Characterization of human extreme heat exposure using an outdoor thermal manikin.

Extreme heat is a current and growing global health concern. Current heat exposure models include meteorological and human factors that dictate heat stress, comfort, and risk of illness. However, radiation models simplify the human body to a cylinder, while convection ones provide conflicting predictions. To address these issues, we introduce a new method to characterize human exposure to extreme heat with unprecedented detail. We measure heat loads on 35 body surface zones using an outdoor thermal manikin ("ANDI") alongside an ultrasonic anemometer array and integral radiation measurements (IRM). We show that regardless of body orientation, IRM and ANDI agree even under high solar conditions. Further, body parts can be treated as cylinders, even in highly turbulent flow. This geometry-rooted insight yields a whole-body convection correlation that resolves prior conflicts and is valid for diverse indoor and outdoor wind flows. Results will inform decision-making around heat protection, adaptation, and mitigation.

Impact of human body shape on forced convection heat transfer.

Predicting human thermal comfort and safety requires quantitative knowledge of the convective heat transfer between the body and its surrounding. So far, convective heat transfer coefficient correlations have been based only upon measurements or simulations of the average body shape of an adult. To address this knowledge gap, here we quantify the impact of adult human body shape on forced convection. To do this, we generated fifty three-dimensional human body meshes covering 1st to 99th percentile variation in height and body mass index (BMI) of the USA adult population. We developed a coupled turbulent flow and convective heat transfer simulation and benchmarked it in the 0.5 to 2.5 m·s-1 air speed range against prior literature. We computed the overall heat transfer coefficients, hoverall, for the manikins for representative airflow with 2 m·s-1 uniform speed and 5% turbulence intensity. We found that hoverall varied only between 19.9 and 23.2 W·m-2 K-1. Within this small range, the height of the manikins had negligible impact while an increase in the BMI led to a nearly linear decrease of the hoverall. Evaluation of the local coefficients revealed that those also nearly linearly decreased with BMI, which correlated to an inversely proportional local area (i.e., cross-sectional dimension) increase. Since even the most considerable difference that exists between 1st and 99th percentile BMI manikins is less than 15% of hoverall of the average manikin, it can be concluded that the impact of the human body shape on the convective heat transfer is minor.

Human body radiation area factors for diverse adult population.

Radiation accounts for a significant fraction of the human body and environment heat exchange and strongly impacts thermal comfort and safety. The direct radiative exchange between an individual and a source or sink can be quantified using the effective (feff) and projected radiation area factors (fp). However, these factors have not been quantified for half of the population of the USA with an above-average body mass index (BMI). Here, we address this gap by developing thirty male and thirty female computational manikin models that cover the 1 to 99 percentile variation in height and BMI of adults in the USA. The radiative simulations reveal that the feff and the fp angular distributions are nearly independent of gender, height, and BMI. Appreciable relative differences from the average models only emerge for manikins with BMI above 80th percentile. However, these differences only occur at low zenith angles and, in absolute terms, are small as compared to variations induced by, for example, the zenith angle increase. We also use the manikin set to evaluate whether the body shape impacts the quality of human representation with several levels of geometrical simplification. We find that the "box/peg" body representation, which is based on the hemispherical fp average, is independent of the body shape. In turn, the fp distributions averaged over the azimuth angle range, representing the rotationally symmetric humans, are only impacted to the same degree as for the anatomical manikins. We also show that the anatomical manikins can be closely approximated by the multi-cylinder and sphere representation, at least from a radiation perspective. The developed anatomical manikin set is freely available and can be used to compute how body shape impacts a variety of external heat transport processes.