Multiscale and Multidomain Digital Twins for Vehicle Thermal Management Applications

Abstract

<div class="section abstract"><div class="htmlview paragraph">Effective thermal management is crucial for vehicles, impacting both passenger comfort and safety, as well as overall energy efficiency. Electric vehicles (EVs) are particularly sensitive to thermal considerations, as customers often experience range anxiety. Improving efficiency not only benefits customers by extending vehicle range and reducing operational costs but also provides manufacturers with a competitive edge and potential revenue growth. Additionally, efficient thermal management contributes to minimizing the environmental impact of the vehicle throughout its lifespan.</div><div class="htmlview paragraph">Digital twins have gained prominence across various industries due to their ability to accelerate development while minimizing testing costs. Some applications have transitioned to comprehensive three-dimensional models, while others employ model reduction techniques or hybrid approaches that combine different modeling methods. The discovery of unknown working mechanisms, more efficient and effective control systems are some of the benefits brought by digital twins.</div><div class="htmlview paragraph">The current competitive market conditions in the automotive industry and in particular in the Indian market requires a fast and cost-effective method to develop and test novel features. This is specially pressing for features with significant energy consumption due to the impact they have on range of electric vehicles (EV) and consequently on the perception of the customers and even the revenue of the company.</div><div class="htmlview paragraph">In this study we present a digital twin of the HVAC system in a vehicle that facilitates a transition between lengthy and computationally expensive CFD simulations of both in-cabin and external flows with fast, real-time capable simulations for embedded applications. We show how this digital twin strategy informs control strategies to reduce operational costs and discover new system behaviors which could lead to further control improvements. The proposed digital twin approach minimizes reliance on the specific vehicle model, allowing for rapid implementation and adaptability.</div></div>