EVALUATION OF VIRTUAL SOURCE BEAM CONFIGURATIONS FOR RAPID TOMOGRAPHIC RECONSTRUCTION OF GAS AND VAPOR CONCENTRATIONS IN WORKPLACES

Citation
Dy. Park et al., EVALUATION OF VIRTUAL SOURCE BEAM CONFIGURATIONS FOR RAPID TOMOGRAPHIC RECONSTRUCTION OF GAS AND VAPOR CONCENTRATIONS IN WORKPLACES, Journal of the Air & Waste Management Association [1995], 47(5), 1997, pp. 582-591
Citations number
13
Categorie Soggetti
Environmental Sciences","Metereology & Atmospheric Sciences","Engineering, Environmental
Volume
47
Issue
5
Year of publication
1997
Pages
582 - 591
Database
ISI
SICI code
Abstract
Beam path average data from an Open Path Fourier Transform Infrared (O P-FTIR) spectrometer can be used to reconstruct two-dimensional concen tration maps of the gas and vapor contaminants in workplaces and the e nvironment using computed tomographic (CT) techniques. However, a prac tical limitation arises because in the past, multiple-source and detec tor units were required to produce a sufficient number of intersecting beam paths in order to reconstruct concentration maps. Such a system can be applied to actual field monitoring situations only with great e xpense and difficulty. A single monostatic OP-FTIR system capable of r apid beam movement can eliminate this deficiency. Instead of many sour ce and detector units, a virtual source arrangement has been proposed using a number of flat mirrors and retroreflectors to obtain intersect ing folded beam paths. Three virtual source beam configurations genera ted for a single-beam steerable FTIR system were tested using 54 flat mirrors and four retroreflectors or 54 flat mirrors and 56 retroreflec tors mounted along the perimeter walls of a typical 24- x 21-ft test r oom. The virtual source CT configurations were numerically evaluated u sing concentration maps created from tracer gas concentration distribu tions measured experimentally in a test chamber Synthetic beam path in tegral data were calculated from the test maps and beam configurations . Computer simulations of different beam configurations were used to d etermine the effects of beam geometry. The effects of noise and peak-r educing artifacts were evaluated. The performance of the tomographic r econstruction strategy was tested as a function of concentration and c oncentration gradients.