• Journal of Korean Institute of Industrial Engineers
• /
• v.39 no.5
• /
• pp.361-366
• /
• 2013

Prediction of process parameters is very important in parameter design. If predictions are fairly accurate, the quality improvement process will be useful to save time and reduce cost. The concept of dual response approach based on response surface methodology has widely been investigated. Dual response approach may take advantages of optimization modeling for finding optimum setting of input factor by separately modeling mean and variance responses. This study proposes an alternative dual response approach based on machine learning techniques instead of statistical analysis tools. A hybrid neural network-genetic algorithm has been proposed for the purpose of parameter design. A neural network is first constructed to model the relationship between responses and input factors. Mean and variance responses correspond to output nodes while input factors are used for input nodes. Using empirical process data, process parameters can be predicted without performing real experimentations. A genetic algorithm is then applied to find the optimum settings of input factors, where the neural network is used to evaluate the mean and variance response. A drug formulation example from pharmaceutical industry has been studied to demonstrate the procedures and applicability of the proposed approach.

• Proceedings of the KWS Conference
• /
• 2006.10a
• /
• pp.253-255
• /
• 2006

Grey relational analysis has been carried out to develop a new approach for optimization of Nd:YAG laser and MIG hybrid welding parameters. The quality of welded material depends on welding parameters. The parameters chosen for current study include wire type, shielding gas, laser energy, laser focus, traveling speed, and wire feed rate. The welding experiments were performed on 6K21-T4 aluminum alloy sheet. Functional demands on products may vary widely depending on their use. The ultimate tensile stress, width, and penetration were chosen as the optimization criterion. Practice based on an orthogonal array which is following Taguchi's method has been progressed. Base on the results of grey relational analysis, the optimal process parameters were obtained. This integrated work was judged and it is observed that the results obtained by using the optimal parameters are much improved compared to those obtained through initial setting.

• Journal of the Korean Vacuum Society
• /
• v.8 no.4B
• /
• pp.541-547
• /
• 1999

Reactive ion etching process for InP using BCl3/O2/Ar high density inductively coupled plasma was investigated. The experimental design method proposed by the Taguchi was utilized to cover the whole parameter range while maintaining reasonable number of actual experiments. Results showed that the ICP power and the chamber pressure were the two dominant parameters affectsing etch results. It was also observed that the etch rate decreased and the surface roughness improved as the ICP power and the bias voltage increased and as the chamber pressure decreased. The Addition of oxygen to the gas mixture drastically improved surface roughness by suppressing the formation of the surface reaction product. The optimum condition was ICP power 600W, bias voltage -100V, 10% $O_2$, 6mTorr, and $180^{\circ}C$, resulting in about 0.15$\mu\textrm{m}$ etch rate with smooth surfaces and vertical mesa sidewalls Also, the maximum etch rate of abut 4.5 $\mu\textrm{m}$/min was obtained at the condition of ICP power 800W, bias voltage -150V, 15% $O_2$, 8mTorr and $160^{\circ}C$.

• Journal of Korean Society for Quality Management
• /
• v.42 no.4
• /
• pp.579-590
• /
• 2014

Purpose: Ball milling is a popular process for obtaining fine powders in the part and material industry. One of important issues in the ball milling is to produce particles with a uniform size. Although many factors affect uniformity of particles, this paper focuses on the choice of ball diameter. Consider a ball milling where balls can be taken with three different diameters. The purpose of this paper is to find a ball mix which minimizes the average particle size. Methods: Ball diameters are selected as 10mm, 3mm, and 0.5mm. In order to find an optimum mixing ratio, the method of mixture experiments is employed in this paper. Taguchi's signal-to-noise ratio (SNR) for smaller-the-better type is also used to analyze experimental data. Results: According to the experimental result, SNR is maximized when the ball mix is taken as either 7:3:0 or 6:4:0. Such mixing ratios can be technically validated in terms of porosity reduction. Conclusion: The ball mixing technique presented in this paper provides a useful way to improve the production efficiency with a low cost.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.2 no.3
• /
• pp.20-25
• /
• 2003

The End-milling has been widely used in the industrial world because it is effective to cutting working with various shape. Recently the end-milling is demanded the high-precise technique with good surface roughness and rapid manufacturing time for precision machine and electronic elements. The cutting working of end-milling such as, cutting direction, revolution of spindle, feed rate and depth of cut have an effect on optimum surface roughness. This study was carried out to decide the working condition for optimum surface roughness and rapid manufacturing time by design of experiment and ANOVA. From the results of this study, the optimum working condition for end milling is upward cutting in cutting direction, 600rpm in revolution of spindle, 240mm/mm in feed rate, 2mm in axial depth of cut and 0 25mm in radial depth of cut. The design of experiment has become an useful method to select optimum working condition mend-milling.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.16 no.3
• /
• pp.125-130
• /
• 2017

In this study, the influence of various factors on surface roughness was investigated using the Taguchi experimental method through high-speed machining processing. Feed rate, pitch, tool diameter, and depth of cut are widely applied to high-speed machining conditions for mold production. Each of these factors was implemented and classified into three levels; then, after high speed machining, surface roughness was measured, the S/N ratio was analyzed, and the influence on the surface roughness of control factors was analyzed quantitatively by ANOVA. Using this information, a mathematical model for predicting surface roughness was derived from multiple regression analysis. This mathematical model enables the surface roughness value after high-speed machining to be predicted at the production stage, before machining, for a wide range of machining conditions.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.30 no.6 s.249
• /
• pp.728-733
• /
• 2006

The reduction of the Vehicle interior noise has been the main interest of NVH engineers. The driver's perception on the vehicle noise is affected largely by psychoacoustic characteristic of the noise as well as the SPL. In particular, the HVAC sound among the vehicle interior noise has been reflected sensitively in the side of psychology. In previous study, we have developed to verify identification of source for the vehicle HVAC system through multiple-dimensional spectral analysis. Also we carried out objective assessments on the vehicle HVAC noises and subjective assessments have been already performed with 30 subjects. In this study, the linear regression models were obtained for the subjective evaluation and the sound quality metrics. The regression procedure also allows you to produce diagnostic statistics to evaluate the regression estimates including appropriation and accuracy. Appropriation of regression model is necessary to $R^2$ value and F-value. And testing for regression model is necessary to independence, homoscedesticity and normality. Also we selected optimum layout of damping material using Taguchi method. As a result of application, sound quality is improved more quietly, powerfully, even though costly, and smoothly.

• Journal of the Korean Society for Precision Engineering
• /
• v.28 no.3
• /
• pp.370-376
• /
• 2011

Recently, the application range of plastic gears is widely expanding by the development of engineering plastics with good mechanical properties. Plastic gears have excellent performances such as light weight, water resistance and vibration absorbing ability for metallic gears. In this study, the optimization of injection molding process was done for the large disk type plastic gears of auto phoropter. Design Of Experiment (Taguchi method) was adopted to find a tendency of molding conditions that influence the flatness of disk type gear. Four main factors for molding conditions were selected based on injection temperature, filling time, packing pressure and mold temperature. Also, Filling, packing and cooling analyses were carried out to evaluate Z directional deflection of large disk type gear by using the simulation software (Moldflow) based on the DOE. From the results, it was found that the injection temperature and packing pressure are the most sensitive parameters for the Z directional deflection of large disk type gears.

• Journal of Korean Society for Quality Management
• /
• v.24 no.3
• /
• pp.111-132
• /
• 1996

Taguchi's parameter design is to determine optimal settings of design parameters of a product or a process such that the characteristics of a product exhibit small variabilities around their target values. His analysis of the problem has focused only on a single characteristic or response, but the quality of most products is seldom defined by a characteristic, and is rather the composite of a great number of characteristics which are often interrelated and nearly always measured in a variety of units. The critical problem in dealing with multiple characteristics is how to compromise the conflicts among the selected levels of the design parameters for each individual characteristic. In this paper, the EXTOPSIS Model using SN ratio which can be optimized by univariate technique is proposed and a parameter design procedure to achieve the optimal compromise among several different response variables is developed. Two existing case studies are solved by the proposed method and the results are compared with ones by the sum of SN ratios, the expected weighted loss, and the desirability function.

• Journal of the Korean Society for Precision Engineering
• /
• v.23 no.4 s.181
• /
• pp.91-98
• /
• 2006

A light guide panel (LGP) is an element of the LCD back light unit, which is used for display devices. In this study, the laser marking process is applied to the fabrication of light guide panels as the new fabrication process. In order to obtain a light guide panel which has high luminance and uniformity, four principal parameters such as power, scanning speed, ratio of line gap, and number of line were selected. A Web-based design tool was developed to generate patterns of light guide panel at any location, and the tool may assist the designer to develop optimized patterns. Topcon-BM7 was used for luminance measurement of each specimen with $100mm{\times}100mm$ area. By Taguchi method optimized levels of each parameters were found, and luminance of $3523cd/cm^2$ and uniformity of 92% were achieved using the laser machined BLU.