Analysis of the Uncertainty in measurements of Polymer Pellets using Microwave Resonant Sensors

The analysis and evaluation of the uncertainty in microwave measurements of some polymer plastic materials in the form of small pellets is presented in this paper. Two different resonant sensors, cavity and planar, operating around 2.45 GHz are used to measure the materials. The presented uncertainty analysis is based on the measured resonant parameters from the sensors and represents a statistical tool capable to generate relevant information such as an adequate number of tests, uncertainty levels, correlation coefficient, covariance matrix, and confidence ellipses, which can be highly useful in the analysis of pellet or grained materials using microwave methods, and for fast and accurate decisions involving materials evaluation. It will be shown that a number of 40 tests for each sample is adequate for a stable uncertainty, and due to the E-field distribution and interaction with the samples, the cavity sensor develops lower uncertainty in resonant frequency compared to the planar circuit, thus, it can be a more reliable sensor for polymer pellet measurements.