Accurate model construction of deformed aero-engine blades for remanufacturing

Remanufacturing for aero-engine blades is an essential operation in current aerospace maintenance industry due to the significant cost savings involved. However, the individual geometric deformation in the damaged blades makes it very difficult to repair them automatically due to the lack of a model for the deformed blade. This paper proposes a new method to adaptively construct an accurate deformed blade model for remanufacturing by finding the deformation rule between the nominal blade model and the measured data collected from the deformed blade via scanning. Considering the precision of the scanned data, the flat regions of the scanned data and nominal model (suction face and pressure face) are segmented and extracted first. To achieve accurate matching results, the segmented scanned data are matched with the segmented nominal model by maintaining local rigidity under the optimal ICP (iterative closest point) framework. Then, the nonflat regions of the nominal model (leading edge and trailing edge as well as the damaged region) are stitched with the scanned data at selected fixed locations to construct the deformed blade model. A twisted compressor blade with some tip damage was selected in the case study to demonstrate the effectiveness of the proposed method. A reference deformed blade model acquired by a computational fluid dynamics software tool was utilized for comparison, and machining simulation was carried out to verify the remanufacturing result.