BEST METHOD FOR DETERMINATION OF THERMAL RESISTANCE OVER HIGH THERMAL CONDUCTIVE

The experimental techniques for thermal characterization of high thermal conductivity dielectrics have developed along last years. Different test methods were standardized for international recognized entities such as ASTM.

EXPERIMENTAL TECNIQUES

The experimental techniques for thermal characterization of high thermal conductivity dielectric can be classified as follows:

a) steady-state techniques allowing direct evaluation of thermal conductivity (D5470),

b) pulsed time domain methods, which deduce the diffusivity from the time required for heat pulse to propagate through a section of the sample (E1461).

c) thermal wave methods

METHOD DESCRIPTION

D5470: This state technique is based on heat conduction between two parallel, isothermal surfaces separated by a test specimen of uniform thickness. The thermal gradient imposed on the specimen causes the heat flow. Apparent thermal conductivity is directly obtained from this data and the thickness of the specimen. Fig 1

E1461: A small, disc specimen is subjected to a high intensity duration radiant energy. The energy of the pulse is absorbed on the front surface of the specimen and the resulting rear face temperature rise (thermal curve) is recorded. The thermal diffusivity value is calculated from the specimen thickness and the time required for the rear face temperature rise to reach a percentage of its maximum value. Fig 2

GENERAL CONSIDERATIONS

Thin-film geometry, microcrystalline or amorphous structure of thin films, and the large number of potential defects due to the microfabrication process lead to inhomogeneity and anisotropic physical properties. As a result, thermal properties of materials in thin-film form differ strongly for those bulk materials. Time scale techniques are influenced due to the thermal transport properties are geometry and homogeneity strongly dependent.

Steady state measurements like D5470, usually yield the highest level of accuracy, on the order +/-5-10%, meanwhile time domain techniques such as E1461, the relative uncertainties are on the order of 15-20%, and at time even larger errors.

Thermal conductivity is a thermal dependent parameter. It is important to consider this dependency because of E1461 method is based on 1-6mm thickness samples measured at 400-600ºC.