Application of an Inverse Design Method to Meet a Target Pressure in Axial-Flow Compressors
Numerous methods have been developed to design axial-flow compressor blades. These methods are generally categorized into inverse or direct approaches. In the inverse design methods, a distribution of an aerodynamic parameter such as pressure or velocity on the blade surfaces is given, and the target blade geometry that can provide the corresponding distribution is to be determined. In the present work, a novel inverse design algorithm called Ball Spine Algorithm (BSA) is developed to design an axial-flow compressor on the blade to blade surface. In the BSA, the blade surfaces are considered as a set of virtual balls that move freely along the specified directions, called `spines'. At first, initial blade geometry is guessed and the blade-to-blade flow field is analyzed by an in-house inviscid flow solver based on the Roe scheme. Comparing the computed pressure distribution (CPD) on the blade surfaces with the target pressure distribution (TPD), gives a guideline in a differential movement for the balls to obtain a modified geometry. For the flow field analysis on the modified geometry, new grids are generated by a combined algebraic-elliptic code. The sequence is repeated until the target pressure is reached. For validation, the approach is applied on an arbitrary blade profile.