An inverse design method for optimizing design parameters of heat sink modules with encapsulated chip

A three-dimensional inverse design problem in estimating the design variables for heat sink modules with an encapsulated chip is solved in the present study by using the Levenberg–Marquardt Method (LMM) and the general purpose commercial code CFD-ACE+ in an irregular domain. Three different types of heat sinks are examined at a fixed fin array volume to determine the most efficient type of heat sink. Moreover, Aluminum and Copper heat sinks are compared to find the optimum design of the module. Results obtained by using the LMM to solve this 3-D inverse design problem are justified based on the numerical experiments and it is concluded that the double row plate fin type heat sink performs best since it can obtain the lowest temperature distribution on the bottom surface of heat sink module. Moreover, larger heat transfer area of heat sink does not guarantee better thermal performance. Due to higher thermal conductivity of Copper heat sink, it also has better thermal performance than the Aluminum heat sink.

► We model three heat sink modules with an encapsulated chip and having same volume.
► The Levenberg–Marquardt Method is then applied to determine the optimal shape for each heat sink module.
► Results show that Type B heat sink always has best performance at a fixed fin array volume.
► Larger heat transfer area does not guarantee better thermal performance for heat sink modules.
► Due to higher thermal conductivity of Copper heat sink, it also has better performance than the Aluminum heat sink.