Vehicle windshield design and design verification using computer aided analysis and neural networks

The design of vehicle windshields is one of the most important factors of vehicle design in terms of aerodynamics and aesthetics. Nowadays, windshield design is a very critical part for passengers and pedestrians because of the increasingly important safety criteria. The scope of this study is that vehicle windshield design and verification processes with computer aided analysis and integration of thermal analysis outputs into artificial intelligence application. The windshield which is considered within the scope of the study has the defrosting and defogging function with an integrated heater coating inside the glass. After designing the 3d data of a windshield with integrated heating coating, head impact and thermal analysis had been performed. Head impact analysis of the design with finite element models were completed and the results were detailed. Thermal analysis were performed using the computational fluid dynamics technique considering the heating function and the flow rate and temperature values of the air coming out of the grill geometries. Thermal analysis results are integrated with artificial intelligence. Artificial neural network models technique was used. Based on the artificial intelligence model trained with the analysis results, predictions were made about the analysis outputs and the thermal analysis results were compared with the outputs.