This work aims at the improvement of measurement accuracy of thermal conductivity and thermal diffusivity using a hot
disk thermal constants analyser. The hot disk technique is based on the transient heating of a double spiral plane
sandwiched between two pieces of investigated material. By researching the temperature change in the sensor surface, it
is possible to deduce both the thermal conductivity and thermal diffusivity of the surrounding material from one single
transient recording, provided the heating power and measuring time are appropriately chosen within the reasonable range
defined by the theory and experimental situation. Based on the engineering application requirement for precision and
efficacy, a new experimental method has been developed for high-accuracy measurement of thermal conductivity and
thermal diffusivity in different experimental conditions. The standardized material Pyroceram 9606, with a thermal
conductivity of 4.05 W/(mK), has been investigated and analyzed using the newly developed method. The measurement
results show that the precision 5% estimated for thermal conductivity and 4% for thermal diffusivity at or around room
temperature and under normal pressure, which indicate that the newly developed method has led to the high-accuracy
measurement of thermal conductivity and diffusivity.
A new computing method is proposed for the measurement of the thermalphysical parameters of specimens with
nonuniform temperature profile distributions. The calculation method is derived from the temperature dependence of the
thermal properties, and can be applied to the measurement of the longitudinal thermal expansion coefficient and
electrical resistivity. The cross section of the entire specimen is uniform at room temperature and the changes during the
experiments are ignored in this method. If the temperatures are measured at equal intervals, the specimen may be
considered as consisting of M equal segments and each of them is S long. The corresponding length and resistance of
these segments at temperature T0 may be of any value. If the changes in length and temperature distribution of the
specimen are measured, the temperature dependence of the thermal expansion coefficient can be worked out using this
method. If the resistance and temperature distribution of the specimen are measured, the electrical resistivity versus
temperature function of the specimen, which is corrected by thermal expansion, can be obtained as well. The validity of
the computing method of thermal expansion coefficient and electrical resistivity is verified through computer simulation.
The maximum 'measurement' error of electrical resistivity is 3.3%.