Header menu link for other important links
X
Vibration based damage detection of rotor blades in a gas turbine engine
S. Madhavan, R. Jain, C. Sujatha,
Published in Elsevier Ltd
2014
Volume: 46
   
Pages: 26 - 39
Abstract
This paper describes the problems concerning turbine rotor blade vibration that seriously impact the structural integrity of a developmental aero gas turbine. Experimental determination of vibration characteristics of rotor blades in an engine is very important from fatigue failure considerations. The blades under investigation are fabricated from nickel base super alloy through directionally solidified investment casting process. The blade surfaces are coated with platinum aluminide for oxidation protection. A three dimensional finite element modal analysis on a bladed disk was performed to know the likely blade resonances for a particular design in the speed range of operation. Experiments were conducted to assess vibration characteristics of bladed disk rotor during engine tests. Rotor blade vibrations were measured using non-intrusive stress measurement system, an indirect method of blade vibration measurement utilizing blade tip timing technique. Abnormalities observed in the vibration characteristics of the blade tip timing data measured during engine tests were used to detect the blade damage. Upon disassembly of the engine and subsequent fluorescent penetrant inspection, it was observed that three blades of the rotor assembly were identified to have damaged. These are the blades that exhibited vibration abnormalities as a result of large resonant vibration response while engine tests. Further, fractographic analysis performed on the blades revealed the mechanism of blade failures as fatigue related. The root cause of blade failure is established to be high cycle fatigue from the engine run data history although the blades were put into service for just 6 h of engine operation. © 2014 Elsevier Ltd.
About the journal
JournalData powered by TypesetEngineering Failure Analysis
PublisherData powered by TypesetElsevier Ltd
ISSN13506307