Thermo-mechanical Solutions In Electronic Packaging: Component To System Level

Since the advent of the transistor and integrated circuit, the performance of electronic equipment has increased significantly while footprint of systems at all levels continues to decrease. Recently, the number of transistors on a high end microprocessor has exceeded a billion. All of the above has necessitated considerable improvement in cooling technology and associated reliability. In this era of high heat fluxes, air cooling still remains the primary cooling solution mainly due to its cost and ease of installation and operation. The primary goal of a good thermal design is to ensure that the chip can function at its rated frequency or speed while maintaining the junction temperature within the specified limit. With the focus on good thermal design, mechanical reliability related issues due to the weight of the heat sink on stresses induced on different components of the package also needs to be considered. With all this thermo-mechanical design going into at the first and the second level of the electronics design, leads to a system level problem. Owing to time-to-market requirements, CFD analysis allows to complete thermal optimization long before the product test can be made available bringing
about financial benefits and timely engineering support during product development. In this study, the development of a heat sink tester, analysis of effect of weight of heat sink assembly on mechanical reliability of WB-PBGA package, followed by a system level thermal solution for a telecommunication cabinet is presented. Part-1 of the thesis focuses on a component level problem, "Experimental and Computational Characterization of a Heat Sink Tester". Use of heat sinks as a thermal solution is well documented in the literature. Previous work exists where uncertainty in heat transfer coefficient for the heat sink tester is calculated by detailed uncertainty analysis based on Monte-Carlo simulations [7]. In this study, the objective is to characterize a heat sink tester experimentally and computationally to see how these results correlate. Experimental characterization for commercially available heat sinks is done according to JEDEC JESD 16.1 [23] standards and is compared against the vendor specifications and computational results. To obtain computational results, a CFD tool IcepakTM is used to carry out the thermal analysis. The results thus obtained from experimental characterization, computational analysis and vendor specifications are used for benchmarking the heat sink tester. Part-2 of the thesis addresses a package level mechanical reliability issue, "Effect of Weight of Heat Sink Assembly on Mechanical Reliability of a WB-PBGA package". In this study, a stress analysis of described package is carried out to study the effect of weight of heat sink assembly on the mechanical reliability of the package. A three dimensional finite element model of WB-PBGA package and Printed Wiring Board (PWB) is solved numerically to predict the stresses induced and assess their impact on the mechanical integrity on different components of the package, which is accomplished by using a commercial analysis tool ANSYSTM. Die and C4 interconnect stresses are examined to evaluate package reliability. Stresses induced within the die and C4 interconnect are examined for different heat sink materials and variation of force developed by different heat sink attachments such as clip-on and screw-on types. Finally recommendations are made regarding choice of heat sink material and clip force for overall heat sink assembly design. Finally, part-3 of this study presents thermal solution to a system level problem, "Compact Modeling of a Telecommunication Cabinet". The objective of this study is to present an overview of techniques to minimize the computational time for complex designs such as heat exchangers used in telecommunication cabinets. The discussion herein presents the concepts which lead to developing a compact model of the heat exchanger, reducing the mesh count and thereby the computation time, without compromising the acceptability of the results. Compact modeling, selective meshing, and replacing sub-components with simplified equivalent models all help reduce the overall model size. The model thus developed is compared to a benchmark case without the compact model. Given that the validity of compact models is not generalized, it is expected that this methodology can address this particular class of problems in telecommunications systems. The CFD code FLOTHERMTM by Flomerics is used to carry out the analysis.