Jinfei Shi

3D Stacked Package Technology and Its Application Prospects

In recent years, the increasing demands for the high performance integrated circuit devices have led to the development of multi-die stacked technology in a single package form. The 3D (three dimension) stacked package technology is developing trend of the integrated circuit advanced high-density packaging, which can easily meet the developing of smaller footprint, lower profile, multi-function, lower power consumption and lower cost for the cell phones and consumer products like digital cameras, MP4, PDA and other wireless devices. Some correlative concepts of the 3D stacked package have been proposed in this paper. Firstly, the development trends and the general classifications of 3D stacked package have been introduced. Furthermore, in order to compare with the traditional 2D package (MCM), the advantages of the 3D stacked package technology have been discussed, and it also briefly states the technical challenges that 3D stacked package technology must be faced. In addition, the potential applications that may take advantage of 3D stacked package technology are discussed in the end.

Surface roughness vision measurement in different ambient light conditions

Machine vision method was used to measure the surface roughness for different Phi38 mm grinding axes in different ambient light conditions. To analyze the effect of ambient light, a method based on the gray-level co-occurrence matrix was applied at the fist step. Then, a new calculation method was proposed to minimize the effect of ambient light during measurement. At the calibration stage, the intensity of ambient light was put into consideration to indicate the relationship between the surface roughness and the corresponding features of the workpiece image. To measure an unknown workpiece, the corresponding features of ambient light and inspected workpiece would be input to calculate the roughness value. Finally, a case study is provided to demonstrate the measurement procedures and effectiveness of the proposed methodology. The experiments show that the proposed method has better accuracy than the gray-level co-occurrence matrix method when the error between inspecting value and its corresponding real surface roughness is set at 0.05 mum.

Dynamic modelling and response analysis of closed high-speed press system

It is necessary to establish a dynamic model to predict and evaluate the dynamic performance of closed high-speed press system during design procedure, when it is impossible to carry out the test of prototype. In this work, a dynamic model for the closed high-speed press system is developed to predict and evaluate its dynamic performance. Unlike in previous models where the press frame is considered to be a single mass and the effects of the transmission mechanism and complex frame structure are neglected, the present model can be used to analyze the dynamic response of a multi-body vibration-impact system treating it as a coupled one. The simulation results reveal that the dynamic responses of press under piling condition include a transient impact and a free vibration process. The vibration acceleration and velocity curves of the upper beam of the closed high-speed press decrease with the increasing speed of rotation. In contrast, vibration acceleration and velocity curves of the lower beam of the press increase with the increasing rotation speed. The experimental results demonstrate that the dynamic model proposed herein is consistent with the actual working state of the closed high-speed press and the validity of the dynamic model is verified.

A neural network-based machine vision method for surface roughness measurement

In our current study, a neural network-based machine vision method is proposed to measure the surfaces roughness for different Ф38mm grinding shafts in different ambient light conditions. Firstly, the effect of ambient light is analyzed using the two approaches, i.e., the approach of standard deviation of gray-level distribution proposed by Luk and that based on gray-level co-occurrence matrix. Then, a new RBF neural network-based method is proposed to measure the roughness by extracting the features of ambient light and work piece. The neural network is trained by five work pieces with known surface roughness, and eleven work pieces are tested by the proposed method. An analytical comparison between the proposed method and the two existing ones mentioned above verifies that our method is of better performance with least variance sum.

A high-precision vision measurement method based on dimension characteristics of sequential partial images

Due to the advantages of continuity, non-contact and non-destructiveness, vision measurement has become an innovative method in dimension measurement. To realize the high-precision vision measurement for large scale machine parts, a new vision measurement method based on dimension characteristics of sequential partial images is proposed. To overcome the corresponding problems arising from the relative rotation between two sequential partial images, a rectifying method based on texture features is used at the first step. Then, the transitional distribution modality of part image edges and its effects on measurement precision are analyzed. To get rid of the effects, a method of edge-pixel compensation is put forward, which availably improve measurement precision. Finally, a case study is provided to demonstrate the analysis procedures and effectiveness of the proposed methodology. The experiments show that the relative error is less than 0.012% using the sequential image measurement method to gauge large scale straight-edge parts. The measurement precision meets the need of the precise measurement .

Workspace and Kinematics Analysis of a 2-DOF Decoupled Spherical Parallel Mechanism

A 2-DOF (degree of freedom) decoupled spherical parallel mechanism is presented. The mechanisms configuration characteristic are studied and summarized. Its branch-chains are different completely, one is a 1/4 circular branch-chain, the other is a 7R+ R branch-chain, and each branch-chain can only control one rotational direction of the movable platform. Furthermore, in order to verify the mechanisms decoupling characteristic, the mechanisms kinematics analysis is completed, which is concerned with the direct/inverse kinematics problems, the input/output speed and the acceleration analysis. Finally, the workspace of the parallel mechanism is analyzed systematically. All these works provided the significant theory for the further study of the mechanisms application. The decoupled spherical mechanism has wide application in the fields of the robots waist, wrist, shoulder, hip and antenna orientation, camera tracking, solar localization et al.

Temperature Control Strategies for Microwave Hot In-Place Recycling of Asphalt Pavements

In order to realize the optimal temperature control for microwave hot in-place recycling of asphalt pavements, the electric field in heating process was solved based on the radio propagation principle, and the equation of heat flow control was established by the theory of permeability plate. The temperature control model was built through coupling between the two equations. The exact solution of the nonhomogeneous linear differential temperature control equation was studied by mathematical analysis method according to the boundary condition during the heating. Effect of electromagnetic and structure parameters on temperature distribution were systematically analyzed. The result indicates that the area ratio between solid and flow dielectric constant have little effect on temperature distribution. However, dielectric loss factor and the value of maximum field strength have remarkable effect on the uniformity of temperature. Temperature field in the asphalt mixtures changes more significant when the parameters are larger. The optimal objective function is built and operation control strategies are researched in two aspects: electrical field, heating method. The result indicates that using 2M210 as heating port is more reasonable than using 2M211. It also shows from experiment that the heating method with 0.5 mins intermission can improve effect of recycling asphalt mixtures.

Surface Roughness Vision Measurement in Different Ambient Light Conditions

Machine vision method was used to measure the surface roughness for different Phi38 mm grinding axes in different ambient light conditions. To analyze the effect of ambient light, a method based on the gray-level co-occurrence matrix was applied at the fist step. Then, a new calculation method was proposed to minimize the effect of ambient light during measurement. At the calibration stage, the intensity of ambient light was put into consideration to indicate the relationship between the surface roughness and the corresponding features of the workpiece image. To measure an unknown workpiece, the corresponding features of ambient light and inspected workpiece would be input to calculate the roughness value. Finally, a case study is provided to demonstrate the measurement procedures and effectiveness of the proposed methodology. The experiments show that the proposed method has better accuracy than the gray-level co-occurrence matrix method when the error between inspecting value and its corresponding real surface roughness is set at 0.05 mum.

A distortion-correction method for workshop machine vision measurement system

The application of machine vision measurement system is developing rapidly in industry for its non-contact, high speed, and automation characteristics. However, there are nonlinear distortions in the images which are vital to measuring precision, for the object dimensions are determined by the image properties. People are interested in this problem and put forward some physical model based correction methods which are widely applied in engineering. However, these methods are difficult to be realized in workshop for the images are non-repetitive interfered by the coupled dynamic factors, which means the real imaging is a stochastic process. A new nonlinear distortion correction method based on a VNAR model (Volterra series based nonlinear auto-regressive time series model) is proposed to describe the distorted image edge series. The model parameter vectors are achieved by the laws of data. The distortion-free edges are obtained after model filtering and the image dimensions are transformed to measuring dimensions. Experimental results show that the method is reliable and can be applied to engineering.

Effect of ambient light on surface roughness inspection

Machine vision is used to assess surface roughness of work pieces in different circumstance light conditions. Different from the traditional stylus way, the machine vision method is a non-contact and non-destructive surface inspection method, and therefore has the prospect in manufacturing application. The effect of ambient light on gray-level distribution parameters was first analyzed. Then a new method, which counts the pixel numbers whose gray value is larger than sum of the Otsu threshold and a given constant, was proposed to represent different roughness. Finally some experiments were conducted to compare the proposed method with other two methods-the first is the method based on standard deviation and root mean square height of the gray-level distribution proposed by Luk, and the second is on gray-level co-occurrence matrix-to evaluate surfaces with different roughness in different conditions. The parameters, such as the constant of the proposed method, the distance and the angle of the gray-level co-occurrence matrix, were optimized to obtain better inspecting performance. Results show that the variance sum between inspecting values of the proposed method and its corresponding real surface roughness is less than half of the other two ways.