Earticle

제조업에 있어 효율적인 제조공정을 위해 기업의 생산성 향상 및 원가절감은 현 제조기업의 절실한 과제 이다. 효율적인 제조공정 설계와 생산성향상을 위해 동작분석 및 가상현실기반의 테스트베드 연계시스템을 활용하 여 불필요한 요소제거를 통해 효율적인 제조공정 방법을 연구한다. 가상의 사전 공정배치 시뮬레이션을 통한 최적 화 검증 후 작업라인을 구축할 경우 추가적인 공정 수정 또는 보완에 소요되는 비용을 절감하고, 이는 결과적으로 기업의 제조경쟁력 향상이라는 실질적 효과를 도출할 수 있다. 본 연구를 통해 설계기간 단축, 설비개선, 자동화, 인간공학 실현, 프로세스 개선 등을 하여 생산성 향상, 품질 향상, 비용절감뿐만 아니라 안전 및 건강에 대한 직원 만족을 달성할 수 있다.

For manufacturing industry to achieve efficient manufacturing process, improving a company’s productivity and reducing cost is a required task of the current manufacturing companies. For design of efficient manufacturing process and improvement of productivity, efficient manufacturing process method is studied through removal of unnecessary elements analyzing motions and using test-bed linkage system based on virtual reality. When building process line after verifying optimization through virtual process arrangement simulation in advance, it is available to save additional cost for process modification or supplement to result in practical effects of improving a company’s manufacturing competition. Reducing design period, improving equipment, making automation, realizing ergonomics, and improving process through this study, it will be available to achieve improvement of productivity, enhancement of quality, cost reduction as well as employees’ satisfaction for safety and health.

Forged part made of Cold heading quality wire materials are used for automotive brake systems. The cost reduction of forged products is a major issue because of the strict shape change. A series of studies were conducted to minimize the cost of EPB spindle process among brake parts. In order to reduce the material cost, heat treatment-abbreviated material was applied and the formability on the processes was verified by the ductile fracture theory. In addition, the causes of shape fixation and die life degradation were analyzed using the numerical simulation. The process cost has been minimized by re-designing process, changing the product shape, and the die material.

Enterprises to achieve global optimization and better diversification and personalized product production scheduling, from the unity of the tasks and resources manufacturing, based on the perspective of global optimization in production scheduling method. With global production scheduling of matching and process model resources as the foundation for the realization of enterprise process production scheduling optimization and level continue working process decoupling point, this paper manufacturing task and manufacturing process is divided into generalization and differential theory , applications from manufacturing process optimization results obtained path, based on the manufacturing process optimization mathematical model to study the production scheduling, and by using the improved genetic algorithm, and finally for example. Above scheduling method simplified the scheduling process, reduce the scheduling of difficulty, shorten the time scheduling, and finally achieve enterprise resource efficient use.

For complexity of process-level manufacturing process in enterprise workshop, Lacking of the researching about the impact on production scheduling of workshop which made by manufacturing process ;To simplify the process of the workshop process-level processing, we can adopt a method that divide the processing operations into two kinds :universal and differentiated method; proposing the concept of process decoupling point, build a process-level manufacturing task decomposition model ;Using the method of directed hyper-graph to get the optimal processing path, achieving the simplification of the production scheduling process on the considering of the ability of enterprise resource ; In order to make the shortest time and the lowest cost and meet the individual needs of clients, making reasonable use of the resources, providing the technical support for the workshop production scheduling.

Black ginseng is derived from a repeated steaming/drying process; however, the black ginseng manufacturing process is not well established. To determine the steaming and drying optimum conditions for producing high levels of biologically active substances such as acidic polysaccharides, ginsenoside Rg3, and polyphenols and low levels of benzopyrene, response surface methodology was used with temperature and time as independent variables. As steaming temperature/time increased, acid polysaccharide, polyphenol, and benzopyrene content increased; ginsenoside Rg3 levels increased with steaming time at temperatures ${\geq}100^{\circ}C$. As drying temperature/time increased, ginsenoside Rg3, benzopyrene, and acidic polysaccharide content also increased. These substances showed a nonlinear canonical form, whereas the phenolic component showed a linear canonical form. Optimum conditions were determined to be steaming at $113.04^{\circ}C$ for 18 h and drying at $100^{\circ}C$ for 8.03 h, resulting in 0.75 mg/g ginsenoside Rg3, 13.72 mg% acidic polysaccharide, 0.26 ppb benzopyrene, and 3.24 mg% polyphenol.

A number of surface mount technology (SMT) process variables including land design are considered for minimizing tombstone defect in flexible printed circuit assembly in high volume manufacturing. As SMT chip components have been reduced over the past years with their weights in milligrams, the torque that once helped self-centering of chips, gears to tombstone defects. In this paper, we have investigated the correlation of the assembly process variables with respect to the tombstone defect by employing statistically designed experiment. After the statistical analysis is performed, we have setup hypotheses for the root causes of tombstone defect and derived main effects and interactions of the process parameters affecting the hypothesis. Based on the designed experiments, statistical analysis was performed to investigate significant process variable for the purpose of process control in flexible printed circuit manufacturing area. Finally, we provide beneficial suggestions for find-pitch PCB design, screen printing process, chip-mounting process, and reflow process to minimize the tombstone defects.

Recently, TFT LCD (thin film transistor liquid crystal display) manufacturing industry is more concerned with the ways of cleaning large TFT LCD´s with high pixed density than ever Ultrasonic cleaners with high frequencies like 1MHz (megasonic cleaners) are effective in removing very small particles without causing mechanical damage to the surface. In this study a megasonic cleaner for TFT LCD manufacturing process is developed and the performance is evaluated through experiments. The experimental results show that the developed magasonic cleaners is effective in removing very small particle from the LCD panel.

Using central composite design, we have designed optimization of the manufacturing of processed butter. And response surface analysis by least-square regression was used Statistical Analysis System(SAS). Central composite design can be achieved by response surface techniques that allow flexibility in modeling and analysis. Response surface methodology(RSM) was used to optimize hardness(%) using as independent variables; the content of butter($X_1$), ranging from 50 to 90(%), the content of soybean oil($X_2$), from 0 to 20(%), and the hydrogenated soybean oil($X_3$) from 0 to 4(%). The results on the regression coefficients calculated for overrun by response surface by least-square regression(RSREG) were followed. It was considered that the linear regression was significant(p<0.01). As for the processed butter, the regression model equation for the hardness(Y, %) to the change of an independent variable could be predicted as follow: $Y=60.88-8.92X_2-{29.3X_2}^2$. The optimal for the manufacturing of processed butter were determined at the content of butter of 88.22%, soybean oil of 6.71% and hydrogenated soybean oil of 2.36%, respectively. Optimum compositions were resulted in hardness of 65.78 N. Finally the reference sample(Butter in the morning, Seoul Dairy Co-op.) and processed butter manufacturing under the optimal conditions were compared with spreadability test. The spreadability scores result from reference sample and butter under optimal conditions was not found a significant difference.

In this study, the directed energy deposition (DED) process was applied to additively manufacture cobalt-chro- mium alloy materials, which are commonly used in molds, dental/orthopedic medical applications, vehicles, and aircraft. The quality of the deposition obtained by using the DED technology is affected by process parameters. Therefore, the aim of this experiment, was to evaluate the microstructure changes caused by varying parameters such as the laser power, laser header moving speed, and powder feeding rate, with the goal of improving and opti- mizing the stacking quality. The shape, heat-affected zone, and dilution ratio of the microstructure were analyzed us- ing an optical microscope. Furthermore, the global energy density(GED) was calculated because it affects the qual- ity of the 3D-printed product. The optimal process conditions for the DED method determined by calculating the GED under various process conditions.

This study presents a case of successfully implementing an AI-based digital twin virtual convergence factory in a 30-year-old automotive parts cold forging company to optimize manufacturing processes. A Unity-based 3D digital twin environment was constructed for five factory buildings, and an LSTM-based AI predictive maintenance model was developed. The system achieved 100% of all PQCD (Productivity, Quality, Cost, Delivery) indicators, reduced equipment failure response time by 85% (from 1-2 hours to within 10 minutes), and decreased VR training time by 50%. The 4-layer architecture (Data & Connectivity, Infra & Platform, Digital Twin & Visualization, AI Analytics & Simulation) enabled real-time data integration within 1 second using OPC-UA protocol. This research demonstrates the feasibility of digital transformation for SMEs in the root industry and provides a practical reference model.

본 연구는 국내 철강기업(S사)의 쾌삭 스테인리스강 중형 압연 선재 제조공정 빅데이터를 활용하여 생산품질 예측 모형을 개발하고, 이를 기반으로 불량률을 최소화시키기 위한 최적 공정조건을 탐색하는 절차를 제안한다. S사의 쾌삭 스테인리스강 압연 선재 제조공정은 크게 제강, 열간압연과 열처리/산세 공정으로 이루어지며, 연구 데이터는 생산관리시스템(MES)으로부터 37종의 공정변수 데이터를 수집하고 문헌조사와 현장 전문가들의 인터뷰를 통해 총 12종의 파생변수를생성하여 반영하였다. 본 연구는 총 49종의 변수에 대한 데이터 전처리 및 탐색적 분석, 차원 축소, 기계학습 모델링, 공정조건 최적화 단계로 수행되었다. 탐색적 데이터 분석을 통해 데이터의 속성과 패턴을 파악하고, 전처리 단계에서는 이상치․결측치 제거를 수행하였다. 차원 축소 단계에서는 생산품질 예측 모형의 다중공선성 등의 문제를 방지하고 효율성을향상시키기 위해 상관분석과 변수 중요도를 이용하여 품질 영향도가 높은 변수를 식별하였다. 기계학습 모델링 단계에서는 다양한 지도학습 모형들에 대한 예비적 실험에서 유의미한 수준의 성능을 보인 로지스틱 회귀(Logistic regression) 모형, 서포트벡터머신(SVM: Support Vector Machine), 랜덤포레스트(Random Forest) 모형, XGBoost(Extreme Gradient Boosting) 모형을 대상으로 겹 교차 검증을 통해 각 후보 모형의 하이퍼파라미터들(Hyperparmeters)을최적화(Tuning)하였다. 또한, 오분류표 기반 성능 지표들과 대수손실(Logarithmic loss)을 종합적으로 고려하여 후보모형들을 비교/평가하고, 가장 우수한 성능을 보인 XGBoost 모형을 생산품질 예측 모형으로 선정하였다. 마지막으로, 최적 공정조건 탐색 단계에서는 개체군집최적화(PSO) 알고리즘이 이용되었으며, 이때, 개발된 생산품질 예측 모형(XGBoost) 이 적합도 함수로 적용되었다.

Based on the manufacturing big data of a free-machining 303 series stainless steel mediumsized wire rod, this study proposes a production-quality prediction model and a methodology for searching operating conditions maximizing the likelihood of producing adequate products. The research data include 37 types of operation variables for 3,542 lots collected from the manufacturing execution system (MES). Based on the MES data, 12 types of derived variables are generated through a literature review and interviews with field experts. Data attributes and patterns are identified, and outliers and missing values are removed in the exploratory analysis and preprocessing. In dimensionality reduction, variables with high-quality influence are identified by variable importance measures to prevent multicollinearity and enhance the efficiency of machine learning models. A preliminary study selected the logistic regression, support vector machine, random forest, and extreme gradient boosting (XGBoost) models as the candidates for the production-quality prediction model in data modeling. Their hyperparameters are tuned by the grid and random search methods based on k-fold cross-validation. In addition, the XGBoost shows the best classification performance for the test dataset. Finally, based on the XGBoost model, a swarm optimization algorithm searches the operating conditions to maximize the likelihood of producing adequate goods.

In this study, we conducted laser welding by using remote scanner that is 5J32 aluminum alloy to observe the mechanical properties and optimize welding process parameters. As the control factors, laser incident angle, laser power and welding speed were set and as the result of weldablility, tensile shear tests were performed. ANOVA (Analysis of Variation) was also carried out to identify the influence of process variables on tensile shear strength. Strength estimation models were suggested using regression alnalysis and 2nd order polynomial model had the best estimation performance. In addition optimal welding condition was determined in terms with wedalility and productivity using objective function and fitness function. Final optimized welding condition was laser power was 4 kW, and welding speed was 4.6m/min.

데이터 시각화 기술은 다양한 데이터와 그 분석 결과를 쉽게 이해할 수 있도록 도와줌으로써 제조현장과 같은 실제 산업현장에서도 그 유용성이 기대되고 있다. 제조현장에서 발생하는 대량의 데이터는 제조 기술의 표준화를 위한 기반 데이터가 될 수 있으며 제조공정의 개선을 위하여 매우 중요한 역할을 수행할 수 있다. 본 논문에서는 실험계획법과 데이터 분석 기반의 제조공정 최적화를 위한 정보 시각화 방법을 제안한다. 데이터 분석 결과의 정보 시각화를 통하여 작업 현장에 이해하기 쉬운 분석 결과를 제공함으로써 다양한 불량원인을 감소시키고 제조공정을 개선시킬 수 있다.

Data visualization technology helps people easily understand various data and its analysis result, so usefulness of it is expected in the real industrial manufacturing sites. The large amount of data which is occurred at the manufacturing sites is able to fulfill very important roll to improve the manufacturing process. In this paper, we propose an information visualization for the manufacturing process optimization based on design of experimental and data analysis. The manufacturing process may be improved and be reduced cause of faulty by providing the easy-process analysis to understand the operation site through the information visualization of data analysis result.

This study aimed to solve the field fatigue failure rate (10%) of SCM440 through-bolts in hydraulic breakers by optimizing the manufacturing process. FMECA results indicated that the root cause was not material limitations but defects in thread machining and heat treatment processes. Consequently, a process improvement was implemented by adopting a 'post-heat-treatment thread rolling' method to induce compressive residual stress at the thread root. To statistically verify the reliability of the improved components, an accelerated life test (ALT) was designed for 300,000 cycles, targeting a B10 life of 165 hours at 80% confidence. Test results showed that all four specimens passed without failure, statistically validating the life target. This proves that manufacturing process optimization is a valid engineering solution to reduce field failure rates.

일반적인 폴리프로필렌(PP) 폼은 랜덤 PP를 사용하여 제조되나 랜덤 PP보다 결정성이 높은 호모 PP를 사용할 경우 PP 폼의 기계적 물성을 향상시킬 수 있다. 그러나 이러한 소재의 결정성 증가는 가공성을 저하시킴으로써 발포 폼을 제조하는데 다양한 공정상의 어려움을 야기한다. 본 연구에서는 자동차 내외장재 용도로 가장 많이 사용되는 PP를 베이스 원료로 하여 전자선 가교와 화학발포제를 이용하여 PP 폼을 제조하는 과정에서 호모 PP를 적용하였으며, 이에 의한 가공성의 저하는 발포제 마스터배치(M/B) 최적화, 압출량 최적화 및 사이드 피딩 시스템을 적용하여 해결하였다.

<P>Typical polypropylene (PP) foam is manufactured using random PP, but the mechanical properties of PP foam can be improved when homo PP, which is more crystalline than random PP, is used. However, the increase in the crystallinity of such a material causes various process difficulties in manufacturing the foamed foam by lowering the workability. In this study, homo PP was applied in the process of making PP foam by using electron beam crosslinking and chemical foaming agent using polypropylene which is the most used for automobile interior and exterior materials. The degradation of the workability in the process was overcome by the optimization of electron beam crosslinking process and side feeding systems.</P>

일반적인 폴리프로필렌(PP) 폼은 랜덤 PP를 사용하여 제조되나 랜덤 PP보다 결정성이 높은 호모 PP를 사용할 경우 PP 폼의 기계적 물성을 향상시킬 수 있다. 그러나 이러한 소재의 결정성 증가는 가공성을 저하시킴으로써 발포 폼을 제조하는데 다양한 공정상의 어려움을 야기한다. 본 연구에서는 자동차 내외장재 용도로 가장 많이 사용되는 PP를 베이스 원료로 하여 전자선 가교와 화학발포제를 이용하여 PP 폼을 제조하는 과정에서 호모 PP를 적용하였으며, 이에 의한 가공성의 저하는 발포제 마스터배치(M/B) 최적화, 압출량 최적화 및 사이드 피딩 시스템을 적용하여 해결하였다.

Typical polypropylene (PP) foam is manufactured using random PP, but the mechanical properties of PP foam can be improved when homo PP, which is more crystalline than random PP, is used. However, the increase in the crystallinity of such a material causes various process difficulties in manufacturing the foamed foam by lowering the workability. In this study, homo PP was applied in the process of making PP foam by using electron beam crosslinking and chemical foaming agent using polypropylene which is the most used for automobile interior and exterior materials. The degradation of the workability in the process was overcome by the optimization of electron beam crosslinking process and side feeding systems.

폐차잔재물중 가연성 성분을 시멘트 제조공정의 열원으로 활용하기 위한 연구가 현재 활발히 진행 중에 있다. 그러나 이러한 폐차잔재물을 시멘트 제조공정의 보조열원으로 사용하기 위해서는 에너지활용에 대한 경제성 및 환경기준과 더불어 시멘트 제조공정의 안정적 조업조건에 대한 검토가 이뤄져야만 하며 특히 폐차잔재물의 연소시 시멘트 제조공정에서의 염소물질의 거동특성에 대한 연구가 반드시 선행되어야만 한다. 따라서 본 연구에서는 이러한 폐차잔재물의 시멘트 제조공정에 대한 열원활용의 최적조건을 도출하고자 현장적용 전 킬른내 염소성분의 제어를 위한 Bypass unit의 Cyclone에 대한 전산모사연구를 수행하였으며 이러한 전산모사결과를 바탕으로 실제 시멘트 제조공정에 대한 폐차잔재물의 현장적용 실험을 수행하였다. 또한 본 연구로부터 시멘트 제조공정에 대한 폐차잔재물의 열원활용 가능성을 확인할 수 있었으며 폐차잔재물의 투입시 시멘트 제조공정의 안정적 조업조건을 확보할 수 있었다.

STS316L 합금의 Laser powder bed fusion 공정 최적화를 위하여 Laser power, Scan speed 및 Hatching distance의 공정조건을 제어하면서 투입 레이저 에너지 밀도와 조형체의 상대밀도와의 상관관계를 연구했고, 최적조건으로 제작된 조형체의 적층 방향에 따른 인장특성 변화를 분석했다. STS316L 분말을 에너지밀도가 55.6 J/mm³, 83.3 J/mm³ 및 111.1 J/mm³인 조건에서 적층 성형한 결과, 투입 레이저 에너지밀도가 83.3 J/mm³이며, Power 및 Scan speed 각각 225 W, 1000 mm/s인 조건에서 가장 안정적으로 고밀도 STS316L 샘플을 제작할 수 있었다. 최적공정조건을 이용해 적층 방향과 인장방향이 각각 0°, 45°, 90°인 인장시험편을 제작하여 인장특성을 비교한 결과 적층 방향과 인장방향이 수직인 시험편의 항복강도, 인장강도 및 연신율이 가장 우수한 것이 확인되었다. 적층 방향과 수직 방향으로의 이방성을 가지는 기공 및 Lack of fusion 결함이 응력집중을 야기하여 인장특성을 열화 시키기 때문인 것으로 추정된다.

To optimize the process parameters of laser powder bed fusion process to fabricate the high density STS316L alloy, the effect of laser power, scanning speed and hatching distance on the relative density was studied. Tensile properties of additively manufactured STS316L alloy using optimized parameters was also evaluated according to the build direction. As a result of additive manufacturing process under the energy density of 55.6 J/mm³, 83.3 J/mm³ and 111.1 J/mm³, high density STS316L specimens was suitably fabricated when the energy density, power and scan speed were 83.3 J/mm³, 225 W and 1000 mm/s, respectively. The yield strength, ultimate tensile strength, and elongation of STS316L specimens in direction perpendicular to the build direction, show the most competitive values. Anisotropic shape of the pores and the lack of fusion defects probably caused strain localization which result in deterioration of tensile properties.

Purpose: To derive the optimal conditions for the Rubber based colloidal suspension manufacturing process, which made using a stirrer, to apply the mixture design method. Methods: We used two process component and one process variable Mixture design to derive the optimal conditions for the process. The response variables were selected for rotational viscometer measures which can represent Rubber based colloidal suspension quality. The input variables were selected as the values of rubber-organic solvent expressed in proportions as process components and stirring amount as a process variable which are controllable factors in the process. Results: Based on the results of the experiment, rubber and organic solvent and the interaction between stirring amount and rubber and the interaction between stirring amount and rubber and organic solvent were significant. Reproducibility of the regression model was confirmed by the observation that the values obtained from the reproducibility experiment fell within the confidence interval. Additionally, the model predictions were found to be in close agreement with the field measurements. Conclusion: In this study, a regression model was developed to predict the viscosity change of colloidal suspensions based on the proportion of rubber based colloidal suspension. The developed regression model can lead to improved product quality.