Electrical Driven Compressor – Vibration Correlation for PCB & Electronic Components
2023-01-0535
04/11/2023
- Features
- Event
- Content
- Electrical driven compressors (EDC) are mounted on the automotive driveline to compress refrigerants in an automotive air conditioning system. The electrical power delivered by the battery is converted to mechanical power by an inverter and an electrical motor. The inverter is composed of a printed circuit board (PCB) with assembled electronic components. For the inverter part of the EDC, the vibration failures are predominant followed by thermal failures on electronic components. Hence robust methodology is necessary to improve the strength of the design under vibration environment through Finite Element Analysis. In this paper, a multi-fidelity methodology to validate automotive electronic components under harmonic loading through Finite Element Analysis is presented. This multi-fidelity methodology allows the risk of failure to be assessed at the earliest stages of product design, when changes are easy to make and have a low financial impact. The main focus is on the last stage of this methodology, which corresponds to the high fidelity model. The main features of this high-fidelity model are reviewed: from the modeling of the Printed Circuit Board which is considered as a composite material, to the glue modeling, through meshing and numerical method used, up to the loading applied to the model. The discussion will mention the main multi-fidelity results up to a vibration fatigue analysis to validate the life of components to check whether it meets the customer requirement. Next, the accuracy of the high-fidelity simulation results are discussed in comparison with the vibration test results. Finally, this robust methodology supports us to minimize the proto samples and quick evaluation of products.
- Pages
- 8
- Citation
- Duraipandi, A., RAJA, A., LEON, R., RIBOT, H. et al., "Electrical Driven Compressor – Vibration Correlation for PCB & Electronic Components," SAE Technical Paper 2023-01-0535, 2023, https://doi.org/10.4271/2023-01-0535.