The content discusses the potential use of carbon fiber composite as a replacement for metal-based battery enclosures in electric vehicles (E.V.s). The advantages of carbon fiber, such as its strength-to-weight ratio and corrosion resistance, make it a suitable candidate. However, the strength of carbon fiber structures depends on various parameters, which need careful consideration.
To address this, the authors used high throughput finite element analysis (FEA) based thermoforming simulation to virtually manufacture battery enclosures with different design and processing parameters. They then conducted virtual crash simulations to assess the crashworthiness of the enclosures, specifically simulating a side pole crash.
Using the data from these crash simulations, the authors developed predictive models using machine learning (ML) techniques to understand the crashworthiness metrics. The ML models demonstrated excellent performance, with an R2 value greater than 0.97, in predicting metrics such as crush load efficiency, absorbed energy, intrusion, and maximum deceleration during a crash.
The authors believe that this FEA-ML framework can assist in selecting process parameters for carbon fiber-based component design and can be applied to other manufacturing technologies as well.