Evaluation of heat loss models for predicting temperature ratings of sleeping bags

Manufacturers label their sleeping bags with temperature ratings to help consumers select the bags that will provide them with thermal comfort during the expected conditions of use. However, there is no consistency among companies regarding their methods for determining temperature ratings. This study evaluated different heat loss models for predicting the lowest air temperature for acceptable levels of thermal comfort during 8 hours of sleep (where the body has incurred heat debt, but not enough to wake the person due to cold discomfort). The insulation values of a rectangular bag and two mummy bags used with thermal underwear, socks, 1 ¾ inch self-inflating pad, and pillow were measured with a heated manikin (ASTM F 1720). The average insulation values for 10 bags of each type were (1) low = 4.3 clo, (2) medium = 5.6 clo, and high = 6.6 clo. Sixty subjects were assigned to three groups with 10 men and 10 women in each group. Each group slept in one type of sleeping bag in an environmental chamber at three or four air temperatures in 5°C intervals. Each subject's skin temperature at the base of the big toe was recorded every 10 minutes throughout the night. Each morning, thermal sensation responses were collected from subjects using questionnaires. The calibration statistical method was used with the thermal sensation response of 4 (slightly cool) to determine the 95% prediction limits for the lowest air temperature for acceptable thermal comfort at three levels of insulation. The results indicated that the KSU model was consistent with the experimental data at each level of insulation, whereas the other models' predictions were lower. In addition, females had lower toe temperatures than males at colder air temperatures. Statistics indicated that the gender difference was due in part to the effects of body size (height and weight). Female subjects gave lower thermal sensation responses for the body and for their feet than the male subjects at the low air temperatures also. Spearman rank order correlation coefficients indicated that the average toe temperature was highly correlated to the subjects' thermal sensation of the feet (0.71).