A modeling approach to predict fretting fatigue on highly loaded blade roots

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

A lifing technique for predicting fretting fatigue on highly loaded blade-disk attachments has been developed and calibrated. The approach combines extensive testing on nickel and titanium based alloys using a specially devised multiaxial fretting test machine and an analytical lifing procedure based on finite element contact calculations and multiaxial shakedown fatigue models. In order to reproduce realistic operational conditions and standardize testing conditions a special fretting fatigue testing machine with high temperature testing capabilities was developed. The machine was employed to perform systematic testing under prescribed load and displacement conditions at representative temperatures. Making use of FEA the rig test results were calculated to identify relevant parameters such as friction coefficient, slip conditions and machine compliance. The computation procedure involves the calculation of several major loading cycles until a stabilized response of the structure is achieved. The material response is assumed to be elasto-plastic and a nonlinear friction law (space and time) was applied. From the computed mechanical variables, a several life prediction models are benchmarked to establish their capabilities to predict fretting-fatigue life. Finally, a most promising life estimation procedure was applied to predict life in a real compressor blade-disk attachment. Predicted failure location and number of cycles to failure are compared against engine tests results. The experimental-analytical approach has the potential to predict fretting fatigue risk during the design phase on highly loaded joints as well as estimating the preventive overhaul intervals for parts already in service.

Original languageEnglish
Title of host publicationProceedings of the ASME Turbo Expo 2009
Subtitle of host publicationPower for Land, Sea and Air
Pages777-787
Number of pages11
DOIs
Publication statusPublished - 1 Dec 2009
Externally publishedYes
Event2009 ASME Turbo Expo - Orlando, FL, United States
Duration: 8 Jun 200912 Jun 2009

Publication series

NameProceedings of the ASME Turbo Expo
Volume4

Conference

Conference2009 ASME Turbo Expo
Country/TerritoryUnited States
CityOrlando, FL
Period8/06/0912/06/09

Fingerprint

Dive into the research topics of 'A modeling approach to predict fretting fatigue on highly loaded blade roots'. Together they form a unique fingerprint.

Cite this