Transient hot strip (THS) method: Uncertainty assessment

The transient hut strip (THS) method can be used to measure simultaneously the thermal conductivity lambda and diffusivity a of dielectrics within a few minutes. However, although the method has been known for 20 years, there is no complete assessment of its uncertainty. First, the underlying complex mathematical model makes any error analysis a tedious and complicated task. Secondly, the ISO Guide to the Expression of Uncertainty in Measurement does not apply directly because of the classical model's implicit character, In the present paper, the combined standard uncertainty II of the THS method has been determined by applying two different models. First, we start fi om the classical nonlinear model. The major sources of errors are analyzed, namely, the ideal model errors, the evaluation errors, and the measurement errors. Next, a newly developed numerical procedure combines all the components in a way that the resultant standard uncertainties of the nonlinear model, u(lambda)/lambda = 2.6% and u(a)/a = 11%, comply as closely as possible with the principles of the ISO Guide. Second, we start from the recently presented linear expression of the THS mathematical model that is briefly discussed. Since this approximation is explicit in both measurands, the uncertainties, u(lambda)/lambda. = 2.5% and u(a)/a = 11%, are determined in full accordance with the ISO guide. Thr uncertainty in thermal conductivity is experimentally assessed against the standard reference CRM 039 (Pyrex). The results obtained are in excellent agreement with the theoretical values.