Author(s): ALKEMADE, P., ZAHNER, M., Institution: GREENTEG AG, Country: SWITZERLAND, Abstract-ID: 2798

INTRODUCTION:
Wearable heat flux (HF) sensors measure dry heat transfer between the skin and environment. However, during exercise and exposure to hot conditions, heat loss occurs predominantly through sweat evaporation, causing HF sensors to underestimate total heat loss. Quantifying total heat loss (i.e., dry and wet loss) could unlock new opportunities in wearable monitoring of skin and core temperature, heat strain, and energy expenditure.
HF sensors occlude underlying skin, preventing local skin cooling through sweat evaporation. Accordingly, local evaporation rate may be reflected by the difference between skin temperature underneath the sensor (i.e., covered; Tcov) and skin temperature next to the sensor (i.e., exposed; Texp) (dT). Local total HF (dry+wet) may then be calculated using a physical model: HFtot = HFcov + (dT/Rskin), with HFcov as covered HF and Rskin as the skin’s thermal resistance.
The aim of this proof-of-concept study is to evaluate if (1) dT is related to evaporation rate, and (2) total HF can be estimated using the proposed physical model.
METHODS:
HFcov and Tcov were measured using a CORE1 sensor. Texp was measured using a contactless sensor attached to the CORE1, directed towards the adjacent skin.
Bench-top experiments were conducted with the prototypes placed on a hot plate (HFref 250 W/m2; ambient temperature ~25°C) at different evaporation rates: “dry”, 60 min at dry plate without convection (<0.3 m/s); “wet still”, 60 min at wetted plate (<0.3 m/s); “wet wind”, 60 min at wetted plate with increased air flow (1.5 m/s).
In-vivo experiments were conducted in the lab (24.6°C, 30%RH, increased air flow), aiming to induce varying sweat rates while minimizing unevaporated sweat. Two participants completed 15 min seated rest, 25 min low- and 25 min high-intensity cycling, and 10min passive recovery. For each stage, dT was measured on the lateral chest, upper back, and thigh, and whole-body sweat rate (WBSR) was calculated from pre- and post-stage nude body mass.
RESULTS:
On a dry hot plate, dT was -0.3°C, with HFcov (247 W/m2) and HFtot (238 W/m2) comparable to HFref (250 W/m2). On a wet hot plate, dT was 3.4°C (still) and 5.9°C (wind). HFcov underestimated HFref (still, 142; wind, 93 W/m2), while HFtot was closer to HFref (still, 232 W/m2; wind, 249 W/m2). dT explained the HFcov underestimation (%) with 0.1*dT + 0.05 (R2=0.99).
Across the four stages, local dT and WBSR ranged -0.7-6.0°C and -0.2-1.2 L/h. In both participants, local dT’s were linearly related to WBSR (R2=0.76-0.99). Across participants, average dT over all positions was linearly related to WBSR normalized to body surface area (R2=0.81).
CONCLUSION:
Preliminary bench-top and human experiments indicated a linear relationship between dT and evaporative heat loss. Bench-top testing further provided proof of concept for using the physical model to estimate total heat loss. Additional experiments are needed to validate the model’s accuracy across various sensor positions, participants, and environments.