ABSTRACT
This paper presents a new way to reduce data in the flash method and the step method of measuring thermal diffusivity. Experimental temperature vs. time data are first properly periodized, then transformed by using the Discrete Fourier Transform (DFT), and the real part of the second term of the image temperature is then fitted with a theoretical formula derived in this paper. The main advantage of this procedure is that the thermal diffusivity calculation actually does not depend on the temperature level before the flash, and can be applied also in the case when the measured temperature signal of the sample is superimposed with an arbitrary linearly rising or falling signal, with no need to know the parameters of this imposed signal. The second advantage is that the procedure allows the treatment of perturbed signals, even in the case when the data are partly statistically correlated, i.e. the noise is not Gaussian. The data reduction procedure has been tested for a correction of the effect of an electronic noise imposed on the temperature vs. time signal in the flash, as well as in the step method. The results show that the reproducibility of our procedure is good, and the accuracy is comparable with other data reduction methods.
