Papers by Author: Hong Tsu Young

Abstract: The product quality of an ultra-precision spindle of milling machine is affected by many factors such as assembly tolerance, assembly process, and working circumstances. The purpose of this study is to find optimal assembly tolerance in order to enhance product quality. This study collects assembly parameters and testing results. In addition, those collecting data are used to train and test by the back-propagation neural network (BPNN) to obtain a quality index predictor so as to predict vibration level and temperature rise level. A variable control chart of SPC is employed to analyze the assembly tolerance data that impact on temperature rise level and vibration level. The results of the propose study not only increase the rotation accuracy and machining efficiency, but also decrease the vibration and temperature rise for ultra-precision spindle during operating.

Abstract: The robust design of chemical etching parameters is dealing with the optimization of the through-silicon via (TSV) roundness error and TSV lateral etching depth in the etching of silicon for laser drilled TSVs. The considered wet chemical etching parameters comprise the HNO₃ molarity, HF molarity, and etching time. Grey-Taguchi method is combining the orthogonal array design of experiments with Grey relational analysis (GRA), which enables the determination of the optimal combination of wet chemical etching parameters for multiple process responses. The concept of Grey relational analysis is to find a Grey relational grade, which can be used for the optimization conversion from a multiple objective case to a single objective case. Also, GRG is used to investigate the parameter effects to the overall quality targets.

Abstract: Through-silicon via (TSV) is an emerging technology for three-dimensional integrated circuit, system in package, and wafer level packaging applications. In this study, a wet chemical etching (WCE) process has been employed to enhance the sidewall quality of TSVs fabricated using nanosecond (ns) laser pulses. Experimental results show that the TSV sidewall roughness can be markedly reduced, from micrometer scale to nanometer scale. We concluded that the proposed method would enable semiconductor manufactures to use ns laser drilling for industrial TSV fabrication as the desired TSV sidewall quality can be achieved by incorporating the WCE process, which is suitable for mass production.

Abstract: Aiming at reducing cost and time of repair, condition-based shaft faults diagnosis is considered an efficient strategy for machine tool community. While the shaft with faults is operating, its vibration signals normally indicate nonlinear and non-stationary characteristics but Fourier-based approaches have shown limitations for handling this kind of signals. The methodology proposed in this research is to extract the features from shaft faults related vibration signals, from which the corresponding fault condition is then effectively identified. With an incorporation of empirical mode decomposition (EMD) method, the model applied in this research embraces some characteristics, like zero-crossing rate and energy, of intrinsic mode functions (IMFs) to represent the feature of the shaft condition.

Abstract: For optical micro- machinery processing, nanosecond laser possess a special advantage in using it as a fabrication method of smaller hole subjected to the minimum thermal distortion. Thus it has become an effective and powerful tool widely used in drilling, cutting and welding process for micro-manufacturing field. To estimate the working performance of pulsed laser, an auxiliary method in numerical skill or semi-empirical technology is usually utilized, where the important parameters including energy intensity, duration and wavelength of laser beam will be taken into account. Nevertheless, several troubles, the unstable numerical iteration for phase change and precise calibration of sensor required in the measuring process, seem to be still inevitable, and which easily makes the numerical calculation become more complicated, even the global ablating behavior will be lost. To compensate the inadequacy mentioned above, an analytic model of optical ablation for pulsed laser, based on the evaporation effect responsible for penetration mechanism, is then derived in this study. Here the penetrating behavior, during the micro-machining process, can be clearly examined with the consideration of plasma absorption. After compared with experimental results made by Chen and Schmidt for copper drilling and steel ablation for Tim, a better agreement of analytic results identifies the accessibility of proposed model which also contributes to the future investigation on pico-or femto- laser material processing.

Abstract: Chemical mechanical polishing (CMP) is a semiconductor process that is used to achieve global planarization. To maintain the stability and the throughput of CMP, the polishing pad needs to be dressed by a diamond dresser. By calculating the distribution of the scratch numbers of diamond grits in the pad, the effect of dressing can be estimated and the pad profile can be predicted. All the parameters of dressing motion were taken into account. The target equipment of this study is Mirra Mesa, which is manufactured by Applied Materials. Different types of dressers, including ring-types and full-type, different pitches and arrangements are used in CMP process. Each of them was simulated in this study. The diamond pitch is relative to the number of diamonds and the scale of the total scratch time, but does not affect the profile of the scratch time. Regular or random arrangement makes no difference in the profile of the scratch times. The width of the ring influences the number of diamonds and the scale of the total counts while the profiles remain similar.

Abstract: An optical non-contact measurement system for micro-scale 3D profiles is developed and analyzed in this research. The core concept of the system is based on Shape-from-Focus method. The methodology and the process of the measurement will be explained in this paper. Briefly, corresponding digital images were taken at different heights with predefined step intervals. The properties of images were enhanced by image processing techniques including discrete Gaussian filtering, histogram equalization and high-boost filtering. Finally, the degree of the focused image was quantified by using Sum-Modified-Laplacian algorithm. The height of each pixel was determined with Gaussian Interpolation. 3D geometry model of the specimens can be reconstructed. This system was verified that it can work on the specimen manufactured by laser beam. Analysis of precision will be revealed and the comparison with other measurement methods will be discussed later.

Abstract: In Laser Beam (L.B.) and Electron Beam (E.B.) drilling, the energy distribution significantly affects both the penetrating efficiency and working performance, both of which are usually estimated by numerical skill or experimental measure. Through the application of a stimulation model, an unstable solution with the finite difference method will result near the solidliquid interface unless much finer grid sizes are set up. To improve on the above defect, nonuniform grids are therefore utilized; this will complicate the built-up of the program and also easily causes the simulated energy distribution to be divergent in the iteration process. In this study, an estimated small Peclet number and observed narrow-deep cavity made the convective and radial diffusion terms small enough to be neglected in the governing equation. From these assumptions, the model was then used to investigate the drilling efficiency where two-phase flow convection could be simplified further into one dimension and thus the analytical solution becomes possible by transferring the penetration velocity into the logarithmic form. When compared with the experiment made by Allmen [1] , the present model shows good agreement in higher energy density and relative errors are no more than 10%.

Abstract: Miniaturization is a worldwide trend in manufacturing industry. Though lithography has been introduced to meet the basic needs, the technology is limited by its process complexity and the parts geometry to be produced. This research attempted to overcome the above obstacles by applying laser machining approaches on general 3D micro-parts. The machining model is based on a layer-by-layer concept. Experimental verification was made on a 1mm stainless steel sheet by applying a diode-pumped Nd:YVO4 laser with Q-switch option. Three main parameters: power, repetition rate and the speed of laser process were set to optimize the process quality. In the research round holes with the diameter ranging from 10 μm to 30 μm were drilled. The following step was the machining of a sloped groove with the area size of 100 μm × 100 μm for validation.

Abstract: The advantages of ductile regime grinding of silicon wafer such as smooth surface roughness (Ra < 10 nm) and minimum subsurface damage layer (< 10μm) have great impact on the production process of wafer. With ductile regime grinding, the subsequent processes such as etching and rough polishing processes can be minimized. To achieve ductile regime grinding, a fundamental concept is the application of grain depth of cut being less than the critical cut depth, dc, of the silicon wafer. However, dc is dependent on material properties, cutting conditions, and crystallographic orientation [1].The objective of this paper is to derive, and to investigate by experiment, the dc value for silicon wafer grinding. Following these key steps, the effects of dc on various major grinding parameters are studied.