Applied Mechanics and Materials Vols. 157-158

Abstract: Three-dimensional discrete element process was adopted to simulate the discharge of hopper and the change of flow pattern of particles and dynamic lateral pressure of walls were analyzed in process of discharge. The results showed that three-dimensional discrete element simulation can obtain much more information. Discrete element method can simulate dynamic discharge problems. The dynamic force of vertical direction is much greater than dynamic force of horizontal direction in the process of discharge in the cone section of hopper. When the stored materials in hopper generate arches in the course of discharge, the inertia of floating stored materials will make the pressure of particles above arches hoist, so the walls of hopper may crack.

Abstract: This paper describes structural optimization design of parallel piezoelectric six-component force sensor. Structure, measuring principle, and load sharing principle are introduced. The influences on load sharing effect of each component of sensor’s structure are researched. Main parameters to influence load sharing effect are obtained, and analyzed by finite element method (FEM). It is clear that the analysis results of FEM are well correspond with the theoretical analysis results. Experiment is designed to verify analysis results. The trend is in consistent with FEM analysis.The paper provides reference basis for design of parallel piezoelectric six-component force sensor.

Abstract: Currently specific discharge point cannot be effectively located using common electrometric methods of generator stator winding partial discharge online monitoring. To solve the problem an electrical-sound joint detection method has been proposed. Since accuracy of location would be heavily affected by time delay of ultrasonic wave propagation, background noise and spectrum of partial discharge ultrasonic signal were analyzed first, and ultrasound field distribution and ultrasonic wave propagation were also stimulated. Moreover, an ultrasonic sensors distribution method was proposed to reduce the delay estimation error. The method was verified by field experiments. Experimental results showed that the theoretical analysis was exact and the sensor distribution method was feasible, thus laying foundation for further realizing large generator partial discharge Ultrasonic wave on line monitoring and location.

Abstract: Owing to the existence of supercavition affect the trajectory as well as the stability of motion of projectile. So modeling and controle problems for supercavitation prjectile are complicated and important. The paper analysed the hydrodynamic of supercavitation projectile. And the nonlinear dynamics equatioons motion in body coordinates were obtained. Inorder to obtain the excellent stability performances of projectile, the robust pole assignment method was used to controle the attack angle of head(cavitator). And the attack angle of head effect to the stability of supercavitation projectile was obtained based on stability equtaions. Several structure parameters of the cavitator were designed to compared the contrlole efficiency. Simulation results show that projectile can have good navigation stability through controle the cavitator attack angle. The controle method is of engineering signification.

Abstract: A GPS signal tracking method utilizing optimized processing in inertial measurement unit (IMU) aided GPS receiver is studied. In order to enhance the sensitivity of baseband processing, low-cost inertial sensors are applied in the GPS and strapdown inertial navigation system (SINS) integration commonly, where the estimation accuracy of Doppler frequency would influence the whole performance of the tracking loops. Stochastic errors of carrier Doppler estimation caused by inertial sensors and local oscillators are corrected utilizing auto regressive moving average (ARMA) method in this paper. And then the accuracy Doppler information is used to correct the local code phase and local carrier frequency to further determine the search space of frequency domain which is unlike to the design of traditional aided loop. Simulation results indicate that the bandwidth of carrier loop and code loop could be decreased significantly in high dynamic environment.

Abstract: Support Vector Machine (SVM) is a new pattern recognition method developed in recent years on the foundation of statistical learning theory. It wins popularity due to many attractive features and emphatically performance in the fields of nonlinear and high dimensional pattern recognition. Due to the complexity of the deep excavation, deformation prediction problem has not been a good solution. In the paper the support vector machine model was proposed to predict the deep excavation deformation. On the basis of deep excavation displacement data measured with real time series, the model of deep excavation displacement with time was built by SVM. Typical deformation data of deep excavation is used as learning and test samples. Comparison analysis is made between calculated values generated by SVM method and observed values. The result shows this method is feasible and effective.

Abstract: In order to test the reducing adhesion and resistance efficient of bionic semi-cylindrical metal non-smooth surface, the direct shear test is experimented on a kind of bionic dredging tools with typical bionic half- circularity metal surface non-smooth surface. The relation of shear force and shear displacement on a certain pressure is presented. The result will help to design the structure of cutting soil tools’ surfaces with less adhesion and resistance.

Abstract: Variable cross-section components are widely used in engineering practice. The paper discussed the reliability analysis, sensitivity numeration and robust design for variable cross-section components obeying normal distribution, built up the reliability-based robust design model, propose a method discussing reliability-based robust design of variable cross-section wedge beam. According to the numeration, obtain the graph of reliability index and reliability of variable cross-section components, obtain the method designing reliability-based robust of variable cross-section components with the certain reliability index.

Abstract: For aluminum hot rolling process with multi-variables, strongly coupling, nonlinear, it is difficult to establish accurate mathematical model to solve transverse thickness distribution. A method based on BP neural network with hot rolling aluminum transverse thickness prediction was proposed; According to multi-channel measurement characteristics of IMS transverse thickness collection system, the BP neural network prediction model for single channel was established; predicting every channel thickness by BP neural network, transverse thickness distribution of aluminum strip was obtained. Simulation results compared with the measured datas, the relative error is less than 2.5%. The results indicate that the BP neural network is suitable for transverse thickness prediction of hot rolling aluminum strip.

Abstract: Introduces a kind of method that creating the point cloud for laser inner carving from texture of 3D model surface. At first, the 3D model surface with texture needs to be converted to the triangle faces. Then texture coordinates which corresponded to the vertex will be obtained by visiting the triangle faces. The transform from texture triangle surface to 3D model triangle surface will be completed next. Then the Bayer dithering algorithm should be applied to handle the texture image. At last, the inner carving point cloud will be gained by transforming from the surface texture space to the 3D model space.