Characterization of temperature trends for predicting heavy-duty drum brake failure through a rotational test rig

Abstract

This research investigates drum brake systems in heavy-duty vehicles at the wheel-axle level using a rotational brake test rig with temperature data acquisition. Conventional methods have limited capability to provide early and continuous detection of brake failure, especially in heavy-duty applications, while most previous researches focus on light vehicles with different thermal characteristics and degradation mechanisms. The results show that temperature-based pre-failure indicators can be identified through axle-level testing by analyzing deviations from a normal temperature threshold of 60°C, determined using the mean plus standard deviation criterion. The tests yielded thresholds of 46°C for vapor lock, 52°C for oil contamination, 60°C for normal braking, 69°C for continuous friction, and 78°C for lining wear. These temperature responses are classified into three condition levels, low, medium, and high, representing warning, normal, and critical braking conditions within a fuzzy logic based Early Warning System (EWS). The deviation patterns indicate changes in thermal behavior caused by brake degradation, as evidenced by reduced braking effectiveness at low temperatures and excessive heat accumulation at high temperatures. The proposed method derives temperature thresholds from statistical analysis of experimental data and translates them into overlapping fuzzy membership functions with reference points obtained from test results. This approach enables gradual and interpreTable interpretable identification of brake degradation and provides a basis for distinguishing different braking conditions. The findings support the development of brake condition monitoring systems for heavy-duty vehicles and provide a basis for implementing early warning systems in real vehicles through instrumentation adaptation and field calibration to improve braking safety