Based On The Vibration Of Large Thread Whirlwind Milling Modeling And Experimental Study Process

High-speed hard whirling for large-scale thread is an advanced machining technology which holds advantages of high efficiency, well processing quality, and green technique without needing cutting lubricant. The thread hard whirling process is obviously distinguished from common cutting methods because of its characters of time-varying interrupted impact and variable multi-point constraints. The on-line monitoring system was constructed to study the whirling process parameters on whirling process. Meanwhile, the influencing factors and prediction methods of the thread surface roughness were also studied, which can provide design basis for the whirling process of screw.Considering the geometric constraints of the clamped and hinged support, theoretical dynamics model of the tool-work piece system was set up based on "Generalized Polynomial Assumed Mode Method" under a moving load. The moving load is interrupted and periodical, motivated by the cutting force in the whirling process. Then the effect of hold devices on the dynamic response of large-scale whirling system was analyzed.The optimization of whirling process parameters was studied based on the vibration signals of the whirling machine in the whirling process. The single factor test was employed to study the influences of process parameters on hard whirling process and chose the normal level for the orthogonal test of the whirling process parameters.The effect of whirling process parameters and vibration on the thread surface roughness was researched. The vibration signal of upper left hold device was chosen to reflect the interaction of the cutting tool, whirling machine and the screw work piece.Finally, the prediction model of screw surface roughness was founded based on the vibration signals of whirling machine in the whirling process and whirling process parameters. The prediction accuracy of multivariate regression model was compared to artificial neural network model.