Curtis O. Pedersen, William P. Bahnfleth
Bibliographic info:
Building Simulation, Vancouver, Canada, 1989, p. 133-138

Experimental studies during the 1940S concluded that heat loss from slab-on-grade floors is proportional to floor perimeter length. More recent numerical investigations, however, indicate that area and shape are also important parameters. Furthermore, results of three-dimensional modelling differ significantly from those of supposedly equivalent two dimensional analysis. Earth-coupled heat transfer processes are increasingly important contributors to building energy consumption, but continue to be poorly understood by most designers. The twin objectives of the present study were to increase fundamental understanding of threedimensional effects on slab-on-grade heat transfer and to provide validation data for simplified models. Over ninety hourly, annual simulations were performed with a detailed three-dimensional finite difference model. Parametric studies considered effects of geometry, climate, soil properties, and boundary conditions. The results of this study show that floor heat flux depends on the ratio of floor area to floor perimeter length (A/P). A potential manual method employing this scaling relationship is described. This model provides a means for characterizing the degree of area dependence of floor heat transfer as a function of building and environmental parameters. A companion paper reports on the parallel development of an improved slab-on-grade model compatible with transfer function based energy analysis programs such as BLAST.