A knowledge of the dynamic behavior of a gas turbine has always been necessary for control system design and development. One of the chief problems is to measure this information from engine tests as comprehensively and accurately as possible in the minimum amount of time. A technique, based on an improved and generalized version of the pseudo-random binary noise (PRBN) method, is described which shows significant improvements compared to other methods of dynamic response testing. The technique involves the injection of a small PRBN or other random disturbance into the fuel flow or variable geometry actuator, the recording of the response of other engine parameters to that disturbance and the subsequent use of a computer to derive the frequency response. The requirements necessary for successful dynamic response testing are discussed and a comparison made between the improved PRBN technique and conventional sinewave testing from actual engine tests. A number of engines have been analyzed using the new method, some of the results from which are presented.

This content is only available via PDF.
You do not currently have access to this content.