Fault detection in electromechanical systems plays a significant role in product quality and manufacturing efficiency during the transition to smart manufacturing. Because collecting a sufficient number of datasets under faulty conditions of the system is challenging in practical industrial sites, unsupervised fault detection methods are mainly used. Although fault datasets accumulate during machine operation, it is not straightforward to utilize the information it contains for fault detection after the deep learning model has been trained in an unsupervised manner. However, the information in fault datasets is expected to significantly contribute to fault detection. In this regard, this study aims to validate the effectiveness of the transition from unsupervised to supervised learning as fault datasets gradually accumulate through continuous machine operation. We also focus on experimentally analyzing how changes in the learning paradigm of the deep learning model and the output representation affect fault detection performance. The results demonstrate that, with a small number of fault datasets, a supervised model with continuous outputs as a regression problem showed better fault detection performance than the original model with one-hot encoded outputs (as a classification problem).

A cold roll-bonding (CRB) process is applied to fabricate an AA1050/AA5052 layered sheet. In the process, commercial AA1050 and AA5052 sheets of 1 mm thickness, 40 mm width and 300 mm length are stacked onto each other, and then reduced to a thickness of 0.5 mm through a 2-pass cold rolling process without lubricant. The roll-bonded AA1050/AA5052 layered sheet is then annealed for 1 h at various temperatures from 200 to 400 °C. The specimens annealed at temperatures below 250 °C showed a typical deformation structure in which the grains were elongated along the rolling direction. However, the specimens annealed at temperatures higher than 300 °C exhibited recrystallization structures in both the AA1050 and AA5052 regions. All the roll-bonded and subsequently annealed specimens showed an inhomogeneous distribution of hardness in the thickness direction, in which the hardness in the AA5052 regions was higher than that in the AA1050 regions. As the annealing temperature increased, the tensile and yield strengths decreased and the elongation increased gradually. The mechanical properties were compared to those of commercial AA1050 and AA5052 materials and CRBed AA5052-2L materials from a previous study.

As environmental concerns escalate, the increase in recycling of aluminum scrap is notable within the aluminum alloy production sector. Precise control of essential components such as Al, Cu, and Si is crucial in aluminum alloy production. However, recycled metal products comprise various metal components, leading to inherent uncertainty in component concentrations. Thus, meticulous determination of input quantities of recycled metal products is necessary to adjust the composition ratio of components. This study proposes a stable input determination heuristic algorithm considering the uncertainty arising from utilizing recycled metal products. The objective is to minimize total costs while satisfying the desired component ratio in aluminum manufacturing processes. The proposed algorithm is designed to handle increased complexity due to introduced uncertainty. Validation of the proposed heuristic algorithm's effectiveness is conducted by comparing its performance with an algorithm mimicking the input determination method used in the field. The proposed heuristic algorithm demonstrates superior results compared to the field-mimicking algorithm and is anticipated to serve as a useful tool for decision-making in realistic scenarios.

본 연구는 자동차부품제조기업종사자들의 조직공정성이 직무만족에 미치는 영향에 대한 조직후원인식의 매개효과를 규명 하는데 있다. 선 행연구를 기반으로 한 설문지를 자동차부품제조기업종사자를 대상로 배 포하여 148부가 회수되었으며, SPSS 20.0 통계 분석 프로그램을 사용 하여 통계 분석을 실시하였다. 분석 결과, 조직공정성은 직무만족에 유 의미한 정(+)의 영향을 미쳤으며, 조직후원인식은 매개 효과가 있는 것 으로 나타났다. 이에 따라 첫째, 조직공정성이 강화될수록 직무만족에 미치는 영향이 있는 것을 알 수 있었다. 둘째, 조직후원인식이 조직공정 성과 직무만족간의 매개효과를 확인할 수 있었고 이는 조직구성원들의 노력과 성과에 가치를 두고 복지를 증대시키면 조직후원인식을 통해서 직무만족을 더 잘 할 수 있음을 의미한다. 이는 조직후원인식을 통하여 조직공정성을 강화함으로써 직무만족을 향상 할 수 있는 가능성을 시사 한 것으로 볼 수 있다.

고분자 분리막의 대표적인 형태 중 하나인 평판형 분리막은 제조가 용이하여 실험실에서 분리막 소재 연구에서부 터 실제 상용 분리막 생산에 이르기까지 널리 활용되는 분리막의 형태이다. 정밀여과 및 한외여과 등에 사용되는 평판형의 다공성 고분자 분리막은 주로 상분리 공정을 통해 제조할 수 있으며, 여기에는 비용매 유도 및 증기 유도 상분리 공정이 활 용된다. 그러나 상분리 공정 특성상 주변 환경과 실험자에 따라 샘플 간 편차가 쉽게 발생하여 재현성의 확보가 어려운 점이 있다. 따라서 개발된 제조기술을 스케일업 및 재현성 확보를 위해 제어된 환경에서 연속식 대면적 제조가 가능한 롤투롤 제 조장치가 필요하며, 본 연구에서는 실험실 스케일의 제조기술을 나이프 및 슬롯다이 롤투롤 공정으로 스케일 업 했을 때 나 타나는 제조 환경 차이에 따른 분리막의 특성 변화를 비교하였다. 최종적으로 연속식 제조공정 인자에 대한 최적화를 통해 대면적 제조 시 분리막의 균일성을 확보하였다.

We report the effect of plastic deformation on the thermoelectric properties of n-type Bi2Te2.5Se0.5 compounds. N-type Bi2Te2.5Se0.5 powders are synthesized by an oxide-reduction process and consolidated via sparkplasma sintering. To explore the effect of plastic deformation on the thermoelectric properties, the sintered bodies are subjected to uniaxial pressure to induce a controlled amount of compressive strains (-0.2, -0.3, and -0.4). The shaping temperature is set using a thermochemical analyzer, and the plastic deformation effect is assessed without altering the material composition through differential scanning calorimetry. This strategy is crucial because the conventional hotforging process can often lead to alterations in material composition due to the high volatility of chalcogen elements. With increasing compressive strain, the (00l) planes become aligned in the direction perpendicular to the pressure axis. Furthermore, an increase in the carrier concentration is observed upon compressive plastic deformation, i.e., the donorlike effect of the plastic deformation in n-type Bi2Te2.5Se0.5 compounds. Owing to the increased electrical conductivity through the preferred orientation and the donor-like effect, an improved ZT is achieved in n-type Bi2Te2.5Se0.5 through the compressive-forming process.