Search

As handheld electronic products are more prone to being dropped during useful life, package to board interconnect reliability has become a major concern for these products. This has prompted the industry to evaluate the drop performance of CSP packages while mounted on printed wiring boards using board level drop testing. Although a new board level test method has been standardized through JEDEC (JESD22-B 111), characterization tests take quite a long time to complete, extending the design cycle. This paper proposes a method to compare and evaluate the drop performance through simulations at the design stage. A global-local approach is used to first determine the dynamic response of the board during drop and then to translate it into stresses and strain energy density in solder joints and intermetallic layers. The dynamic response of the board is validated by using data from actual board level testing as per JEDEC standard. The solder joint and intermetallic stresses are then related to drop to failure test data to derive a prediction model. The method is then applied to quantify the effect of package design parameters on the drop performance. Factors considered include material set, thickness of various material layers, pad size, and ball size. The same factors were tested in board level drop to further validate the prediction model. Experiments were also conducted to quantify the effects of package ball pad finish on the drop performance through board level testing according to JESD22-B111. The results indicate that the drop performance can be increased by a factor of 4 or more by changing package design and material variables.

Automated design space exploration was implemented and demonstrated in the form of the multidisciplinary optimization of the design of a space-based telescope. Off-the-shelf software representing the industry standards for thermal, structural, and optical analysis were employed. The integrated thermal/structural/optical models were collected and tasked with finding an optimum design using yet another off-the-shelf program. Using this integrated tool, the minimum mass thermal/structural design was found that directly satisfied optical performance requirements without relying on derived requirements such as isothermality and mechanical stability. Overdesign was therefore avoided, and engineering productivity was greatly improved. This ambitious project was intended to be a pathfinder for integrated design activities. Therefore, difficulties and lessons learned are presented, along with recommendations for future investigations.

Productivity bottlenecks for integrated thermal, structural, and optical design activities were identified and systematically eliminated, making possible automated exchange of design information between different engineering specialties. The problems with prior approaches are summarized, then the implementation of the corresponding solutions is documented. Although the goal of this project was the automated evaluation of coupled thermal/optical/structural designs, significant process improvements were achieved for subset activities such as stand-alone thermal, thermal/ structural, and structural/optical design analysis.