Laminar convection of horizontal and vertical simulated printed-circuit board assemblies

Experimental investigations have been performed to determine the heat transfer coefficients on both a horizontally and a vertically orientated simulated printed-circuit board (PCB) assembly in a laminar-flow region. The horizontal simulated PCB assembly consisted of two parallel horizontal plates, and the lower one was mounted with uniformly spaced, electrically heated rectangular ribs which were used to simulate the electronic components. The influences of Reynolds number, the ribs'pitch-to-height ratio (H/B), and the width-to-height ratio (L/B) on the convection from the PCB assembly to the ambient air were investigated. The horizontal PCB assembly was turned by 90° to become a vertical PCB assembly and similar investigations were performed. For both horizontal and vertical PCB assemblies, it was found that the variation of the ribs'heat transfer is strongly dependent on the Reynolds number, and the heat transfer coefficient for L/B = 4 is always greater than that for L /B= 3. This difference in thermal performance becomes larger with increase of Reynolds number. Comparing the heat transfer coefficients between the horizontal and vertical PCB assemblies reveals the fact that when L/B = 3, the horizontal orientation is better than its counterpart, but when L/B = 4, a similar result is obtained only for small values of H/B; for larger values of H/B, a converse result is obtained. Consulting two ribs of the same volume, the one with a larger top surface area has a better ability to enhance heat transfer. Correlations used to determine the heat transfer coefficient in terms of H/B, L/B, and Reynolds number were obtained from the experimental results.