Experimental study of optimum spacing problem in the cooling of simulated electronic package S. Chen, Y. Liu, S. F. Chan, C. W. Leung, T. L. Chan Abstract An experimental investigation has been performed to determine the effects of different arrange- ments of obstacles on the cooling of simulated electronic package. The considered simulated electronic package consisted of a channel formed by two parallel plates. The bottom plate is attached with ®ve identical electrically heated square obstacles, which are perpendicular to the mean air¯ow and arranged with different side-to-side distances. The experimental results show that the conventional equi-spaced arrangement might not be the optimum option and should be avoided. A better thermal performance could be obtained when the side-to-side distances between the obstacles followed a geometric series. For example, at Re  800, the highest temperature of the optimum arrangement could be reduced by 12% compare to the equi-spaced arrangement and the maxi- mum temperature difference among the ®ve obstacles is lower than that of equi-spaced arrangement by 32.1%. List of symbols A surface area i.e. top and two sides) of each obstacle exposed to air, m 2 B obstacle height above the base of the simulated electronic package, m C clearance between the upper surfaces of the ribs and the ceiling of the channel, C  H  B E electric power supplied to each obstacle, W F view factor for thermal radiation from an obstacle to its surroundings h convective coef®cient for heat transfer from an obstacle to the air¯ow, W/m 2 á K) H vertical free height within the unencumbered duct, m k thermal conductivity of the ¯uid, W/m á K) Nu Nusselt number Pr Prandtl number, v/a Q steady-state rate of heat loss from the simulated electronic package, W Re Reynolds number, UC/v T temperature, K U mean air speed in the duct, m/s a thermal diffusivity e mean surface emissivity with respect to thermal radiation m kinematic viscosity of the ¯uid, m 2 /s r Stefan±Boltzmann constant, W/m 2 K 4 ) Subscripts a air¯ow ave average value c forced convection i ambient environment l conduction n number of obstacle r via thermal radiation s the obstacle's surface 1 Introduction The electronic devices miniaturization is characterized by high rate of heat dissipation per unit of component area and it demands the electronic system has enough heat dissipation capability in order to avoid its temperature from rising signi®cantly, which may lead to malfunction and breakdown of the entire device. Improvements of cooling techniques are crucial to meet the design and development of rather complex circuits and dense electronic boards. Fluid ¯ow around obstacles mounted on a channel wall forms a fundamental basis for studies of the cooling of electrical devices. Mixed convection in horizontal channels is of interest in cooling of local heat sources encountered in electronic devices. Investigations of the cooling of electronic components have been widely studied both numerically and experimentally Incropera, [1]). Kennedy and Zebib [2] reported mixed convection between horizontal parallel plates with a local heat source ¯ash-mounted on the horizontal plate. They showed the heat transfer characteristics and ¯ow pattern resulting from four cases of heat source locations. Incropera et al. [3±5] has performed a series of numerical and experi- mental studies in a rectangular horizontal channel. Mixed convection in the entrance region was extensively discussed, and they visualized four ¯ow regimes along the bottom plate-laminar, mixed, transitional, and turbulent regimes. Consequently, signi®cant improvement in heat transfer was found to be due to the buoyancy-driven Heat and Mass Transfer 37 2001) 251±257 Ó Springer-Verlag 2001 251 Received on 17 December 1999 S. Chen, Y. Liu &), S. F. Chan, C. W. Leung, T. L. Chan Department of Mechanical Engineering The Hong Kong Polytechnic University Hung Hom, Kowloon, Hong Kong Support given by The Hong Kong Polytechnic University under Central Research Grant No. G-YB87 is gratefully acknowledged.