Range of process parameters of thin-walled tube NC bending

WANG Yong-an,WU Jian-jun,LI Fei-fan,HUI Yu
DOI: https://doi.org/10.3969/j.issn.1007-2012.2018.01.016
2018-01-01
Abstract:Taking including the maximum wall thickness variation rate and the maximum cross-section distortion rate,the quality evaluation parameters of large diameter thin-walled tube NC bending,as random variables,and converting the two related random variables to independent random variables,the normal distribution characteristics of the independent random variables were obtained through the Bayesian analysis.Using the SNR based multi-objective decision-making theory,the best ellipsoid domain of quality evaluation parameter under the confidence level 1-α was obtained,which centers on the optimal parameter vector.The optimization range of process parameters was obtained by neural network mapping for the mapping relationship between quality evaluation parameters and process parameters.The bending of tube of Φ50 mm × 1.0 mm × 100 mm was made,and it is shown that this method has good practicability and validity.
What problem does this paper attempt to address?