Pose Estimation and Damage Characterization of Turbine Blades during Inspection Cycles and Component-Protective...

Area of Science:

• Mechanical Engineering
• Quality Assurance
• Robotics

Background:

• Automated inspection of confined spaces, like aero engines, is difficult.
• Assessing turbine blade condition and damage is crucial for engine maintenance.
• Current methods lack automation and precision for in-situ inspection.

Purpose of the Study:

• To demonstrate a high-precision inspection approach for movable turbine blades in confined spaces.
• To automate the assessment of turbine blade condition and damage.
• To assist engine disassembly decisions by providing metric geometry deviations.

Main Methods:

• Borescopic inspection using point cloud data.
• Feature extraction of film-cooling holes for point cloud pre-alignment.
• RANSAC-based feature matching and multi-stage registration.
• Damage assessment based on geometric deviations.

Main Results:

• Successful high-precision pose estimation of turbine blades.
• Automated identification and alignment of turbine blade features.
• Metric geometry deviations derived for damage assessment.
• Fast inspection of blade roots for disassembly damage.

Conclusions:

• The proposed borescopic inspection approach enables precise and automated quality assurance in confined spaces.
• This method aids in making informed engine disassembly decisions and preventing further damage.
• The technique is applicable to the inspection of movable turbine blades and blade roots.