Economic tolerance design for folded normal data

Abstract

An important issue for design engineers is how to assign tolerance limits economically. Most work related to tolerance design is for nominal-is-best (N-type) quality characteristics and restricted by a normality assumption. However, smaller-is-better (S-type) quality characteristics and larger-is-better (L-type) quality characteristics are common in real applications. The practical distributions for S-type data or L-type data are typically skewed, and the normality assumption is violated. Determining tolerance with non-normal data using methodologies based on a normality assumption is not appropriate. This study considers the case in which measurements are recorded without their algebraic signs. The folded normal distribution works well to fit these absolute data. Based on the statistical properties of the folded normal distribution, this study develops an economic model encompassing quality loss, manufacturing costs, and re-work costs to determine tolerances. By minimising total costs, a procedure based on the Newton-Raphson method is utilised to obtain the optimal solution. Finally, a welding machine experiment is carried out to demonstrate the applicability of the proposed model.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 16 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.