Experimental Measurement of Multiple Thermal Properties by Error Minimization

Common thin film thermometry techniques are usually based on transient heat diffusion within a sample and its surroundings and are therefore sensitive to the film’s thermal conductivity (k) and heat capacity (C). This presents a problem of under-constraint in the numerical fitting models when both k and C of a given film are unknown. A number of approaches and assumptions have been studied to eliminate this dual dependence or estimate C analytically. However, they often amount to little more than fitting parameters, experimental assumptions, and rough estimates for many composite and polymer films that are emerging in the microelectronics and MEMS industries. The effect that the uncertainty in one property has on the prediction of the other is discussed in the framework of the polymer film PVDF used in many microsensor and actuator applications. An error surface analysis is used to describe the link between assumption and prediction for thermoreflectance and temperature phase measurement techniques. A methodology is presented that combines the results of two thermal tests through an error minimization algorithm to solve for both k and C with no analytical assumptions or approximations. This approach is demonstrated with an experimental test case, validated with synthesized data, and generalized to any system variable and a multitude of thin film thermometry variable or thin film thermometry technique.