Filtering effects on involute gear form measurement

Traditional gear measurement results are used to classify a gear tolerance grade in accordance with ISO 1328-1:2013 but as more stringent requirements are placed on gear performance it is important to also characterise performance related features properly on the gear flank surface. However, the measurement uncertainty of the measurement machines is not improving at the same rate as the tolerance requirements, and it is not uncommon to have less than 10 mu m tolerances to be measured with a 1-3 mu m measurement uncertainty capability. This breaks the 'rule of thumb' that the uncertainty should be smaller than one tenth of the value you are trying to measure. It is therefore important to review the measurement and evaluation process to identify if there are any areas for improvement within our current methods. Additionally, the measurements are being used to refine gear tooth contact analysis performance models and the development of optical and high speed tactile scanning measurement methods is allowing the full 3D gear flank surface measurement a feasible option on the shop floor. Greater understanding of the involute flank beyond the standard evaluation limits is thus required. These additional requirements have prompted this review of the effect that the Gaussian filters specified in ISO 1328-1:2013 have and that their characteristics are fully understood by those who must perform and interpret the measurement results. This paper examines the effect that these filters have on measurements where micro geometry corrections, such as tip relief, or tooth tip and end face chamfers are present. Methods of minimising these effects are reviewed with specific reference to strategies defined in ISO 16610-28:2016. Recommendations are offered to minimise the effects when evaluating the gear tolerance class and quantifying gear flank micro geometry.