Description

Typical turbo-machinery aerofoil profiles have a large variation of curvature and are non-analytical functions by virtue of a round leading edge, where the slope is infinite. Many methods have been devised for geometric modelling of aerofoil shapes such as, by use of discrete coordinates, B-Splines, thickness distribution along camber line etc. Representation of these geometries by discrete coordinates leads to a large number of design variables, especially near the leading edge, and hence, is numerically expensive. B-Splines and polynomial fit methods fail to represent the round leading edge and thus depend on circular arcs to close the profile. Moreover, in above mentioned methods, the desired passage throat area and location becomes difficult to control and is therefore achieved iteratively. The methodology presented in this paper attempts to construct the blade shape by first constructing the desired flow passage using two interpolation curves, each for suction and pressure sides. These are then used to arrive at the preliminary blade aerofoil, which then is manipulated by superposing a thickness function. This thickness function has three control parameters which are used to vary the leading edge roundness, midsection curvature and trailing edge boat-tail angle. The proposed method, unlike the conventional methods, provides a better control over the location of throat in a reaction blade and also gives a smooth control over the local curvature to avert separation and minimization of trailing edge wake. In addition, the method ensures the closure of blade profile without the need of an additional arc.