Feature Based Reverse Engineering of Compressor Blades

Reverse engineering has progressed in significant ways in the decades since Pierre Bezier first started with that is arguably the first modern reverse engineering process in industry. Advances in computer processing power, interactive graphics and image based scanning have changed the landscape for reverse engineering. Modelling philosophies have also progressed from reconstructing surface patches to full blown parametric feature based reverse engineering. This paper will investigate the latest modelling approach on a compressor impellor. This has become almost a de facto case study for feature based reverse engineering software. Small features on parts, such as the tip radius of a compressor impeller blade, can be significant for the performance of the product. This paper investigates the reverse engineering of such a blade for CFD analysis. The blade consists of three functional surfaces, the pressure surface, suction surface and tip surface. The tip surface is conical and tangential with the other two surfaces. In this case the tip radius was less than 1 mm, making it difficult to scan points on its surface and separating it from points on the pressure and suction surfaces. Instead, a constrained optimisation strategy was employed to simultaneously approximate the surfaces as features and find the tip radius without actual measurements on it. The strategy is presented here as well as some results on the accuracy of the measurements. It was found that the tip surface could be modelled in this way.