Purpose - To introduce an interval arithmetic (IA)-based approach to tolerance analysis and design and to yield evaluation. The technique distinguishes itself by its reliability in the detection of even non convex and/or non simply connected regions of acceptability. Design/methodology/approach - Range methods, and IA among them, inherently allow the evaluation of an overestimation of the true range spanned by a performance function of a given system because of variation of its parameters values. Such a feature is of great help in performing the robust design of any system, that is, in doing the placement of the design set within the region of acceptability, even in the presence of tolerances and parameters values drifts. Findings - The IA-based branch and bound numerical procedure proposed in the paper has shown its reliability and robustness during the tolerance analysis of systems exhibiting quite involved regions of acceptability. Research limitations/implications - Many remarks are reported in the paper to complete the research and extend it to a larger class of problems. Practical implications - The procedure might be built into simulation environments in order to help the designer in dominating tolerances effects. Originality/value - This paper shares in suggesting the use of a powerful tool, i.e. range methods, in the area of tolerance design and reliability.
Detection of acceptability regions by means of an interval arithmetic-based algorithm
SPAGNUOLO, Giovanni
2006-01-01
Abstract
Purpose - To introduce an interval arithmetic (IA)-based approach to tolerance analysis and design and to yield evaluation. The technique distinguishes itself by its reliability in the detection of even non convex and/or non simply connected regions of acceptability. Design/methodology/approach - Range methods, and IA among them, inherently allow the evaluation of an overestimation of the true range spanned by a performance function of a given system because of variation of its parameters values. Such a feature is of great help in performing the robust design of any system, that is, in doing the placement of the design set within the region of acceptability, even in the presence of tolerances and parameters values drifts. Findings - The IA-based branch and bound numerical procedure proposed in the paper has shown its reliability and robustness during the tolerance analysis of systems exhibiting quite involved regions of acceptability. Research limitations/implications - Many remarks are reported in the paper to complete the research and extend it to a larger class of problems. Practical implications - The procedure might be built into simulation environments in order to help the designer in dominating tolerances effects. Originality/value - This paper shares in suggesting the use of a powerful tool, i.e. range methods, in the area of tolerance design and reliability.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
