Exploiting double-flank roll testing spur gear measurements to determine gear parameter deviations through numerical simulation of free-form meshing gears

Even though double-flank testing has been one of the most common inspection methods for gears, there is as yet no method to correlate the measured radial composite errors with the geometrical deviations of the gear tooth flanks that produce them. In order to understand the relation of the two, a numerical model is developed to simulate the double-flank inspection process that is valid for gears with any combination of geometrical profile deviations. The results of the model are in line with the theoretical and experimental findings in the literature and are further validated with original experimental data. Finally, a novel inverse numerical model is proposed that is able to derive the geometrical parameters of the test gear from the variation of the composite error as obtained by double-flank inspection using a genetic algorithm. The model can predict the deviation of the gear geometrical parameters with high accuracy even in the presence of measurement errors and profile deviations from the nominal involute.