Dielectric Property Retrieval with Open-Ended Coaxial Probe for Solid Materials

Aydinalp C., Joof S., Akıncı M. N., Akduman İ., Yilmaz T.

IEEE USNC-URSI Radio Science Meeting / Joint IEEE Antennas-and-Propagation-Society (AP-S) International Symposium, Colorado, United States Of America, 10 - 15 July 2022, pp.13-14 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.23919/usnc-ursi52669.2022.9887510
  • City: Colorado
  • Country: United States Of America
  • Page Numbers: pp.13-14
  • Istanbul Technical University Affiliated: Yes

Abstract

Determining the dielectric properties of materials based on their microwave features is an important research topic in various disciplines and industries. Accurate retrieval of solid material dielectric properties is one of the challenges in non-destructive measurement approaches. In this work, the dielectric property of three different flat-surface solid materials (kestamid, delrin and alumina) were retrieved from reflection coefficients through deep learning model from 0.5 to 6 GHz. The deep learning model was designed based on Debye parameters and reflection coefficients computed from the open-ended coaxial probe admittance model. The results were compared with commercially available Speag Dielectric Assessment Kit (DAK) software and the calculated percentage dielectric property differences are 5.5%, 6.8% and 7.5% for kestamid, delrin and alumina, respectively.