Abstract
Experiments performed by Arya and Keyhani (1990) measured the temperature of 12 vertical heated rods within a constant temperature, internally finned cylindrical enclosure. Measurements were performed with air and helium in the enclosure for ranges of rod heat generation rate and gas pressure. In the current work, steady three-dimensional computational fluid dynamics simulations of conduction, natural convection and radiation heat transfer within the experiment were conducted to benchmark the simulation techniques. In the computational model, different thermal conductivities were applied to a spacer plate between a plate that held the heaters, and one of the enclosure endplates. This was done to model a range of contact resistance between the plates. This was necessary because the experimental endplate conditions were not completely documented. The calculations accurately reproduced the local and average temperatures when a high contact resistance was modelled. These results emphasise that conditions far from data measurement locations can affect experimental results. Those conditions must be well documented if they are to be used to benchmark computational methods.