Posted by Nick Miller on May 12, 2003 at 09:13:27:
As part of an ongoing investigation we are carrying out here into the geometrical accuracy of an aero gas turbine transmissions part, we have measured lead errors in a 59-tooth internal involute spline. So far I have results from a sample of three parts. In each instance, when I plot the individual lead errors against tooth number as a polar or 'radar' type graph I get a highly distinctive heart-shaped curve. Can anyone explain, in terms of the manufacturing process, how this arises?