Correlative Method for Diagnosing Gas-Turbine Tribological Systems

Area of Science:

• Tribology
• Mechanical Engineering
• Materials Science

Background:

• Oil sampling is a long-standing diagnostic technique for lubricated tribosystems, particularly in gas turbine main-shaft bearings.
• Interpreting wear debris analysis is challenging due to complex power transmission systems and varied test method sensitivity.

Purpose of the Study:

• To develop and validate a correlative model for analyzing oil samples from M601T turboprop engines.
• To establish customized alarm limits for iron concentration based on aluminum and zinc levels.
• To assess engine health by correlating wear debris concentrations with engine condition.

Main Methods:

• Oil samples from M601T turboprop engines were analyzed using optical emission spectrometry.
• A correlative model was developed, incorporating customized alarm limits for iron based on binned aluminum and zinc concentrations.
• Two-way analysis of variance (ANOVA) with interaction analysis and post hoc tests were employed to determine statistical correlations.

Main Results:

• A strong positive correlation was found between iron and aluminum concentrations in the oil samples.
• A weaker, yet statistically significant, positive correlation was observed between iron and zinc concentrations.
• Application of the model to a specific engine identified accelerated wear through deviations from established iron limits, preceding critical damage.

Conclusions:

• The developed correlative model, supported by ANOVA, provides a statistically validated method for assessing engine health in turboprop engines.
• Early detection of accelerated wear is achievable through oil analysis, enabling proactive maintenance and preventing catastrophic failures.
• The study demonstrates the effectiveness of integrating spectrographic oil analysis with statistical modeling for tribosystem diagnostics.