Engine Distress Detection in Gas Turbines With Electrostatic Sensors

Abstract

Failures in gas turbines such as fretting at combustor assembly interfaces, blade rub, Thermal Barrier Coating (TBC) spalling, minor amounts of domestic or foreign object damage can be detected by mechanical vibration or gas turbine performance degradation, but it is usually too late for damage control by the time the failure is significant enough to be detected with these methods. Electrostatic charge sensors present a potential method for identifying the failure modes at an earlier stage before significant damage has occurred. In a gas turbine, there are potentially two sources of electrostatic charge in the exhaust gas flow stream: ionized plasma that is a natural byproduct of high temperature combustion, and any form of debris that has originated either in the compressor, combustor, or turbine sections of the gas turbine engine as a result of vibration or fatigue. For this reason, the electrostatic charge monitor becomes a very useful device for monitoring both combustion performance problems as well as potential damage related to debris in the exhaust stream. Electrostatic sensing technology has been proven to work in detecting ingested debris and engine debris on aerospace jet engines. However, the use of the sensors in industrial applications has shown that much research is still required especially in the areas of sensor placement and failure identification. This paper discusses results from testing conducted to identify the optimal placement location for the electrostatic charge sensors in a gas turbine exhaust stream. The results are presented for various sensor locations on small (112 kW) and medium (1.185 MW) frame gas turbines to evaluate distance, velocity, radial location, and gas turbine geometry effects. These tests are completed with the gas turbine ingesting varying amounts of TBC upstream of the compressor and administering power level changes to the gas turbine. The results of this experimental program demonstrate a clear sensitivity to sensor placement along the exhaust duct of a gas turbine as well as the radial location. There are variations in the particle flow pathlines in the exhaust duct at different gas turbine operating conditions. These variations influence the sensors response. Best results are obtained when the sensors are placed at the location with the fastest and hottest exhaust gas. Multiple sensors may be required to obtain comprehensive coverage for practical event detection.