Guardbanding specifications with SPC (statistical process control) application
A reasonable assumption in the manufacture and measurement of a product characteristic is that test set variation contributes to the observed random variation of the characteristic in addition to the product variation. A Supplier may use test limits (guardbands) set inside specification limits to preclude the sending of nonconforming product to the Buyer. The Supplier's goals are for very few or no misclassifications of product (especially bad called good) in the Supplier's accepted product which would be submitted to the Buyer. Meeting these goals depends on several factors. This paper reviews an example; quantifies the influencing factors of the product percent defective, the ratio of gage variation to product variation and the guardband limits; provides tables of the misclassification percentages to illustrate the effects of changing these factors; and discusses SPC ratios that may already be available to a user to estimate these factors for the further understanding of the practice of guardbanding. With estimates of these influencing factors, one could make a reasonable setting of a guardband. 6 refs., 7 tabs.
- Research Organization:
- Allied Corp., Kansas City, MO (USA). Bendix Kansas City Div.
- DOE Contract Number:
- AC04-76DP00613
- OSTI ID:
- 6111870
- Report Number(s):
- BDX-613-3781; ON: DE88000834
- Resource Relation:
- Other Information: Portions of this document are illegible in microfiche products. Original copy available until stock is exhausted
- Country of Publication:
- United States
- Language:
- English
Similar Records
Evaluation of Guardbanding Methods for Calibration and Product Acceptance
Benefits of SPC for the hydrocarbon process industries