Influencing factors and a prediction model for thermal performance of protective clothing based on the AHP–CRITIC method

Abstract

The thermal protection performance (TPP) of fire clothing is crucial to the safety of firefighters. The total burn area of burning dummies is an important index to evaluate the TPP of fire clothing. The prediction of the total burn area of burning dummies through the TPP experiment of fabrics and the physical properties of fabrics is the key to evaluating the TPP of fire clothing. This work employed the analytic hierarchy process (AHP)–Criteria Importance Through Intercriteria Correlation (CRITIC) combination weighting method and correlation analysis to determine the effect of various factors on the total burned area of the dummies. The results of the study showed that the weight of the fabric, TPP, garment shrinkage rate, and the flash fire exposure time had a greater influence on the total burned area of the burning dummies, while the thickness of the fabric had less effect on the total burned area of the burning dummies. In this work, the combined assignment method was used to determine the weights of the influencing factors in the influencing factor analysis, and this method made the decision results more comprehensive, accurate, and reliable. The weight vector of each influencing factor was calculated as $\text{α}$  = (0.1831, 0.2120, 0.3271, 0.2258, 0.0520), thus it can be obtained that the weight value of TPP is the highest, whereas the weight values of thickness and flash time are relatively low. In addition, a prediction model for the total burned area of the combustion dummy value was established using the multiple linear regression fitting method. The mean relative error (MRE) was employed to validate the accuracy of the predictive model, yielding an MRE of 18.39%, which falls within the acceptable range. This method of predicting the total burn area of a burning dummy through small-scale fabric TPP experiments eliminates the need for full-scale burning dummy experiments, which reduces the cost and complexity of the experiments and is more time-efficient.