2½D Temperature Mapping: The Tomographic “Blanket”

Sergio Garcia C., Krikor B. Ozanyan

Department of Electrical Engineering and Electronics, UMIST, PO Box 88, Sackville Street, Manchester, M60 1QD, UK

ABSTRACT

A tomographic method and implementation is described for imaging of the temperature distribution in a non-planar surface based on direct-current resistance measurements. The temperature field is reconstructed from measurements of the temperature-induced resistance changes in a grid of non-interacting chains of sensors forming the sensor head. Resistance information is collected at the periphery of the grid to obtain the temperature distribution. In the simplest sensor head configuration the chains are crossing without contact in a 10 rows x 10 columns arrangement, which may be non-planar and can be made to assume a variety of shapes. Each chain is sequentially multiplexed, to a DC bridge measurement circuit with a software programmable gain amplifier and a low pass filter. This process is handled by a microcontroller, which also routes the time-multiplexed output signal for subsequent digitisation and image reconstruction. Iteratively reconstructed cross-sections of heated air flow through a square sensor head are shown. The estimated accuracy of the system in terms of resistance is around ±0.02%.

An interesting aspect of the reported work, namely the demonstration of the principle of Guided-Path Tomography, is also discussed; by confining the electric field along well defined routes, it is possible to obtain a forward transformed object by DC electrical measurements, and subsequently reconstruct the object by a typical hard-field inversion technique, such as ART.

Keywords Sensor head, hard-field, temperature distribution.