This paper presents a detailed analysis that was performed for the design of a “uniform” temperature boundary condition imposed on a boundary of a three-dimensional cubical experimental apparatus for benchmark natural convection heat transfer study. The three-dimensional experimental apparatus was constructed with plates which were assembled to act as boundary conditions to the enclosure walls. Test measurements revealed that temperature non-uniformity along one of the plates (boundary) was significant enough that the benchmark study could not be carried out to the desired accuracy of about 1% error. A subsequent detailed mathematical analysis revealed that the temperature non-uniformity on the plate was a result of the effect of thermal spreading/constriction resistance. Modifications to the original design of the apparatus were made to reduce the temperature non-uniformity on the plate by adding a heat source around the plate where the uniform temperature setting was desired. Before the addition of this heat source, a careful mathematical analysis shows a significant reduction in temperature non-uniformity from about 4% (based on the initial design) to less than 1% (for the modified design). By examining the temperature difference between two locations on the plate, the predicted temperature difference obtained through mathematical analyses show excellent agreement with the measured temperature difference.

This content is only available via PDF.
You do not currently have access to this content.