Nondestructive relaxation spectrometry of dielectrics

Determination of relaxation and retardation spectrum by inverse functional filtering

“…Our experience [12,13,17,18] has shown that the best results give the functional filters designed by the identification method [13]. The name of the method comes from the term 'system identification' as a technique for obtaining a mathematical model of a system from the behaviour of a system (output function) and external influence to a system (input function).…”

“…Restriction (17) does not allow MAF to achieve the maximum performance. Thus, input window ranges d xMAF of MAF are also restricted to be the discrete values: MAF have been derived to recover mean values of the relaxation spectrum F( m ) ln q L .…”

Digital Emulation of Dielectric Relaxation Functions for Capacitive Sensors of Non-Destructive Dielectric Spectrometry

“…We have developed several NDT dielectric spectrometers for solving various quality testing problems based on the ideology [2] that a lot of parameters of users' interest correlate with the dielectric characteristics, and, so many quality testing problems can be solved through NDT dielectric spectrometric measurements. Our spectrometers are classical frequency-domain instruments working over relatively low frequency band 10 Hz-1 MHz with exciting test object with electrical field generated by sinusoidal voltage signals at discrete frequencies, usually altered geometrically with the common ratio 2.…”

“…Dielectric measurements [1] have long been of increasing interest for the characterization of non-metallic materials in many areas of science and technology, such as material science, electronics, biology, medicine, agriculture, food processing industry, etc. The potentialities of the dielectric method were significantly increased with developing one-side access (OSA) capacitive sensors [2] allowing non-destructive testing (NDT) and the characterization of materials without preparation of specimens.…”

Digital estimators of relaxation spectra

“…operating with equally spaced data on a logarithmic scale, which manifest as the data spaced according to geometric progression u n = u 0 q n on a linear scale (logarithmic sampling [5][6][7][8][9][10]). For the logarithmically sampled data, Eq.…”