Collaborative improvement of profile accuracy and aerodynamic performance in robotic grinding of transonic compressor blade leading edge

Blade machining accuracy is critical to engine service performance. However, the intricate curvature characteristics and exceedingly thin dimensions of the leading edge present formidable challenges in achieving precise grinding. This study introduces a novel robotic belt grinding framework aimed at collaboratively enhancing the profile accuracy and aerodynamic performance of the leading edge. The irregular contact area and pressure distribution of the narrow rigid-flexible grinding interface are calculated based on the Semi-Hertz model. Additionally, a time-varying material removal contour (MRC) model is developed for precision grinding with the pyramid-structured abrasive belt. The non-linear equations governing dwell time are transformed into a convex quadratic optimization problem, and the solution is obtained using the fastest gradient method with constraints. Subsequently, the dwell time vector is converted into a feed rate value for robot programming. Experimental results demonstrate that the size and shape accuracy at three typical sections of the leading edge are effectively controlled within the desired range. The mean value of the line profile error is reduced to 0.019 mm, and the standard deviation of the surface profile error shows a noteworthy 32% improvement compared to the previous method. Furthermore, aerodynamic performance analyses are conducted for the ground airfoils using NUMECA. Numerical simulation results reveal that shape accuracy significantly influences the flow field state within the channel, given the precondition of qualified dimensional error. The isentropic efficiency is increased most by 4.67% by this dwell time control grinding.