Comparison of Brake Cooling System on Bus: Simulation and Evaluation of Natural vs Forced Air through Machine Learning Processing

Abstract

Intense braking can cause heat accumulation that can reduce the braking system’s efficiency and damage the braking system components due to changes in material structure. The high temperature in the braking system of heavy vehicles requires an air-cooling system that produces high-pressure air. This study compares the cooling process results between natural air and forced air. CFD modeling was chosen as the method to simulate the cooling process of bus drum brakes. The CFD method works by involving a network of elements in a system called a mesh. In CFD, the mesh is a component that focuses on dividing the fluid domain into small elements to analyze fluid flow, heat transfer, and dynamic phenomena. Using a cooling process, the brake system temperature gradually decreased as the wind speed increased. The examination proved that natural air could reach the highest temperature of 168.73 °C and the lowest temperature of 63.01 °C. Meanwhile, forced air could reach the highest temperature of 73.8 °C and the lowest temperature of 30.73 °C. Statistical models and machine learning were carried out to predict the accuracy of simulation from SOLIDWORKS. The values of average temperature after the cooling process are examined by employing the Gradient Boosting algorithm with the highest accuracy of prediction (R²) value of 0.999895 in natural conditions and 0.999822 in forced conditions. Overall, forced air shows better temperature reduction than natural air. By lowering the operating temperature, drum brake systems can extend their lifespan.