Optimized volumetric error compensation for five-axis machine tools considering relevance and compensability. (January 2016)

• Record Type: Journal Article
• Authors: Givi, Mehrdad; Mayer, J.R.R.
• Is Part Of: CIRP journal of manufacturing science and technology. Volume 12(2016:Jan.)
• Page Start: 44
• Page End: 55
• Publication Date: 2016-01
• Subjects: Machining -- Compensation -- Volumetric error -- Error compensability -- Error relevance; Manufacturing processes -- Periodicals; 670.5
• DOI: 10.1016/j.cirpj.2015.09.002
• Languages: English
• ISSNs: 1755-5817

Abstract

Highlights: Two new notions, i.e., error relevance and error compensability are introduced. Irrelevant errors are filtered out using filtration matrix for each process. Uncompensable errors are flagged using a compensability ratio. Only relevant and compensable errors are compensated in optimized compensation. Minimal but effective commands modifications are made in validation test. Abstract: Compensation of a machine tool's volumetric errors can be achieved by adjustment of the axes original commands. Depending on the machine topology, the volumetric errors and the axes commands, a full compensation may not be achieved due to kinematic singularities. Furthermore, corrections of volumetric errors which do not affect part quality, by virtue of the machining process and machined feature, may lead to excessive corrections causing surface degradation. The paper addresses these problems through the introduction of two notions, error relevance and error compensability, leading to an optimized compensation in which minimal but effective command modifications are made. Mathematical definitions are presented together with application examples for different processes. A specially designed test part containing four common features, i.e. hole, curved slot on a spherical surface, cone frustum and flat surface is machined, using up to five-axis simultaneous machining, for the experimental validation. Three parts are machined using uncompensated G-code, compensated G-code and optimized compensated G-code, respectively, which are then measured on a coordinate measuring machine. Simulation results show a reduction (up to 75%) in the 1-norm of the linear and angular compensations while the relevant errors are effectively corrected by the proposed optimized strategy.