Influence of Uncertain Factors on Automotive Electronics Thermal Simulation

Abstract

<div class="section abstract"><div class="htmlview paragraph">The electronic content in automotive has increased over recent years and expected to contribute about 50% of the total vehicle cost by 2030. The semiconductor research indicates that focus is on enhancing the functionality of single device and miniaturizing of components to reduce the electronic module size. It is to be ensured, that devices in automotive electronic modules should be within its allowable temperature limit while operating at harsh environment. The accurate virtual simulations using CAE tools prior to proto build can assist in understanding the design risks upfront and aids in arriving at a reliable thermal mitigation solution. The prediction accuracy of thermal simulation is driven by the inputs and modeling approach used in the analysis. Current automotive electronic product development trend indicates, chip development and thermal design of electronics module goes in parallel. Hence the access of device intricate details and accurate power estimation for thermal simulation is not feasible during initial design phase. An electronic module will have multiple devices and modeling each device in detail along with explicit PCB trace modeling requires huge computational resource and time consuming. This indicates that simplified modeling techniques assisted with engineering assumptions are essential for virtual simulation that creates inevitable uncertainty in temperature prediction. In this study a generic automotive electronics system is used as an example to showcase the influence of uncertainty factors such as modeling techniques, power input load and boundary condition on thermal simulation. High fidelity thermal simulation model is used as reference and DOE studies are performed at different levels for the identified uncertainty factors. The uncertainty factors are studied individually and as combination to understand its influence in temperature prediction and its findings are highlighted. The approach showcased in this work can be adapted to understand the influence of uncertainty factors on thermal simulation for any automotive electronic module.</div></div>