| Thin-walled parts,such as aerospace structural components and large engine lades,are the key parts in the field of aeronautics and astronautics.In general,the design dimensions,contour accuracy and surface quality of these parts are very high,and high-speed CNC milling technology is still the main processing technology.However,because the thin-walled workpiece is a typical weak-rigid component,the chatter phenomenon between workpiece and cutter easily occurs during the milling process,which leads to accelerated tool wear and deterioration of the workpiece surface quality.Therefore,the milling stability diagram is established in advance to select the initial process parameters without chatter,and the on-line vibration monitoring and control algorithms are aided in the milling process.They have important theoretical guiding significance and engineering application value for realizing high-efficiency and high-quality machining of weak-rigid parts.In this paper,the process characteristics and dynamics of the weak-rigid components are studied from the aspects of milling force modeling,machining stability prediction,milling chatter detection and variable speed chatter suppression.The main research work of this paper is summarized as follows:(1)On milling force modeling: Based on the unequal distance model of the main shear zone and the friction model of the second deformation zone,the analytical expressions of the micro-element forces in oblique cutting process are derived.Considering the system vibration and tool eccentricity,a numerical method of solid intersection-surface erasure is presented to accurately calculate the instantaneous dynamic cutting thickness under weak-rigid cutting conditions.On this basis,a cutting force model with high precision is established for milling cutter with arbitrary profile,and a coefficient identification method for high precision cutting force model is proposed based on the differential evolution optimization method.Experimental results show that the proposed cutting force model can accurately predict the cutting force of non-standard milling cutter,such as variable tooth pitch or variable helix angle,and the average force error is less than 10%.The proposed identification method can estimate both the cutting force coefficient and the tool eccentricity parameters by only one set of experiments,and it has strong versatility and high identification efficiency.(2)On milling stability prediction: Combined with the characteristics of the flexible milling dynamic equations,the state term of the milling delay differential equations is approximated by the higher order barycentric formula of the Lagrange interpolation polynomial.An improved Chebyshev-Gauss differential matrix operator is proposed to solve the state derivative term,and the delay equation is transformed into an ordinary differential equation form by combining the weighted residual technique and the numerical integration method.Then a novel algorithm for milling stability analysis is proposed.Simulation results show that the proposed algorithm has spectral convergence accuracy,and compared with the first-order semidiscrete method,Runge-Kutta method and Chebyshev collocation method,the computational efficiency of the proposed algorithm can respectively be increased by more than 90%,17% and16% in the single-degree-of-freedom milling model,and by more than 85%,22% and 14% in the two-degree-of-freedom milling model.The experimental results show that the stability diagram predicted by the proposed algorithm can guide the operator to select high efficiency and stable cutting process parameters.(3)On 3D stability diagram of thin-walled parts: In view of the fact that the tool and workpiece aer easy to deform in the milling process of weak-rigid thin-walled parts,the dynamics equation of multi-nodes and multi-modes milling system is established.A novel data fusion method for updating the dynamic parameters of thin-walled workpiece is proposed based on the modal data of hammering test and finite element model.On this basis,a threedimensional stability lobe diagram considering both the material removal effect and the change of tool position is established by the proposed stability algorithm.The results of finite element analysis show that the material removal effect has a great effect on the natural frequency of the thin-walled workpiece,and the mode shapes of the nodes in the tool-workpiece contact area are different when the tool position and axial cutting depth are different.The experimental results of thin-walled workpiece milling show that the predicted boundary of the three-dimensional stability diagram is in good agreement with the actual chatter location.(4)On milling chatter detection: Based on the mechanism of chatter regeneration,it is deduced that milling vibration signal can be composed of period,chatter and noise in theory.After the thorough analysis of the relationship between the acceleration vibration signal and the chatter evolution,it is found that the energy of the vibration signal will gradually converge to the chatter frequency band when chatter occurs.Furthermore,a cyclostationary energy ratio indicator is proposed for early weak chatter feature extraction.After monitoring the tendency of milling chatter,a neural network model based on the exponential damping signal is establised to accurately estimate the main chatter frequency.The results of numerical simulation and milling experiment show that the proposed chatter indicator is simple and efficient,and its size can directly reflect the chatter degree of milling system.The proposed frequency estimation method can adaptively estimate the main frequency order of the weak chatter component.The sample number of the proposed method is about 1/10 of that of the classical FFT method under the same frequency estimation accuracy,and the frequency estimation error is about 1/27 of that of FFT method under the influence of different background noise.(5)On variable speed chatter suppression: Based on the secondary development of the synchronous action function module of Siemens 840 D CNC system,a milling platform which can realize early chatter detection and variable spindle speed chatter suppression at the same time is built.A variable speed milling chatter suppression algorithm suitable for arbitrary periodic waveform is proposed,which transforms the variable time delay constant coefficient differential equation into constant time delay variable coefficient differential equation by introducing angle domain variable substitution.Simulation results show that,compared with the mainstream improved semi-discrete method and fixed step size numerical integration method,the proposed algorithm has fast and stable exponential convergence accuracy,and the computational efficiency has been improved by more than 86%.The continuous and discrete variable speed strategies are applied on the proposed milling platform,and the experimental results show that the two variable speed strategies can effectively suppress the chatter phenomenon and improve the surface quality of the workpiece. |
PDF Full Text Request