Optimising laser absorption tomography beam arrays for imaging chemical species in gas turbine engine exhaust plumes

This paper will describe the application of resolution matrices to the design of an optimised 126 beam, 6m absorption tomography array for imaging concentrations of CO2 in the exhaust plume of a Rolls-Royce Trent 1000 gas turbine engine. The resolution matrix will be used to define a fitness value, which is a function the beam configuration, and is minimised by the optimal beam arrangement. Constraints ensure that the optimised beam arrangement can be implemented in a real tomography system. Genetic algorithms are used to determine the optimal array design from the large problem set. Results for image reconstructions of a quasi-realistic phantom of the exhaust plume for each of the array designs are presented with indications of the reconstruction errors. From the results, conclusions are drawn on the suitability of applying resolution matrices to the design of beam arrays for real limited-data tomographic systems.