The miniaturization and high reliability for automotive electronic components has been strongly requested. Generally, electronic component and printed wiring board are connected using solder joint. The reliability of solder joint has widely dispersion. For the dispersion reduction of solder joint reliability, not only design factors but manufacturing factors should be optimized. The evaluation of manufacturing factors for solder joint reliability was very difficult by experimental evaluation alone. Therefore, the reflow process simulation was established. The simulation was reenacted soldering process on chip component, which was the most severe reliability in automotive electronic components. The novelty of simulation was the coupled analysis of flow and rigid for simulating self-alignment of chip component. In this simulation, contact angle and surface tension was very important factor. So, these characteristics were measured based on Spread test and Wetting balance tests using the specimens. In the result, the solder joint shape of analysis was agree with the one of specimens using the measured contact angle and surface tension. Next, the effect of manufacturing process dispersion for solder joint shape was evaluated. The factors were mount offset and length unbalance of electrodes on chip component. As a result, the mount offset was not affected solder joint shape of chip component until a certain level. Also, the unbalance of electrode of chip component was not almost affected for solder joint shape of chip component because a part was moved to the center of part by surface tension of solder joint. Finally, the relation between the estimated solder joint shape and fatigue life of solder joints is evaluated using crack propagation analysis based on Manson-Coffin’s law and Miner’s rule. When the value of mount offset was large, the crack propagation mode was changed and the fatigue life of solder joint was decreased. As mentioned above, it was able to evaluate the relation between manufacturing factors and solder joint reliability. Accordingly, this simulation was very useful for consideration on the miniaturization, high reliability and appropriate margin for design of electronic components.
Investigation of reliability assessement in power electronics circuits using machine learning
