Browse Prior Art Database

Predicting Process Capability with Variable Geometric Tolerance Specifications

IP.com Disclosure Number: IPCOM000010088D
Publication Date: 2002-Oct-17

Publishing Venue

The IP.com Prior Art Database

Abstract

Current statistical formulas used to predict process capability from continuous data regard all dimensional specifications as having constant tolerance while many of them are variable. Geometric tolerances and datum feature designations followed by the tolerance modification symbols ? or ? permit a variable "bonus" or "datum-shift" tolerance based upon feature size. Process capability predictions from discreet data (measurement with attribute gauges) include the variable portion of tolerance in the gauge design but predictions from continuous data (measurement with variables gauges) ignore the variable portion of tolerance. The disparity betweent he capability prediction methods affect product acceptance and rejection, process precision requirements, gauge preferences, process optimization (targeting feature sizes), and design practices (designation of modifiers and the proportioning of variable tolerance between size and geometric form, orientation, or location). Since most quality control procedures require continuous data from variables gauges for manufacturing process analysis and control the problem is common throughout the manufacturing industry.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 7% of the total text.

Page 1 of 14

Predicting Process Capability with Variable Geometric Tolerance Specifications

Abstract

Current statistical formulas used to predict process capability from continuous data regard all dimensional specifications as having constant tolerance while many of them are variable. Geometric tolerances and datum feature designations followed by the tolerance modification symbols ~ or ~ permit a variable "bonus" or "datum-shift" tolerance based upon feature size. Process capability predictions from discreet data (measurement with attribute gauges) include the variable portion of tolerance in the gauge design but predictions from continuous data (measurement with variables gauges) ignore the variable portion of tolerance. The disparity between the capability prediction methods affect product acceptance and rejection, process precision requirements, gauge preferences, process optimization (targeting feature sizes), and design practices (designation of modifiers and the proportioning of variable tolerance between size and geometric form, orientation, or location). Since most quality control procedures require continuous data from variables gauges for manufacturing process analysis and control the problem is common throughout the manufacturing industry.

This paper presents a solution to the problem of predicting the process capability of variable geometric tolerances. It also demonstrates the rationale for targeting feature size to optimize the process performance of geometric specifications and it cautions against the designation of tolerance modification symbols for reasons other than static fit, clearance, or boundary conditions (gauging preferences should not drive modifier selection).

Page 2 of 14

Predicting Process Capability

There is a problem with the statistical formulas used to characterize the quality of manufacturing processes. One of the most basic instruments in the quality arsenal "Process Capability Prediction" is broken and needs to be fixed. The problem simply stated is that the statistical formulas widely used to predict manufacturing process capability from continuous data regard engineering tolerances as constant while many of them are variable. The problem lingers, I

suspect, because those that are expert in statistics are not often expert in geometric dimensioning and tolerancing and visa-versa. If the quality improvements envisioned in "Six Sigma" are to be realized one of its foundational measurement tools "Process Capability Prediction" needs to be re-defined.

This paper explores a potential solution to the problem of predicting process capability with variable geometric tolerance specifications by addressing the following questions. What are variable geometric tolerances? How is process capability currently measured? How and why should variable tolerances be analyzed differently? How do variations in feature size affect the capability predictions? Can variable tolerance from Datum-Shift be included in the capability a...