Chen Zichen

Design of an ultra-precision vibration isolation system by imitating the special organic texture of woodpecker's brain

In this paper, a bionic structure model of ultra-precision vibration isolation system is presented by imitating the special organic texture of woodpeckers brain. The vibration isolation system mainly consists of giant magnetostrictive actuators, air-springs, a pedestal and a rubber layer. The air-springs are used to reduce high frequency vibrations, and the giant magnetostrictive actuators are used to reduce low or ultra-low frequency vibrations. In order to reduce thermal distortion caused by temperature rise of coil, a constant temperature cooling system is developed for giant magnetostrictive actuator. In consideration of complex vibration environment and self-nonlinearity of vibration isolation system, a two dimensional fuzzy active control system with two regulation factors is designed. The experiment results show that the bionic vibration isolation system has good performance against floor disturbances. It can be applied to the vibration isolation system of the ultra-precision measuring and manufacturing devices.

AXIS-BASED LOOK-AHEAD INTERPOLATOR FOR MACHINING OF SURFACES REPRESENTED BY NURBS CURVES

The dynamics influence on each axis doing high speed machining is significant but always ignored in most of the interpolators. The proposed Axis-based Look-ahead NURBS Interpolator (ALANI) strictly confines the component acceleration and jerk on each axis according to the servo system and cutting status, so that the acceleration and jerk on axes are kept under the limit in order to avoid immoderate vibrations and shocks. Chord error is also confined during interpolating. ALANI may trace back and recalculate previous data while encountering dangerous points. The recalculation algorithm is composed of acceleration decreasing algorithm (ADA) and synchronously decreasing algorithm (SDA), which are able to choose secure feed rate as well as to attain high efficiency. The simulation and experiment are presented in this paper.

Demolding Defects and the Design of Demolding Device in Micro Hot Embossing Process

The defects caused by auto/manual demolding are studied to get better imprint quality in micro hot embossing process. Finite element method(FEM)is used to simulate the demolding process of high aspect ratio microstructures.The results show that the stress concentration mainly occurs at the bottom of microstructures when the auto demolding device is used.The maximum demolding stress is generated during the initial stage and the defects are mainly necking and root fracture.Great stress concentration will be caused by concentration little demolding direction deviation in manual demolding,and that stress concentration will reach the maximum in the late stage of the demolding process.190 MPa stress can be caused due to 5°direction deviation,then the demolding defects are mainly microstracture topping and root fracture.So an auto demolding device is needed when fabricating high aspect ratio structures.A new pneumatic demolding device is designed to overcome the warp distortion of polymer substrate caused by the traditional demolding system.Successful fabricating of microstructures with an aspect ratio of 3.3 verifies the improvement of the demolding device.

Simulation on Diesel/DME Mixed Injection System

Dimethyl ether (DME) is a promising and excellent alternative fuel for compression ignition DI diesel engines. In addition to the significant merit of no soot emission, DME would be of better ignition quality, lower auto ignition temperature and better atomization compared with diesel. The paper presents models and simulates fuel injection system of diesel/DME engine. Simulation calculation proven that DME could injected into mixed cavity propitious and mixed rather equality with diesel. Injecting characteristic of fuel system was impacted by blended DME, and the incidence came along with DME proportional.

Thrust ripple optimization and experiment for a permanent magnet linear synchronous motor

Thrust ripple such as end force, slot force and normal force are key factors that affect the properties of permanent magnet linear synchronous motors (PMLSM). According to different mechanics and analytical models, end force resulting from open magnetic circuit of PMLSM was greatly decreased by optimizing the length of the PMLSM mover. Slot force caused by slot effect was greatly reduced by using fraction slot structure, and normal force was calculated through the finite element method (FEM). After thrust ripple was calculated, its uniform formula was obtained through Fourier series nonlinear regression. An experimental system was set up to measure thrust ripple, and experiment results demonstrated that experimental force ripple was quite in line with that calculated by the fitting formula. The optimal theory and analysis method is effective, and the obtained formula can be utilized to compensate thrust ripple in practical applications and improve the motion performance of PMLSM.

Web-based application services composition for e-business

Web service is becoming an important approach for enterprises to implement their business applications on Internet. However, most of Web services are built based on special business function, they are isolated and do not address the information communication for integrating the business processes. In this paper, we propose a new composition framework, which comprises the Web-based semantic definition for business messages, the methodology of constructing visualized integrated processes based on Business Process Execution Language (BPEL), and the strategy of vocabulary-based business rules for e-business. In addition, we present a case study that demonstrates the way of composing different kind of application services to a uniform service

Sliding Mode Based Controller for Magnetostrictive Actuator

Magnetostrictive actuator, for its merits of relatively low voltage and high force, has been increasingly applied in many applications, such as vibration control, aviation, positioner, etc. At low drive level, magnetostrictive actuator presents linear relation between strain or displacement and input voltage or input current, while non-linear appears when applied moderate or high drive level. To achieve accurate control for high drive level, non-linear, including saturation and hysteresis, must be compensated. Sliding mode control, a robust control scheme, can handle these non-linear. As magnetostrictive actuator modelled in Jiles-Atherton model, the relation of magnetic field H and bulk magnetization M, hysteresis, is divided into anhysteresis and deviation from anhysteresis. Saturation can be compensated by inversion of anhysteresis (free-hysteresis) and then, hysteresis, represented as the deviation from anhysteresis, is compensated with sliding mode control. In section I, introduction of magnetostrictive actuator characters and compensation methods for its non-linear is presented. Then, Jiles-Atherton model based on domain wall theory is introduced in section II. Next, the design procedures of sliding-mode controller are given. Control law for GMA position tracking is derived as well. Simulation and its results are shown in Section IV. And the conclusion is made at Section V.

Neural network approach for modification and fitting of digitized data in reverse engineering

Reverse engineering in the manufacturing field is a process in which the digitized data are obtained from an existing object model or a part of it, and then the CAD model is reconstructed. This paper presents an RBF neural network approach to modify and fit the digitized data. The centers for the RBF are selected by using the orthogonal least squares learning algorithm. A mathematically known surface is used for generating a number of samples for training the networks. The trained networks then generated a number of new points which were compared with the calculating points from the equations. Moreover, a series of practice digitizing curves are used to test the approach. The results showed that this approach is effective in modifying and fitting digitized data and generating data points to reconstruct the surface model.