FAQ5 CALIBRATION AND NORMALISATION

For most practical ECT sensors, there is not a simple linear relationship between the capacitances measured between the electrodes and the permittivity of the material inside the sensor. The relatively large number of different measurements required and the fact that the relationship between capacitance and permittivity may be different for each of these measurements, creates potential calibration and operating problems for ECT systems.

The method which is commonly used to overcome these problems is to restrict the use of ECT to the case where the sensor contains mixtures of two materials of differing permittivities and to operate the ECT system between the range of permittivities of these two materials.

This is done by calibrating the sensor before any measurements are commenced and involves first filling the sensor with the lower permittivity material and measuring all of the inter-electrode capacitances and then repeating this operation with the higher permittivity material. All subesequent capacitance measurents are then referenced (or normalised) to the values measured at calibration. For example, all of the capacitances have normalised values zero when the sensor contains the lower permittivity material and one when the sensor is filled with the higher permittivity material. For all other conditions, the capacitances will have values which nominally lie between these two measurement limits. The image pixel values are also normalised in a similar manner so that they have the values zero and one when the sensor contains the lower and higher permittivity materials respectively. Further details sabout normalisation and calibration are given in subsequent FAQs (see below).

When a mixture of 2 dielectric materials is to be imaged, ECT systems are normally calibrated by measuring two reference sets of inter-electrode capacitances, CL and CH with the sensor filled with the lower and higher permittivity materials in turn. These values are then used to normalise the subequent capacitance measurements as described in the following FAQ. Back to top

Once the sensor has been calibrated, all subsequent capacitance values CM are normalised to have values CN between zero (when the sensor is filled with the lower permittivity material) and 1 (when filled with the higher permittivity material) by applying the formula:

CN = (CM - CL) / (CH - CL)

The pixel values in the permittivity image are normalised in a similar way to the capacitances, so that they have the value 0 when the sensor is filled with the lower permittivity material and 1 when the sensor is filled with the higher permittivity material. Back to top

Last updated 18-06-2002

Copyright 2001 PTL