Benefits Obtained Using AI Based Test Equipment on Installed Turbine Engines 941172
Since 1986, the U.S. Army has been evaluating a digital data recording system known as a Portable Engine Analyzer Test Set (PEATS) for the detection, diagnosis and prognosis of faults on turbine engines installed in helicopters. On-board use of digital data systems makes test cell quality instrumentation available for testing installed engines. Electronic data recording and transfer of accurate performance information into a personal computer allows the use of Artificial Intelligence methods to form and interpret the significance of the data. The resulting information is being successfully used to provide previously unobtainable diagnostic capabilities for maintenance decisions. Standardized engine testing and formatting of data helps low experience technicians see the relative internal condition of a gas turbine engine at a glance.
The computational capability of the personal computer allows expert logic to be applied in data reduction to yield Referred Engine Diagnostic Data (REDD), Patent 5,018,069, May 21, 1991. REDD compares the engine under test to a normal, installed specification engine. REDD may not detect all engine problems experienced but has shown its ability to identify many hidden faults, thereby enhancing propulsion system management. Field testing has suggested that installed indicating systems which display inaccurate readings cause many maintenance actions (or decisions to take no action) to be made in error.
This paper presents results from field testing of helicopter turbine engines with digital data recording systems by a maintenance contractor. Engine maintenance management based on the demonstrated capability of REDD to provide knowledge of each turbine engine's internal condition has essentially improved operational reliability, maximized flight safety, reduced Remove and Replace (R & R) troubleshooting, while producing an excellent return on investment.