Proud Liberal
Quite Involved in Discussions
I'm measuring axial and radial surface finish (16 Rz) on a broached internal radius. The internal geometry is somewhat complex (one-way clutch mechanism) and the broach tool can leave grooves in the surface (see picture). Due to the complex geometry, the surface needs to "nibbled" in sections (vs the entire pattern cut with a single tooth in the broach). This means its extremely difficult to get a micron level tooth alignment from one section to another (thus the grooves).
Consequently, it's easy to run up against the tolerance limits if I measure the "worst" sections of the profile. Per ISO4288 Annex A, there exists a 16% rule with a series of inspections that appears to provide a loop-hole for accepting small areas outside the surface finish spec. My understanding of the rule and its application is limited to what is published in the Mitutoyo "QUICK GUIDE TO SURFACE ROUGHNESS MEASUREMENT" as I do not own a copy of ISO4288.
Does the 16% rule assume something about the homogeneity of the surface (turned, ground, etc vs broached)? There isn't anything that addresses the width of a "non-conforming" section permitted. Would that be limited to a sampling length? Or an evaluation length?
I'm in need of some expert advise as I feel unqualified at this level of interpretation of the ISO specification.
Consequently, it's easy to run up against the tolerance limits if I measure the "worst" sections of the profile. Per ISO4288 Annex A, there exists a 16% rule with a series of inspections that appears to provide a loop-hole for accepting small areas outside the surface finish spec. My understanding of the rule and its application is limited to what is published in the Mitutoyo "QUICK GUIDE TO SURFACE ROUGHNESS MEASUREMENT" as I do not own a copy of ISO4288.
Does the 16% rule assume something about the homogeneity of the surface (turned, ground, etc vs broached)? There isn't anything that addresses the width of a "non-conforming" section permitted. Would that be limited to a sampling length? Or an evaluation length?
I'm in need of some expert advise as I feel unqualified at this level of interpretation of the ISO specification.
