• Journal of Korea Technical Association of The Pulp and Paper Industry
• /
• v.34 no.5
• /
• pp.49-55
• /
• 2002

A new method is proposed for the on-line measurement of the fiber orientation of sheet materials. The measurement of fiber orientation is very important in manufacturing paper sheets, non-woven fabrics, and glass sheets, because fiber orientation strongly affects product properties represented by, for example, dimensional stability of paper. A method developed in this research utilizes anisotropy of dielectric constants of sheet materials as a key characteristic to determine the fiber orientation. The new on-line sensor, consisting of 5 microwave dielectric resonators set in different directions, was designed to detect the fiber orientation while paper is running with high speed on a paper machine. This sensor can determine the direction and the degree of fiber orientation from the measured direction of the maximal dielectric constant and its variation, respectively. The fundamental performance of this system was examined by the static measurement of printing grade paper, which gave a satisfactory result. Then, the dynamic measurements were done at a speed of 1,000 m/min by using a high-speed test-coating machine.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.21 no.7
• /
• pp.34-42
• /
• 2022

Minimizing the backlash of gears and reducers is important for their proper and precise functioning. In this study, an automatic backlash measuring system was developed for the mass production and quality control of a military-grade reducer. The developed automatic backlash measurement system eliminates human error during the backlash measurement process. It also reduces the manufacturing time and digitizes the backlash number. The system was tested for an aircraft machine gun control reducer that required low-torque and high-precision conditions. The test results show that the torque range was 0.820-4.788 Nm. The maximum torque error is less than 0.231 N·m at 2.943 N·m, and 1.2 arcmin of the maximum backlash error with ± 0.3 arcmin of repeatability. The developed system satisfies all required conditions: torque of 1-3 Nm, torque accuracy within ± 0.5 N·m, and backlash accuracy of ± 3 arcmin.

• International Journal of Naval Architecture and Ocean Engineering
• /
• v.13 no.1
• /
• pp.848-857
• /
• 2021

This paper proposes an efficient approach based on a machine learning technique to predict the local stresses on mooring chain links. Three-link and multi-link finite element analyses were conducted for a target chain link of D107 with steel grade R4; 24,000 and 8000 analyses were performed, respectively. Two serial Artificial Neural Network (ANN) models based on a deep multi-layer perceptron technique were developed. The first ANN model corresponds to multi-link analyses, where the input neurons were the tension force and angle and the output neurons were the interlink angles. The second ANN model corresponds to the three-link analyses with the input neurons of the tension force, interlink angle, and the local stress positions, and the output neurons of the local stress. The predicted local stresses for the untrained cases were reliable compared to the numerical simulation results.

• Steel and Composite Structures
• /
• v.47 no.1
• /
• pp.103-117
• /
• 2023

This paper presents an investigation on the shear resistance of corrugated web steel beams (CWBs) with a circular web opening. A total of five specimens with different diameters of web openings were designed and tested with vertical load applied on the top flange at mid-span. The ultimate strengths, failure modes, and load versus middle displacement curves were obtained from the tests. Following the tests, numerical models of the CWBs were developed and validated against the test results. The influence of the web plate thickness, steel grade, opening diameter, and location on the shear strength of the CWBs was extensively investigated. An XGBoost machine learning model for shear resistance prediction was trained based on 256 CWB samples. The XGBoost model with optimal hyperparameters showed excellent accuracy and exceeded the accuracy of the available design equations. The effects of geometric parameters and material properties on the shear resistance were evaluated using the SHAP method.

• Transactions of the Korean Society of Machine Tool Engineers
• /
• v.14 no.4
• /
• pp.1-6
• /
• 2005

In this study, on automatic ultrasonic testing system is used to detect defects of flawtest specimen. We study the effects of the difference of ultrasonic damped rate of the different materials on the judgment of defects. The results indicate that the difference of sensitivity compensating quantity is 2dB, and the judgment is correct over $90\%$ when a specimen is judged as a defect when it exceeds third grade.

• Journal of The Korean Association of Information Education
• /
• v.25 no.2
• /
• pp.327-335
• /
• 2021

This study verified the effectiveness of machine learning education programs focused on data labeling as an educational method for improving computational thinking of elementary school students. The education program was designed and developed based on the results of a preliminary demand analysis conducted on 100 elementary school teachers. In order to verify the effectiveness of the developed education program, 17 sixth-grade students attending K Elementary School were given 2 classes per day for a total of 6 weeks. In order to measure the effect of the training on improving computational thinking, the educational effects were analyzed by conducting pre-post-inspection using the "Beaver Challenge". According to the analysis, machine learning education focused on data labeling contributed to improving computational thinking of elementary school students.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.8
• /
• pp.49-58
• /
• 2023

Dropouts of students not only cause financial loss to the university, but also have negative impacts on individual students and society together. To resolve this issue, various studies have been conducted to predict student dropout using machine learning. This paper presents a model implemented using DNN (Deep Neural Network) and LGBM (Light Gradient Boosting Machine) to predict dropout of university students and compares their performance. The academic record and grade data collected from 20,050 students at A University, a small and medium-sized 4-year university in Seoul, were used for learning. Among the 140 attributes of the collected data, only the attributes with a correlation coefficient of 0.1 or higher with the attribute indicating dropout were extracted and used for learning. As learning algorithms, DNN (Deep Neural Network) and LightGBM (Light Gradient Boosting Machine) were used. Our experimental results showed that the F1-scores of DNN and LGBM were 0.798 and 0.826, respectively, indicating that LGBM provided 2.5% better prediction performance than DNN.

• Proceedings of the Korea Technical Association of the Pulp and Paper Industry Conference
• /
• 2003.11a
• /
• pp.230-248
• /
• 2003

In this work an integrated model for paper plants combining wet-end and dry section is developed and a model predictive control scheme based on the plant model is proposed. Closed-loop process identification method is employed to produce a state-space model. Thick stock, filler flow, machine speed and steam pressure are selected as Input variables and basis weight, ash content and moisture content are considered as output variables. The desired output trajectory is constructed in the form of 1st-order dynamics. Results of simulations for control of grade change operations are compared with plant operation data collected during the grade change operations under the same conditions as in simulations. From the comparison, we can see that the proposed model predictive control scheme reduces the grade change time and achieves stable steady-state.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.12 no.8
• /
• pp.3430-3438
• /
• 2011

The purpose of this study is to set the Korean mandarine quality standards based on consumer preference survey. Until now the Korean mandarine's quality standards has been based on the fruit size. The Korean mandarine's quality in agricultural cooperative, citrus agricultural cooperative federation, and some agricultural corporation has been selected in accordance with its own brand of quality grade using a non-destruction sorting machine. But, setting the Korean mandarine's quality standards has been based on the convenient and routine method rather than the scientific and objective method, consumer's preference. According to the grade contents, the highest grade brand was required more than sugar $12^{\circ}Bx$ and less than acid 1.0% and the following grade brand was required more than sugar $11^{\circ}Bx$ and less than acid 1.0% uniformally. Thus, in this study, based on the consumers' preference of Korean mandarine, 4-level grades of sugar and 4-level grades of acidity were divided into the total 16-level grades. Based on them, 5-level grades were set.

• Geomechanics and Engineering
• /
• v.24 no.4
• /
• pp.323-335
• /
• 2021

This paper aims to estimate the range of the excavation damaged zone (EDZ) formation caused by the tunnel boring machine (TBM) advancement through dynamic three-dimensional large deformation finite element analysis. Large deformation analysis based on Coupled Eulerian-Lagrangian (CEL) analysis is used to accurately simulate the behavior during TBM excavation. The analysis model is verified based on numerous test results reported in the literature. The range of the formed EDZ will be suggested as a boundary under various conditions - different tunnel diameter, tunnel depth, and rock type. Moreover, evaluation of the integrity of the tunnel structure during excavation has been carried out. Based on the numerical results, the apparent boundary of the EDZ is shown to within the range of 0.7D (D: tunnel diameter) around the excavation surface. Through series of numerical computation, it is clear that for the rock of with higher rock mass rating (RMR) grade (close to 1st grade), the EDZ around the tunnel tends to increase. The size of the EDZ is found to be direct proportional to the tunnel diameter, whereas the depth of the tunnel is inversely proportional to the magnitude of the EDZ. However, the relationship between the formation of the EDZ and the stability of the tunnel was not found to be consistent. In case where the TBM excavation is carried out in hard rock or rock under high confinement (excavation under greater depth), large range of the EDZ may be formed, but less strain occurs along the excavation surface during excavation and is found to be more stable.