Numerical heat transfer predictive accuracy for an in-line array of board-mounted PQFP components in forced convection

Numerical predictive accuracy is assessed for component-Printed Circuit Board (PCB) heat transfer in forced convection using a widely-used, Computational Fluid Dynamics (CFD) based software dedicated for the analysis of electronics cooling. This is achieved by comparing numerical predictions with experimental benchmark data for an in-line array of fifteen, equally spaced, PCB-mounted 160-lead Plastic Quad Flat Pack (PQFP) components. Test case complexity is incremented in controlled steps, from a single component, to components individually powered on the fully populated PCB, to a simultaneously powered configuration exhibiting a high degree of component thermal interaction. Benchmark criteria are based on component steady-state junction temperature and component-PCB surface temperature profiles, measured using thermal test chips and infra-red thermography respectively. These measurements were taken with the test vehicle mounted in a wind tunnel. Component numerical modeling is based on nominal package dimensions and material thermal properties. In the absence of a dominant length scale for describing the fluid flow regime in non-dimensional form, the fluid domain is solved using both laminar and turbulent flow models. Component junction temperature prediction accuracy for the fully powered, populated PCB is typically within ±1°C to ±10°C (up to 25%). The full complexity of component thermal interaction is shown not to be fully captured. Neither the laminar or turbulent flow model could resolve the complete flow field, suggesting the need for a turbulence model capable of modeling transition. Overall, component junction temperature prediction accuracy would not be sufficient for the predictions to be used as boundary conditions for subsequent reliability and electrical performance analyses.