• 산업공학
• /
• 제22권3호
• /
• pp.252-262
• /
• 2009

It is crucial to prevent low yields in the semiconductor industry. Since many factors affect variation in yield and they are deeply related, preventing low yield is difficult. There have been substantial researches in the field of yield prediction. Many researchers had used the statistical methods. Many studies have shown that artificial neural network (ANN) achieved better performance than traditional statistical methods. However, despite ANN's superior performance some problems such as over-fitting and poor explanatory power arise. In order to overcome these limitations, a relatively new machine learning technique, support vector machine (SVM), is introduced to classify the yield. SVM is simple enough to be analyzed mathematically, and it leads to high performances in practical applications. This study presents a new efficient classification methodology, Stepwise-SVM (SSVM), for detecting high and low yields. SSVM is step-by-step adjustment of parameters to be precisely the classification for actual high and low yield lot. The objective of this paper is to examine the feasibility of SVM and SSVM in the yield classification. The experimental results show that SVM and SSVM provides a promising alternative to yield classification for the field data.

• 한국생산제조학회지
• /
• 제20권6호
• /
• pp.679-684
• /
• 2011

Recently, STS 301L joints of side panels to frames for stainless steel rolling stock have been made out by using existing welding methods including Resistance Spot Welding, Laser Welding and Arc Welding. Most of the processes were jointed by spot welding because it is faster at welding and comparatively less expensive for investment in welding facilities than other methods. During spot welding, however, indentation of the metal surface was made due to pressure and melting property of welding. Moreover, since the melting metal was forced to periphery of the plate as indentation was made, sheet separation, which cracked apart between jointed sheets, was carried out. A slight deformation which resulted from sheet separation deteriorated the emotional quality of railway vehicles. This paper suggests that by processing conventional spot electrode tip appropriately, melting metal is able to push up around the processed part (Hollow Spot Electrode Tip) and prevent from being dislodged from first place to periphery. Consequently, sheet separation is remarkably decreased. Also, the emotional quality of appearance of stainless steel rolling stock is enhanced considerably.

• 한국기계가공학회지
• /
• 제11권4호
• /
• pp.25-30
• /
• 2012

The transportation is almost dependent on a single fuel petroleum with transportation energy dilemma. Hybrid Electric Vehicle(HEV) technology holds more advantages on efficiency improvements for petroleum consumption at the transportation. And bearing is recognized as the important component of gearbox. Gearboxes for HEV transmission have been ensured the highest reliability over some years in withstanding high dynamic loads. At the same time, the demands of lightweight design and cost minimization are required by thought-out design, high-quality material, superior production quality and maintenance. In order to design a reliable and lightweight gearbox, it is necessary to analyze bearing rating life methods between standard and different bearing companies with calculation methods for modification factors. In this paper, the influence of life time of bearings will be pointed out. Bearing contact stress and load stress distribution of HEV gearbox are obtained and compared with Romaxdesigner and BearinX. And the unequal wear of the left bearing for the gearbox intermediate shaft is investigated between simulation and test.

• 한국생산제조학회지
• /
• 제21권3호
• /
• pp.502-506
• /
• 2012

Air bearing is characterized by its extremely low friction and cleanliness such that it is widely used especially for spindles with ultra-high rotational speed at several tens of thousands rpm. This paper contributes to design of a static radial air bearing suggesting numerical analysis to anticipate its performances. The numerical analysis is an iteration method based on finite difference formulation of the Reynolds equation. A prototype air bearing has been designed and manufactured. Its load capacity has been measured and compared with the numerical solutions. The result shows good consistency between the experiment and theory, which informs that the numerical analysis can be used as an useful tool to anticipate the performances. Effects of design variables on the bearing performance have been examined by Taguchi's experimental methods using orthogonal array. Number of holes for supplying pressurized air, clearance between shaft and bearing, the hole diameter and bearing length are chosen for the design variables. The result shows that the clearance and the bearing length are the most influential variables while the others can be considered as almost negligible.

• 한국생산제조학회지
• /
• 제21권6호
• /
• pp.878-884
• /
• 2012

In this paper, we consider the sway suppression control problem for container cranes with the frictions between the trolley and the rail. If the friction effects in the system can be modelled, there is an improved potential to design controllers that can cancel the effects. The proposed control improves the trolley positioning and sway suppressing against various frictions. The proposed synthesis combines a variable structure control and the adaptive control to cope with various frictions including the unknown constants. First, the variable structure control with the simple switching action is designed, which is based on a class of feedback lineariztion methods for the fast stabilization of the under-actuated sway dynamics of container. Second, the adaptive control with a parameter estimation is designed, which is based on Lyapunov stability methods for suppressing the oscillation of the trolley travelling, especially due to Coulomb friction in the vicinity of the target position. The asymptotic stability of the overall closed-loop system is assured irrespective of variations of rope length. Simulation are shown under initial sway, external wind disturbances, and various frictions.

• 한국의류산업학회지
• /
• 제19권1호
• /
• pp.79-89
• /
• 2017

Sensory attributes and preferences that contribute to consumer satisfaction with artificial leather were measured by subjective evaluation, and subjective hand and preferences were analyzed in relation to its use. Using tactile and visual senses, 50 experts in fashion and textile industry evaluated leathers classified into two categories, suede and polyurethane coated, according to different manufacturing methods. They answered questions on subjective hand and preferences of different artificial leathers of various fashion items (jackets, purses, bags, shoes, boots, furniture, etc.), using specific adjectives to describe the hand properties. As a result, it was found that the subjective hand properties of artificial leathers were related to 'Thickness', 'Fullness/softness', 'Surface contour', 'Stickiness', and 'Elasticity'. The leather type from different manufacturing methods influenced their perceived hand and preferences relating to use. By use, different hands were preferred. The preferences for jackets and furniture of suede type leathers were related to their surface properties, whereas the preferences for items of the other type of leathers were associated with their resilience. On the other hand, in the case of polyurethane coated leathers, the preferences for jackets were significantly affected by their thickness, while those for the other items were influenced by their resilience and surface properties.

• 품질경영학회지
• /
• 제34권4호
• /
• pp.33-39
• /
• 2006

These days manufacture technology and manufacture environment are changing rapidly. By development of computer and enlargement of technique, most of manufacture field are computerized. In order to win international competition, it is important for companies how fast get the useful information from vast data. Statistical process control(SPC) techniques have been used as a problem solution tool at manufacturing process until present. However, these statistical methods are not applied more extensively because it has much restrictions in realistic problems. These statistical techniques have lots of problems when much data and factors are analyzed. In this paper, we proposed more practical and efficient a new statistical design technique which integrated data mining (DM) and statistical methods as alternative of problems. First step is selecting significant factor using DM feature selection algorithm from data of manufacturing process including many factors. Second step is finding optimum of process after estimating response function through response surface methodology(RSM) that is a statistical techniques

• 펄프종이기술
• /
• 제40권4호
• /
• pp.1-9
• /
• 2008

The utilization of non-woody fibers with the fast growing annual plants has occurred in the paper industry to replace wood and preserve environment of the earth. The non-woody fibers generally used for papermaking are paper mulberry, gampi, manila hemp, rice straw, bamboo, and coton linter etc.. Recently Kenaf has been spot-lighted for the same application. Kenaf is an annual plant of Hibiscus species of Malvaceae family. Kenaf, a rapid growing and high harvesting non-woody fiber plant, was identified as one of the promising fiber sources for the production of paper pulp. This study was carried out to investigate the pulping characteristics of Kenaf bast fiber for Hanji (traditional Korean paper) manufacturing by different pulping methods, such as alkali, alkali-peroxide and sulfomethylated pulpings. It was possible to make superior grade of Hanji. Especially sulfomethylated pulping was resulted in superior pulp in terms of higher yields and qualities in comparison to those of the other pulping methods. Hanji from sulfomethylated pulp was shown the highest brightness of over 60% and higher sheet strength. In addition, the morphological features of pulp fibers (pulp compositions) affect to the sheet properties. Therefore the effect of fiber distribution index(FDI) which was calculated from the data of Confocal laser scanning microscopy(CLSM) on the sheet properties of Kenaf Hanji was also discussed.

• 한국기계가공학회지
• /
• 제13권6호
• /
• pp.66-72
• /
• 2014

Incremental forming is a relatively novel sheet forming process, in which parts can be formed without the use of dedicated dies. In this paper, the influence of the process parameters (tool diameter, step size, feed rate, existence of a die, forming methods, and kinds of tool path) on surface roughness in the case in which parts are processed by incremental forming was discussed. Al 1050 material is used in the experiments. A table of orthogonal arrays is used to design the experiments and the ANOVA method is employed to statistically analyze the results. The obtained results show that the process parameters of tool diameter, step size, and the existence of a die have a significant effect on the surface roughness, whereas the feed rate, forming methods and kinds of tool path are insignificant.

• 한국생산제조학회지
• /
• 제23권5호
• /
• pp.498-504
• /
• 2014

Displacements and strains are very important for material evaluation as critical factors to a machine's life cycle and safety. Typically, the strain gauge has been employed to measure displacement and strain. However, this contact-type measurement method has disadvantages that are not quantified under the test conditions of a specific object shape, surface roughness, and temperature. In this paper, the measurement of deflection and flexural strain due to the three-point bending test is presented, employing Digital Image Correlation (DIC) methods. In order to ensure measurement reliability, DIC and universal test machine methods were compared by measuring the deflections and flexural strains developed by such bending tests.