal fluctuations outdoors, two fluctuation modes are defined, namely, the highly-dynamic mode and the weakly-dynamic mode. A classification method was developed based on the empirical critical values of the derivative of local skin temperature (dTsk,i/dt), categorizing data according to the percentage of values exceeding the critical threshold within a 30 s window. The step-up phase of the dynamic model was developed via linear regression, utilizing the derivative of mean skin temperature (dTsk,m/dt) as the independent variable for the weakly-dynamic mode, while both dTsk,i/dt and dTsk,m/dt were used for the highly-dynamic mode. The dynamic model exhibits a satisfactory performance, with an average Accuracy of 74.2% for the highly-dynamic mode, and 73.8% for the weakly-dynamic mode. The step-down phase of dynamic model provides thermal sensation prediction when the human body experiences negative changes in skin temperatures. This model is developed based on experiments involving 142 human subjects who experienced step-change thermal environments, during which both environmental parameters including skin temperatures of 12 body parts and subjective votes via questionnaire survey, in terms of TSV, thermal comfort vote (TCV), thermal acceptability (TA), thermal pleasure (TP), and thermal stay willingness (TSW), were collected. Around 700 subjective surveys were used in the step-down phase model development. The steady-state part prediction of the abdomen, foot, and neck has been updated in this study. Furthermore, a combined sun and wind condition index (SWI) correction term was added to quantify different effects of wind and solar radiation on TSV. Skin temperature data were classified into the highly-dynamic and weakly-dynamic modes. Lower correlations were found between dTsk,i/dt and dynamic local TSV and between the derivative of Physiology Equivalent Temperature (dPET/dt) and dTsk,i/dt in the weakly-dynamic mode than those in the highly-dynamic mode. Comparing the two phases, higher associations between dPETs and dTSVs in the step-up phase were observed than those in the step-down phase. The step-down phase model was developed using linear regression, utilizing dTsk,i/dt as the independent variable in both the highly-dynamic and weakly-dynamic modes. For most body parts, the model exhibits satisfactory results. The average Accuracy for the highly-dynamic model is 62.1%, and 69.0% for the weakly-dynamic mode. In addition, among all the subjective thermal environment assessment scales TSV, TA, TP and TSW, TSW illustrates the largest rs with PET during both the transition period (exposure time ≤ 120s) and the prolonged exposure period (exposure time > 120s). TSW appears to be more suitable for the assessment of outdoor thermal conditions. Finally, as an extension of the previous findings, thermal alliesthesia and short-term adaptation phenomenon in outdoor conditions are explored. This study examined the dynamic thermal perceptions and physiological responses of 50 subjects exposed to transient outdoor environments, including underneath an elevated building (UEB) and sunlit areas. Results show that when thermal environment changed, overshoot in thermal sensation persisted for at least 5 mins, and its influence on thermal comfort increased and then diminished at around 5 min, at which point thermal adaptation began to occur. The current thermal comfort was affected by the preceding thermal status: in the strong alliesthesia zone, a 1.00 scale decrease in changes in thermal sensation vote (dTSV) resulted in a 0.44 scale increase in changes in thermal comfort vote (dTCV); whereas in the moderate alliesthesia zone, a slight dTSV within 1.00 scale could positively affect dTCV. Skin temperature of exposed segments correlated better with TSV than that of unexposed ones. Skin temperatures had lower correlations with TSV compared with experiments at static conditions conducted by other researchers. Besides thermal alliesthesia and adaptation effects, sweat accumulation and evaporation are possible reasons for the low correlations. For the outdoor thermal sensation model developed in this study, there are four main differences from other mainstream models. Firstly, this study employed a series of semi-controlled experiments and field experiments to cover various outdoor settings for model development. Secondly, the model accounts for individual differences and thermal sensation overshoot effect outdoors. It predicts thermal sensation using three values - mean, 10th percentile, and 90th percentile - covering about 80% of individuals. Thirdly, in contrast to indoor models, the dynamic part of this model captures natural environmental fluctuations in outdoor settings by classifying data into two fluctuation modes, highly-dynamic and weakly-dynamic modes. Lastly, the complete model provides three sets of local TSV predictions for 11 body parts, as well as three values of overall TSV predictions based on local skin temperatures and the derivatives of local and/or mean skin temperatures. It is designed to be integrated with a multi-nodal human thermoregulation model to numerically assess thermal comfort in outdoor spaces during the design and renovation stages. The findings provide insights into thermal evaluation scales for outdoors and are beneficial for improving thermal comfort assessments in outdoor settings. |
| Rights: | All rights reserved |
| Access: | open access |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/14182