Experimental Verification Trials for Fast Spectro-tomography Sensing in Process Reactors

M. Nahvi1, B. S. Hoyle2 and Y. Zhao3

1Department of Electrical and Electronic Engineering, Faculty of Engineering, University of Guilan, Rasht, Iran Email: nahvi@guilan.ac.ir

2,3Institute of Particle Science and Engineering,

School of Process, Environment and Materials Engineering, University of Leeds, UK Email: B.S.Hoyle@leeds.ac.uk

ABSTRACT

The augmentation of process tomographic sensing with spectroscopic data, obtained in the same measurement interval, offers large potential benefits through increased knowledge of the specific trajectory and/or state of a target process, in essence offering: ‘spectro-tomography’. This added spectral dimension can be gained through Electrical Impedance Spectroscopy (EIS), which is able to discriminate between different materials or components. The paper provides a background review of previous research which exploits EIS to create a fast spectral sensing method. This uses a linear frequency modulated (LFM) ‘chirp’ excitation signal, and ‘wavelet’ extraction of spectral data from the detected responses. The method is shown to offer benefits of fast spectral sensing over a wide bandwidth, designed to excite features of interest in the process. The resulting data can be reconstructed classically to create multiple spectral image sets for each tomographic ‘spatial time slice’. The paper focuses on practical and materials aspects of the method through an experimental trial of the monitoring of the crystallization of L-glutamic acid (LGA). This material provides a convenient and safe stimulant for a range of crystallization reagents. The paper offers a review of the relevance of this type of material and its typical processing in stirred mixing vessels, which have featured in many 2D and 3D industrial process tomography studies. The experimental arrangements used in the verification trials are described. Results are presented which contrast experimental data which show spectroscopic characteristics at a variety of temperature values, with those obtained from a spectrum analyzer. The results offer verification of the fast spectroscopic sensing ‘dimension’ in spectro-tomography, and illustrate its potential for the characterisation of a variety of industrial process materials and components.

Keywords: Process spectroscopy, Electrical impedance spectroscopy, Spectro-tomography, Wavelet transform, Chirp waveform.