Systems of contacting spheres are common in engineering applications where the heat transfer analysis can be quite cumbersome due to the transient behavior and the complex geometrical arrangement. As a result, most of the previous works, in this area, have adopted the porous media approach. However, this approach requires the length scale of the representative cell to be roughly three orders of magnitude larger than the size of the spheres. Constriction resistance relations could be useful in accurately computing the temperature distribution within systems of contacting spheres, however many of the requisite relations are not available. Thus, the objective of this study is to develop these relations. In this study, the transient, three-dimensional conduction equation was solved using a ®nite volume scheme and a non-uniform grid. From the resulting temperature distributions, the steady-state and transient constriction resistance of one-sphere and twosphere systems were computed and correlated. The results also showed, for the ®rst time, the critical parameters below which the transient variations must be considered.List of symbols a constriction alleviation factor, R ss kr c A area, m 2 C speci®c heat, J/kg-K Fo Fourier number, asar 2 c k thermal conductivity, W/m-K M total number of spheres in system q heat¯ux, W/m 2 Q total heat transfer, W r,h radial and angular variable in spherical coordinate, m, rad R thermal constriction resistance, K/W T temperature, K U internal energy, òqCT dV V volume, m 3Greek letters a thermal diffusivity, m 2 /s c ratio of contact radius to sphere radius, r c /r s q density, kg/m 3 s time, s Subscripts b bulk temperature c contact i initial value s sphere, or length scale with r s ss steady state value
IntroductionHeat transfer within systems with contacts are common in engineering applications. These include insulated cables (cf. Liu et al., 1998), composite materials (cf. Lewinski and Kucharski, 1992) and packed-sphere systems (cf. Chan and Tien, 1973). For these systems, the heat transfer analysis can be quite complicated due to the transient behavior and the complex geometrical contact arrangement. For the packed-sphere system, the boundary condition of each sphere as well as the number of conduction path are affected by the number of contacting spheres, the orientation between these contacts, and the size of the contact areas. Furthermore, the air gap between the spheres also affects the heat transfer by introducing convective and radiative effects.Due to these complexities, most of the previous works in the packed-sphere system have adopted the porous media approach. This approach uses lumped-parameter models to either measure or compute the effective thermal conductivity of a representative cell (cf. Hsu et al., 1994). However, this requires the length-scale of the representative cell to be roughly three orders of magnitude larger than the size of the spheres (Kaviany, 1995). Consequently, the temperature solution obtained from the porous media approach may not be suitable for...