Turbine Pyrometry: An Equipment Manufacturer’s View

Abstract

Optical pyrometry is potentially a valuable technique for temperature measurement and control in gas turbines. However the problems associated with obtaining a clean signal interpretable as a metal temperature can be formidable. The difficulties are examined here in the light of more than 18 years’ involvement, by the author’s company, with turbine pyrometry. A number of “ground-rules” are formulated which, it is hoped, will be useful to new and prospective users of the method. Experienced users may find the perspective adopted valuable. Some possibilities arising from recent technical developments are highlighted. If certain “ground-rules” can be followed in design of the pyrometer and its installation then this model is trivial. If the rules are contravened the model rapidly becomes intractable. In fast (blade profiling) applications the finite bandwidth of the pyrometer system must be considered. In slow systems the “one-sided” nature of most optical noise cannot be ignored. Finally provision of an adequate optical signal and maintenance of a clean lens are obvious prerequisites.