Study On Development And Application Of Magnetorheological Finishing Expenrimental Machine

With the development of magnetorheological fluid (MRF), magnetorheological finishing has become an advanced ultra-precision process method. In the magnetic field, MRP becomes a Bingham fluid, which is similar to half solid when magnetic field is applied to it and its stiffness is proportional to field strength directly. MRF becomes flowing fluid as soon as the magnetic field disappears. For the feature of controllable viscosity, MRF can be used in ultra-smooth surface polishing technology, with which surface precision can be nanometer level. The controllable flexible polishing is a high-efficient high-precision process method in the field of laser fusion, spaceflight, aerospace detection, military spy and so on.Based he mechanism of the magnetorheological finishing, a magnetorheological finishing experimental machine is introduced. With the vertical placing polishing mode, the magnetorheological finishing device is combined with VM-1360 CNC. Four DOF in translation of x, y, z-axis and rotation of z-axis and control system are shared. The device is divided into magnetic field, circulating, part clamp, polishing wheel and transmission. The magnetorheological finishing experimental machine can ultra-precision polish the small plane of the part.Research on the effect of several parameters in magnetorheological finishing is one of key technologies to deterministic process. After a small quantity of MRF experiment, the laws of the effect on the removal efficiency and the surface quality by those parameters, including the intensity of magnetic field, polishing wheel rotation speed, part rotation speed and size of gap between.the optical part and polishing wheel, are given with the quadrature analyses.Finally, a magnetorheological finishing optimization model with multi-input, single output and nonlinear technological parameters based on Support Vector Regression (SVR) theory is given by the sample data from the experiment. The model realizes the bidirectional optimization among the magnetorheological finishing parameters. With the accumulation of sample data, the model can solve the optimization of multi-input, multi-output and nonlinear parameters.