Electric vehicles are practically computers on wheels. New innovations such as active and passive safety systems, electric propulsion, and semi and fully autonomous vehicles have all contributed to an increase in the usage of electronics in automotive applications. More importantly, automotive designers must still adhere to the same size and packaging constraints to ensure vehicles’ size and weight does not increase. To resolve this dilemma, automotive designers often rely on components being tightly placed on both sides of the Printed Circuit Board (PCB) to ensure the most efficient use of board space.
To ensure electronic packages can withstand harsh environments that automotive products are often subjected to, electronic designers perform qualifications tests by exposing electronic packages, such as the widely used Ball Grid Array (BGA), to accelerated thermal cycling. Designers must consider numerous component features and their locations on the PCB that can impact the reliability of electronic packages under harsh environments. Performing thermal cycling tests on numerous combinations is a costly and time-consuming step of the qualification process.
When time to market is more critical than ever before, every minute spent on physical thermal cycling tests may put automotive suppliers at a significant competitive disadvantage. Ideally, you want to determine design robustness prior to initiating physical qualification tests. Virtual testing of various components, combinations of components, and their locations makes testing faster and provides more insight into the ultimate combinations for highest reliability.
Let’s get back to BGAs and their increased use in the automotive industry and reliability issues associated with them. Depending on where the components are placed and what type of coating is used can greatly affect time to failure of these electronic packages. However, the time required to perform thermo-mechanical simulations that require detailed models consisting of the electronic package, printed circuit board (PCB) and housing can be time-consuming and tedious. Additionally, simulation results only tell you time to failure but do not provide insights into the reasons for that failure.
Reliability physics analysis (RPA) provides a complementary tool to thermal cycling testing by combining finite element analysis (FEA) and knowledge of failure mechanism into virtual qualification process that can determine design robustness prior to initiating physical qualification testing. Sherlock Automated Design Analysis software generates a full-scale detailed model of PCBs and components and runs them through thermo-mechanical analysis in a matter of minutes. Engineers can virtually test temperature, shock, vibration dependent material properties by running them through incremental changes and tracking their time to failure. Additionally, engineers can calculate fatigue life of solder joints by changing underfill, coating, their location, and pressure applied to the board.
Let’s say an automotive supplier wants to predict low cycle fatigue of its electronic packages subjected to thermal mechanical cycling. With Sherlock they could build a detailed board model that features BGA components and specifies their location on the board and any conformal coating type. With the board built, they could define the life cycle for the board including specific vibration temperature and shock loads. Engineers could then run thermal mechanical analysis on multiple variations of the BGAs with multiple coatings, and location. Running such analyses early in the design process allows to determine the optimal design before building a physical model and testing it in the field environment.
When running such an analysis recently, an automotive supplier determined that mirrored components resulted in a higher probability of failure when compared to unmirrored ones. Additionally, they found out that conformal coating also increased the probability of failure, but the effect was affected by the components and the location. Due to these insights early in the design process, engineers knew they needed to further investigate and understand how the coating used, its thickness and the components coated could impact the final probability of failure.
Running these analyses virtually early in the design process provides a lot of insights that are helpful in building an ultimate physical model capable to withstand physical field tests and doing it quickly and precisely.