CN109128413A - A kind of electrode clamping mechanism and its method applied to five axis electric machining - Google Patents

Abstract

The invention discloses a kind of curved surface EDM parameter methods corresponding with Milling Process parameter established using neural networks principles, this method mainly utilizes now existing 5-shaft linkage numerical control lathe and its digital control system on the market, by the clamping device for designing a kind of line tracking electric machining with wire electrode collocation structure, Application of Neural Network Technology establishes the corresponding relationship in traditional numeric-control processing between numerical control milling parameter and EDM parameter.At the interface of CNC milling machine, Milling Process parameter is inputted, is the EDM parameter needed by Milling Process Parameter Switch by the neural network having built up, final realize completes curved surface electric machining process on the basis of five number of axle Milling Machining centers.The method of the present invention is simple, easy to accomplish, under less expensive cost, by the electric machining that the repacking of existing five-shaft linkage numerically controlled machine may be implemented more deep camber metal surface.

Description

A kind of electrode clamping mechanism and its method applied to five axis electric machining

Technical field

The present invention relates to five-axle linkage spark erosion technique domain variabilities to be related to neural network algorithm technical field. A kind of electrical discharge machining clamp structure with collocation structure is specifically designed, and is designed on this basis a kind of based on five axis The method that the corresponding relationship of Milling Process parameter and EDM parameter is established at linkage motion cutting center.

Background technique

Since nineteen forty-three former Soviet Union scientist La Zhalinke Mr. and Mrs' invention electric discharge machining method, process generation more than half The research and development of discipline, electrical discharge machining has become a kind of important manufacturing process, and in subsequent development, it is grand long that Japan increases pool Professor proposes that line electrode electric spark grinding technology becomes the turning point in electrical discharge machining development history.It is multiple in Modern Manufacturing Science Miscellaneous Machining of Curved Surface, difficult-to-machine material processing (such as electronic ceramics, high temperature alloy, hard alloy), Complex Different Shape face are (such as: three-dimensional Curved surface, special-shaped hole slot etc.) manufactures of components has become the hot spot of research, push the development of micro EDM.

Micro EDM has many advantages, such as low stress, impulse- free robustness, processable high hardness material, in micro processing field In be widely adopted, have become one important developing direction of micro processing field.

Existing 5-shaft linkage numerical control lathe mainly has: overhead five-shaft linkage machine tool, five-axle linkage ULTRASONIC COMPLEX electrical fire Spend micro- retrofit numerically-controlled machine tool.Overhead five-shaft linkage machine tool is more mature, but structure is complex, it is difficult to realize complicated bent Face processing, difficult-to-machine material processing either curved surface low stress shell processing；Five-axle linkage ULTRASONIC COMPLEX electric spark is micro- finely to be added Number control machine tool feed direction is stablized, and electrode fine feeding processing on complicated route may be implemented, realize to a certain extent multiple The processing of miscellaneous Special-Shaped Surface, but whole electric spark machine tool system is huge, manufacturing cost is high to be difficult to be widely popularized.

Summary of the invention

In view of the deficiencies of the prior art, the present invention devises a kind of electrical discharge machining clamp structure with collocation structure, And provide a kind of curved surface EDM parameter method corresponding with Milling Process parameter established based on neural networks principles.It utilizes The electrical discharge machining clamp structure with collocation structure of design and the curved surface EDM parameter of foundation and Milling Process parameter pair The electric machining of more deep camber metal surface may be implemented complicated bent in the repacking of existing five-shaft linkage numerically controlled machine in the method answered EDM micro elements are carried out on face.This method structure is simple, low in cost, it is easy to accomplish.

The technical scheme adopted by the invention to solve the technical problem is that: first on existing 5-shaft linkage numerical control lathe Micro EDM cutter is installed；And the collocation structure of micro EDM is designed, the compensation of tool-electrode creep feed Mechanical device include upper self-clamping module, lower self-clamping module, elongated modules, four part of tool-electrode；Then pass through physical relation Correspondence obtain Milling Process parameter, spindle power, spindle speed and feed engagement, be obtained by calculation input response ginseng Under several, the metal removal rate of workpiece to be machined, while using the method for sample difference, it is established by neural networks principles Corresponding relationship between metal removal rate and electrical parameter, to obtain the corresponding pass between EDM parameter and Milling Process parameter System.

Of the invention has the advantage that 1) method is simple, easy to accomplish；2) the new micro EDM knife designed Tool has compensation mechanism, it is ensured that during electric machining, electrode tip is with workpiece surface gap relative distance always Definite value guarantees electrode tip energy stable discharging, guarantees the steady progress of processing；3) EDM parameter and Milling Process parameter designed Correspondence algorithm can be considered as the black box between five-shaft linkage numerically controlled machine and micro EDM cutter, filled by this black box It sets, the Milling Process parameter of input terminal can be directly transmitted in micro EDM Tool Control with EDM parameter The heart, and then realize the repacking that 5-shaft linkage numerical control Micro Edm Machine Tools are completed on any five-shaft linkage numerically controlled machine.

Detailed description of the invention

Fig. 1 is the schematic diagram that the collocation structure of micro EDM is designed in the present invention, 101- tool electricity as shown in the figure Self-clamping module under pole, the upper self-clamping module of 102-, 103- elongated modules, 104-；

Fig. 2 is micro EDM cutter spindle master-plan structural blast schematic diagram in the present invention, as shown in the figure Axis, 204- clip-on cap, 205- cramping body, 206- outside 201- medial axis ring support, 202- coil and jackshaft connector, 203- Jackshaft, 207- voice coil motor coil；

Fig. 3 is the assembling schematic diagram of micro EDM cutter spindle master-plan structure in the present invention；

Fig. 4 is the corresponding relationship schematic diagram designed between EDM parameter and Milling Process parameter in the present invention；

Fig. 5 is that the neural networks principles figure that relationship is established between EDM parameter and metal removal rate is designed in the present invention.

Fig. 6 is that the present invention is based on the network simulation flow charts that neural network RBF neural network model is established.

Specific embodiment

By taking the five-shaft linkage numerically controlled machine of certain model as an example.Micro EDM is installed in the five-shaft linkage numerically controlled machine Cutter, due to the loss situation in edm process there are electrode, if wire electrode cannot keep closing with for workpieces processing The discharging gap of reason, electric machining process can stop because of export license, so needing to design the electrical spark working with collocation structure Work clamp structure, collocation structure motion principle as shown in Figure 1, the motion process of the wriggling collocation structure include: 1) on clamp Module (102) unclamps micro tool electrode (101), releases the clamping to micro tool electrode (101)；2) elongated modules (103) Elongation drives micro tool electrode (101) and lower self-clamping module (104) to feed down certain distance；3) self-clamping module (102) on It clamps micro tool electrode (101)；4) self-clamping module (104) unclamps release micro tool electrode (101) under, releases and clamps；5) Elongated modules (103) backhaul, and lower self-clamping module (104) is driven to retract original position；6) self-clamping module (104) clamps fine again under Tool-electrode (101).The electrical discharge machining clamp structure with collocation structure designed under this movement mechanism, electrical discharge machining Inside and outside nested structure is used in tool spindle device, as shown in figure 3, i.e. upper self-clamping module (102) tool-electrode collet is direct It is nested in lower self-clamping module (104) inside.Inside and outside nested structure explosive view, as shown in Figure 2.Upper clamping device (102) is corresponding Structure is jackshaft coil (201), and tool-electrode (101) corresponding clip-on cap (204), cramping body (205), jackshaft (206) are stretched The corresponding coil of long module (103) and jackshaft connector (202), the corresponding external axis (203) of lower self-clamping module (104), voice coil electricity Machine (207) is mainly that the movement of wire electrode provides control movement.Lower self-clamping module had both been used as one of creep feed device to clamp Module implementation tool electrode timesharing clamping function, and the precision positioning collet as tool-electrode realize the essence to tool-electrode Close positioning.Since electric spark has the loss of wire electrode in process, when export license needs to compensate, utilize Fig. 3's Structure completes electrode creep feed electrode compensation according to the motion mode in Fig. 1.This structure can be ensured entire processed Cheng Zhong, electrode keep an identical distance with workpieces processing always, are able to achieve the steady progress of electrical discharge machining.

In the five-shaft linkage numerically controlled machine for having installed micro EDM cutter, need to propose a kind of equivalent method, Enable total Milling Machining that can establish with electrical discharge machining to contact.Principle is as shown in figure 4, be metal removal rate be equivalent relation Node, respectively using Milling Process parameter as input point, EDM parameter is output point.Form input, Milling Process parameter, metal Removal amount, EDM parameter, output equivalent model.Wherein from Milling Process parameter, metal removal rate using linear public affairs Formula derivation, directly by Milling Process parameter (Milling Speed, the cutter number of teeth, working depth, tool radius and its processing thickness) It calculates metal removal rate (mm3/min), and from metal removal rate, EDM parameter (between current peak, pulsewidth, arteries and veins and its capacitor Value) sample point interpolation is then carried out using Lagrange and Gauss interpolation function, finally obtain the equivalent result of electrical parameter.

Because EDM parameter and metal removal rate do not have specific derivation of equation relationship, electric machining ginseng cannot be directly established Several and metal removal magnitude relation.Based on this problem, a kind of curved surface EDM parameter established based on neural networks principles is proposed Method corresponding with metal removal rate, as shown in figure 5, radial basis function (the radial basis based on new development in recent years Function, RBF) network constructed the metal removal rate prediction model under different electrical parameter processing, and it is pre- using the model Metal removal rate has been surveyed, and different machining parameters can have been fitted to the overall merit of metal removal rate, electrical fire in microfabrication RBF neural in the corresponding relationship of flower electrical parameter and milling parameter, this is a kind of three-layer forward networks.Input layer is by inputting Node composition, the second layer is hidden layer, and for neuron number depending on the needs of described problem, third layer is output layer, it The effect of input pattern is made a response.Transformation from input sheaf space to implicit sheaf space is nonlinear, and from hidden layer Transformation to output sheaf space is linear.The transforming function transformation function of hidden layer neuron is radial basis function, it is a kind of part point The non-negative nonlinear function of the central point radial symmetric of cloth.The curved surface EDM parameter and metal established based on neural networks principles The corresponding method of removal amount, concrete methods of realizing is as shown in fig. 6, the process that RBF network model is established mainly includes trained sample Acquisition originally, the pretreatment of data, the design of network.(1) acquisition of training sample；(2) pretreatment of data；It need to be to sample number According to being normalized, normalized the purpose is to reduce the training times of network, but for feedforward network, due to interior Portion does not have feedback layer, can be without normalized.(3) design of network；With peak point current, pulse width, pulse spacing, punching Oil pressure, electrode sections product, input of 5 factors as model, using metal removal rate as output.(4) it tests and shows survey Test result is compared with pre- period parameters.Training by neural network to sample, the electric machining after available one group of training The corresponding relationship chart of parameter and Milling Process parameter.Applied mathematics tool can further verify the data obtained corresponding relationship Accuracy.

Claims (3)

1. a kind of electrical discharge machining clamp structure with collocation structure, it is characterised in that: clamping device is compensated with wire electrode Structure uses inside and outside nested structure in main shaft device, i.e., upper self-clamping module tool-electrode collet is directly nested in lower clamping mould Inside block, lower self-clamping module had both been used as a self-clamping module of creep feed device, implementation tool electrode timesharing clamping function, again As the precision positioning collet of tool-electrode, the precision positioning to tool-electrode is realized.

2. a kind of clamping device structure of line tracking electric machining with wire electrode collocation structure according to claim 1, Wriggling is characterized in that: upper and lower two self-clamping module can form relative motion, outer shaft and intermediate spindle in up and down motion direction Between by upper and lower self-clamping module double bond passing movement, while therebetween can opposite sliding up and down, when main shaft fine motion module When for compensation for electrode wear, needing to drag self-clamping module and move downward, self-clamping module is oriented to for tool-electrode at present for this, Upper and lower self-clamping module successively clamps in order and unclamps electrode, and the two cooperation makes electrode creep feed electrode compensation.

3. according to claim 1 with a kind of folder of the line tracking electric machining with wire electrode collocation structure as claimed in claim 2 Apparatus structure is held, using the curved surface EDM parameter method corresponding with Milling Process parameter that neural networks principles are established, foundation The physical relation and neural networks principles of Milling Process parameter and metal removal rate establish Milling Process and EDM parameter Equivalent model, in the nerve network system of foundation, input layer is Milling Process parameter, and Milling Process parameter setting is milling speed Degree, the cutter number of teeth, working depth, tool radius and its processing thickness, output layer EDM parameter are set as current peak, arteries and veins Between wide, arteries and veins and its capacitance.