Further improvements to the diffusive heat pulse technique

The diffusive heat pulse technique is used for the simultaneous determination of specific heat capacity (C) and thermal conductivity (κ), from a single measurement on a single sample. As such, accurate analysis is a very important part of the experimental process, and any distortions of the sample response should be accounted for in the determination of C and κ. In this paper we describe a powerful computational fitting technique for use with this type of data, based on non-linear least squares fitting using a "Levenberg-Marquardt" minimisation scheme. We give the information necessary to use this fitting method in the presence of the three distorting effects we have encountered; clamp boundary resistance, finite heat pulse duration and instrumentation time constant. We have found this new technique to be highly successful, even in situations where all three forms of distortion are significant.