The procedure for the estimation of heat losses from a composite panel that must be known in order to optimize repairs is described. Many papers for estimating parameters describe either simulated experiments or experiments conducted under very controlled situations and the results based upon models of the experiment are usually favorable [1,2]. In contrast when experiments are conducted in the field or under adverse conditions, even when the system is a simple one and amenable to modeling, the estimation can be unexpectedly difficult. The paper describes experiments to estimate the heat losses from the back side of a composite panel undergoing repairs and the modeling of the system. Although the temperatures based on the estimated parameters successfully match the measured steady state temperatures, the model does not accurately predict the transient behavior, raising questions about its ability to determine the magnitude of the observed heat losses and, more importantly, what must be done to improve the model.

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