Simulation Based Design Optimization for Microelectronics Packaging Product.

摘要

Reduction in size of portable products such as cellular phones and camcorders has led to the miniaturization of integrated circuit packages. Fine-pitch BGA (fpBGA) packages have been gaining popularity due to being compact in size and relatively low cost. The fpBGA is essentially a smaller version of the BGA package in which the pitch, bond-pads and balls are reduced in size. With further down-sizing in package height, reliability issues like thermal warpage and solder joint fatigue could be big challenges during the production engineering design process.;Based on the root cause analyses from observed failures of microelectronics during different life cycles, it is found that thermo-mechanical (thermal, mechanical and thermo-mechanical) related failures account for about 65% of total failures in microelectronics. It is also clear that most of the thermo-mechanical reliability problems originate from the product/process design phase. However, within the electronics industry, microelectronics design and qualification still largely depend on the designer's experience, or trial-and-error method. Functional fulfillment and integration are seen as the only concerns during the original production prototyping stage. Quality, robustness and reliability are usually dealt with after physical prototyping, wherein reliability qualification testing with duration of several months is no exception. This experience-based design and qualification method cannot lead to competitive products with shorter-time-to-market, optimized performance, low costs, and guaranteed quality, robustness and reliability.;In this study, a relatively new and popular package, fine pitch ball grid array package (fpBGA), is adopted for demonstration. For the sake of enhancing the thermal-mechanical reliability performance of fpBGA, a systematic methodology of packaging design and optimization based on computational prototyping is proposed. Thermal induced package warpage and solder ball fatigue failure are investigated and characterized because they are considered to be the major failure mechanisms during BGA manufacturing, assembling and testing process.;After carefully characterizing the packaging materials properties needed in simulation by instruments, 3D non-linear finite element models including appropriate information of geometrical profile and loading conditions are constructed to predict the thermal-mechanical behavior of fpBGA during different production and testing processes. Real samples are fabricated and subjected to shadow Moire inspection and Thermal Cycle Test (TCT) to verify these finite element models. After selection of six design variables or control factors, a screen experiment process is conducted to determine the key factors which have a significant impact on the outputs (package warpage after molding and solder joint fatigue life) based on these verified simulation models.;From the sensitivity analysis results, it can be found that the thickness and CTE (coefficient of thermal expansion) of the molding compound and the thickness of substrate play the most important role on both package warpage and solder joint fatigue life. Then a response surface methodology (RSM) is applied to establish explicit regression models for the warpage value and thermal fatigue life in terms of the selected key design factors with a well-established design of experiment (DOE) scheme. By employing the gradient-based numerical techniques, required optimal designs searching process can be performed based on the well constructed and verified surrogate models (RSM). The design task is to identify the optimal fpBGA design specification by varying several package input parameters so that the maximum fatigue life of the solder joints can be achieved and, at the same time, satisfying design requirements and package warpage performance criteria. Finally, the optimal design parameter set is determined according to the outputs of simulation based design optimization process.