Effect of initial microstructure and heating rate on microstructure evolution during intercritical annealing and mechanical properties in ferrite based microalloyed steel

홍석환 포항공과대학교 철강에너지소재대학원 2023 국내석사

철강공정에서의 메타스테이블 가스의 효율적 활용 : Effective Use of Metastable Gas Species for Metallurgical processes

박우일 포항공과대학교 철강대학원 2014 국내박사

Certain elements and compounds can produce the metastable products, enhancing reaction rates or easing the operation conditions. By applying the metastable reactants, the composition of products may deviate appreciably from that formed with the stable alone. Activated carbon from methane decomposition and gaseous ammonia might be considered as such metastable reactant species that lead to reaction product compositions differing from the equilibrium under certain conditions as well as possibly enhancing the reaction rates in the metallurgical processes. This study presents several experimental results giving support to the existence of the influences of these species on the reaction products and rates. First, gaseous ammonia enhanced solid solution nitriding. Second, CH4 containing gas enhanced carburization rate. Finally, CH4 containing gas enhanced reduction rate in the suspension reduction process. Chapter II investigates the NH3 and N2 solution nitriding of Fe-Cr alloys under various conditions. Comparing the N contents after 1 h nitriding, NH3 nitriding is found to be about 5 times faster than that of N2 (0.1 MPa) nitriding under the present experimental conditions. Chapter III investigates the carburization rate of iron in N2-11% CH4 gas mixture at 1573K. The rate was about 40 times faster than that by 50% CO gas mixture. Chapter IV investigates the reduction and carburization of the fine Chinese Benxi magnetite ore in the suspension reduction with CH4-containing gas. A very high fractional reduction was achieved by CH4-containing gas at 1573 K compared with that by H2 or CO gas. These selective experimental results and discussions show that we can extend the thermodynamic limitation as well as enhance the reaction rates by involving metastable species in the metallurgical processes. 특정 원소나 화합물은 메타스테이블 상태의 생성물을 만들며, 이는 반응속도를 향상시키거나 공정의 조건을 완화하는데 이용된다. 이런 메타스테이블 상의 반응물을 공정에 적용함으로써 단순히 안정화된 반응물만을 사용했을 경우보다 생성물의 조성이 상이하게 된다. 특히, 메탄과 암모니아에 의한 Activated 탄소나 질소의 경우 이처럼 평형상태에서의 생성물의 조성이 상이하며 철강공정 상의 반응속도를 향상하는데 이용될 수 있다. 이번 연구는 몇 가지 실험 결과들을 토대로 이런 메타스테이블 가스가 실제 생성물과 반응속도에 미치는 영향을 살펴보았으며, 첫째로 암모니아 가스를 활용한 고용체상에서의 침질과정을 통해 반응속도 향상을 살펴보았으며, 둘째로, 메탄가스를 통한 가스 침탄 과정을 통해 철의 침탄속도 향상을 확인하였다. 마지막으로 메탄을 포함한 환원가스의 조성을 통해 서스펜젼 환원 공정에서의 그 효용성을 확인하였다. Chapter II는 이 중 첫 번째 내용으로 암모니아 가스와 질소가스의 비교를 통해 Fe-Cr 합금의 다양한 상태에서의 침질과정을 살펴보았이며, 1시간의 암모니아를 통한 고용체 상에서의 침질과정을 질소 1기압의 침질과정과 비교해본 결과 약 5배 정도의 반응속도 향상을 확인하였다. Chapter III는 11% 메탄가스를 이용하여 1300도에서 철의 침탄과정을 살펴보았으며, 같은 시간의 약 900초 이 후에 11% 메탄을 이용하여 침탄된 탄소의 양이 50% CO를 이용한 침탄과정에 비해 약 40배 정도 높은 탄소의 농도 차이를 보였다. Chapter IV는 서스펜젼 환원 공정을 가상하여 중국 본계광의 환원과 침탄과정을 살펴보았으며, 메탄을 이용한 환원이 수소나 일산화탄소에 비하여 훨씬 효과적이라는 것을 확인 할 수 있었다 이런 몇 가지 선택적인 실험결과나 논의과정을 통하여 우리는 열역학적인 평형상태에서의 한계를 극복하거나 반응속도를 향상시키는데 이런 메타스테이블한 상들의 활용할 가치가 있다는 점을 확인하였고, 이를 위해 분명히 반응 전에 메타스테이블 상이 Cracking되어 반응성을 잃는 부분, Sooting과 같은 카본 deposition 문제 등 공정상의 문제점, Nitride나 Carbide, Graphite의 생성과 같은 반응성의 컨트롤 등, 기술적으로 해결해야 할 부분이 많다는 점을 확인하였다. 하지만, 이를 고려하더라도 향 이런 Metastable 물질 활용에 대한 연구는 향 후 새로운 철강공정을 개발하는데 중요한 밑거름이 될 것이라 여겨진다.

A Study to Correlate between Electrochemical Desulfurization of Molten Iron using CaO-A2O3-MgO Molten Slag and Applied Electric Current

김동현 포항공과대학교 철강에너지소재대학원 2022 국내박사

현대의 철강 산업에서 극청정강 제품에 대한 수요가 꾸준히 증가함과 동시에, 저급 원료 사용량 증대로 용선/용강 (용철) 내 S 등 다량의 불순물을 초래하는 문제를 낳고 있다. 한편, 기존 KR 등의 탈황능은 이미 한계에 다다랐고, 2차 정련에 또한 복황 저감에 대한 기술이 필요한 상황이다. 따라서, 극청정강 제조를 위한 철강 공정 내 새로운 정련 기술 도입이 요구된다. 기존의 용철 탈황 반응의 경우, 생석회(CaO) 기반의 슬래그에 의해 용철 중 S이 환원되어 슬래그 내 CaS 형태로 용해되게 된다. 이를 바탕으로 전기에너지를 이용한 탈황의 전기화학적 반응을 극대화하는 방법이 등장하였고, 많은 선행 연구자들이 전기화학적 탈황의 가능성에 대해 연구해왔다. 대부분의 연구에서 전기 인가에 따른 추가 탈황을 확인하였고, 반응 기구 규명과 속도론적 해석이 중점적으로 이루어졌다. 하지만, 전기화학적 탈황에 미치는 슬래그의 전기적 특성과 실질 전기에너지에 대한 정성/정량적 논의가 부족하였다. 본 연구에서는 일련의 전기화학적 탈황 실험과 열역학적 계산을 통해 CaO-Al2O3-MgO 슬래그에 의한 용철의 탈황능과 인가 전기에너지와의 상관 관계를 확인하고자 하였다. 용강 내 S가 슬래그로 평형 분배될 시점에 특정 전류를 일정하게 인가하여 일반 탈황 대비 전기 탈황 후의 평형 S 분배비 변화를 확인하였고, Nernst 식을 통해 전기화학적 전위차(ΔΦ_S) 추정하였다. 실험 결과, 같은 전류 인가 조건임에도 슬래그 조성에 따라 ΔΦ_S가 변화함을 확인하였다. 다시 말해, 슬래그 내 CaO 함량이 많을수록 전기 탈황 효과가 떨어짐을 의미한다. 이는 슬래그 내 CaO 함량이 많을수록 일반 평형에 유리하다는 정설과는 반대되는 경향으로, 전기화학적 탈황은 슬래그의 전기적 특성에 영향을 받는 것으로 추정된다. 또한, 본 연구에서는 전기화학적 탈황에 쓰이는 저항을 정의하고, 해당 슬래그의 비저항 문헌 보고값과 저항 측정값을 통해, R_DeS (전기화학적 특성)와 R_slag (물리화학적 특성)의 선형 관계를 확인하였다. 전기화학적 탈황 정도를 나타내는 ΔΦ_S는 용융 산화물 내 황화물 용해를 위해 개발된 열역학 모델을 통해 예측할 수 있었다. 전기화학적 탈황은 용철 내 S의 화학적 포텐셜이 감소하는 반면, 슬래그 내 CaS의 화학적 포텐셜은 증가하게 된다. 이를 이용한 열역학 계산 결과, ΔΦ_S는 동일한 전류 인가 조건에서 CaO/Al2O3 함량비가 증가함에 따라 감소하는 것으로 나타났다. ΔΦ_S 예측 모델은 실험 데이터와도 잘 일치하였고, 전기화학적 탈황에 의해 용철의 목표 S 함량 도달에 필요한 전류(또는 전류 밀도)를 추정하는 데 활용할 수 있었다. 본 연구는 또한 전기화학적 탈황의 실조업 적용을 위한 전극재료 및 적용 공정을 제안하고, 예상되는 전기에너지 비용 계산을 시도하였다. 는 용융 산화물 내 황화물 용해를 위해 개발된 열역학 모델을 통해 예측할 수 있었다. 전기화학적 탈황은 용철 내 S의 화학적 포텐셜이 감소하는 반면, 슬래그 내 CaS의 화학적 포텐셜은 증가하게 된다. 이를 이용한 열역학 계산 결과, ΔΦ_S는 동일한 전류 인가 조건에서 CaO/Al2O3 함량비가 증가함에 따라 감소하는 것으로 나타났다. ΔΦ_S 예측 모델은 실험 데이터와도 잘 일치하였고, 전기화학적 탈황에 의해 용철의 목표 S 함량 도달에 필요한 전류(또는 전류 밀도)를 추정하는 데 활용할 수 있었다. 본 연구는 또한 전기화학적 탈황의 실조업 적용을 위한 전극재료 및 적용 공정을 제안하고, 예상되는 전기에너지 비용 계산을 시도하였다. Sulfur (S) should be removed from molten iron during the steelmaking process since the demand for high-quality steel products is growing. Desulfurization of molten iron by molten slag is an electrochemical reaction where S in molten iron dissolves into molten slag as sulfides by gaining electrons. Several previous studies reported the feasibility of electrochemical desulfurization based on its electrochemical character. These investigations confirmed that the application of electricity could decrease the final S content in molten iron below that obtained by normal desulfurization. They primarily focused on the kinetic analysis, reaction mechanism, and electricity efficiency of electrochemical desulfurization. However, the previous studies are lacking regarding the effect of the electric potential on the electrochemical equilibrium and the properties of slag when used for desulfurization. In the present study, the electrochemical desulfurization of molten iron by molten slag was correlated to the electricity applied to the system by conducting a series of electrochemical desulfurization experiments and thermodynamic analyses. A molten iron containing C and S was allowed to react with CaO-Al2O3-MgO_sat. slag at 1400 ºC. It was subsequently desulfurized further by applying electricity of constant current. S distribution coefficients (L_S = (pct S)/[pct S]) were obtained after the normal and the electrochemical equilibria, respectively. The obtained results were interpreted by employing the Nernst equation to extract the electrochemical potential difference (ΔΦ_S) for the electrochemical desulfurization. It was confirmed that applying electric current (I) increased the L_S after the electrochemical desulfurization, which resulted in the increase of ΔΦ_S. A resistance, R_DeS = ΔΦ_S/I, or a specific resistivity, ρ_DeS = ΔΦ_S/(I/A), for the electrochemical desulfurization were defined, which can be used to characterize the susceptibility to the electrochemical desulfurization. The R_DeS was found to be independent of the I within the range of this investigation, to decrease as (pct CaO)/(pct Al2O3) (C/A) increased, and to be proportional to the resistance of the slag (R_slag). On the other hand, the effect of the C content on the ΔΦ_S was negligible when the same slag was used. Favorable conditions for electrochemical desulfurization were not the same as those for normal desulfurization. The ΔΦ_S, which is an indicator of the extent of electrochemical desulfurization, was independently predicted by employing a thermodynamic model for the oxysulfide slag. The enhanced desulfurization by the electricity was interpreted as increasing the chemical potential of S in molten slag (in the form of CaS) relative to that of S in molten iron. It was found that the ΔΦ_S predicted by the thermodynamic calculation decreased with increasing C/A under the same applied current. Moreover, it increased with increasing current at the same C/A ratio. The model prediction was in good agreement with the experimental data. The thermodynamic prediction of ΔΦ_S and the experimental data can be practically used to estimate the electric current or current density (for a set cathodic area) for a desired electrochemical desulfurization. The present study also presented graphite and cermet electrode materials and discussed steelmaking processes for the practical application of electrochemical desulfurization.

합금원소가 철강 영률에 미치는 영향

황지현 포항공과대학교 철강대학원 2020 국내박사

The demand to increase the efficiency and performance of structural members motivates the development of innovative structural materials for various applications. An increase in elastic modulus can also decrease the thickness of material and thus provides significant weight saving in various application along with a prevention of buckling in structural components. Therefore, understanding the microstructural and physical effects on elastic modulus is the most important to develop high modulus materials. This paper focused on microstructural influences on elastic modulus of steels. Preliminary research was performed to determine the effect of Al2O3 particles on elastic modulus of aluminum metal matrix composite (Al MMC). For main subject, carbon, which is the most important alloying element in steels, is selected to investigate its effect on elastic modulus of steels after various thermal mechanical treatment. Other alloying elements (W, Cr, Ti, and Nb) that form the particles having a higher elastic modulus than steels are selected as well to investigate their effects on elastic modulus of steels after solid solution and aging treatment. The demand to increase the efficiency and performance of structural members motivates the development of innovative structural materials for various applications. An increase in elastic modulus can also decrease the thickness of material and thus provides significant weight saving in various application along with a prevention of buckling in structural components. Therefore, understanding the microstructural and physical effects on elastic modulus is the most important to develop high modulus materials. This paper focused on microstructural influences on elastic modulus of steels. Preliminary research was performed to determine the effect of Al2O3 particles on elastic modulus of aluminum metal matrix composite (Al MMC). For main subject, carbon, which is the most important alloying element in steels, is selected to investigate its effect on elastic modulus of steels after various thermal mechanical treatment. Other alloying elements (W, Cr, Ti, and Nb) that form the particles having a higher elastic modulus than steels are selected as well to investigate their effects on elastic modulus of steels after solid solution and aging treatment.

Effect of HER property in bainite-martensite complex steels according to Si and Al compositions with isothermal heat treatment temperatures

신석모 포항공과대학교 철강·에너지소재대학원 2022 국내석사

Various elements are used to make high strength and high elongation alloy steels. Among the lots of alloy elements, Si is the representative element to improve strength of steels because it has solid solution strengthening effects to the ferrite matrix. On the other hand, Si makes wettability problems in the hot-dip galvanizing process, so Al is being considered one of the elements discussed to replace it. To make bainite and martensite mixed complex steels by performing isothermal heat treatments before/after Ms temperature through Q&P process, two different types of chemical composition alloy steels are used. One is high Si contents alloy steels (Si7Cr2 alloys), and the other one is Al contents steels (Al2Cr2 alloys) in which Si contents are replaced partially by Al. As a result of confirming the stretch-flangeability through the HER (Hole Expansion Ratio) test, there is no significant differences in HER because chemical compositions between two alloy steels. But, in common, they show that the HER is decreased as the isothermal heat treatment temperature is increased. By EBSD (Electron BackScatter Diffraction) analysis of HER specimens, it is found that the smaller EGs (Effective Grain size), the higher HER. And the larger aspect ratio (the microstructures are longer and lath shape), the higher HER. Through this, it is found that the chemical compositions of alloy elements by replacing parts of the Si contents with Al contents have a little effect on the HER characteristics. And also, it is confirmed that the most important factors affecting to the HER characteristics are the size and shape of the microstructures according to the isothermal heat treatment temperatures. 고강도/고연신 합금강을 만들기 위하여 다양한 합금원소가 사용되고 있다. 그중에서 Si은 대표적인 합금원소로 ferrite matrix에 고용강화 효과를 부여하여 강도 향상에 기여하지만, 용융도금공정에서 낮은 젖음성 문제를 일으켜 이를 대체하기 위하여 Al이 거론되고 있다. 높은 함량의 Si를 가진 합금강(Si7Cr2 alloys)과 Si의 일부를 소량 Al로 대체한 합금강(Al2Cr2 alloys)을 Q&P 프로세스를 통해 Ms 온도 전/후로 등온 열처리를 진행하여 bainite와 martensite가 혼합된 복합조직강을 만들었다. HER(Hole Expansion Ratio) 시험을 통해 Stretch-flangeability 특성을 확인한 결과, 두 강종 간에 합금원소 함량 차이에 의한 HER 차이는 크게 없었으며, 공통적으로 등온 열처리 온도가 증가함에 따라서 HER이 감소하는 경향을 보였다. HER 시편에 대하여 EBSD(Electron BackScatter Diffraction) 분석 결과, EGs(Effective Grain size)가 작을수록 HER이 증가하고, Aspect Ratio가 클수록(microstructure가 길쭉한 lath 형태일수록) HER이 증가한다는 것을 알 수 있었다. 이를 통해 HER 특성에 대하여 Si 함량 일부를 Al로 대체함에 따른 합금원소 영향은 거의 없으며, 등온 열처리 조건에 따른 microstructure의 크기와 형상이 HER에 영향을 미치는 주요 인자라는 것을 알 수 있었다.

Effect of FeO/MgO-containing materials on melting behavior of sinter mix containing high Al2O3 ore

김성진 포항공과대학교 철강·에너지소재대학원 2022 국내석사

현재 철강 산업에서는 고품위 철광석의 단가가 상승함에 따라 많은 맥석 성분을 함유한 저품위 철광석의 배합의 필요성이 대두되고 있다. 특히 호주, 인도 등에서 수입한 철광석은 다량의 Al2O3을 함유하고 있어 조업 시 소결광의 융화 거동과 소결강도 전반적으로 악영향을 미친다. 이에 본 연구는 고알루미나 철광석을 배합한 소결광을 제조하여 Al2O3의 악효과를 상쇄하는 방법으로 FeO 및 MgO원을 첨가하여 융화거동을 실험 및 비교하였다. 시약급 FeO를 소결광에 첨가 시 DSC 열분석에 의한 소결광의 융액생성온도는 감소하였고 Fluidity test를 통한 IFL (Index of fluid of liquid)는 증가하였다. 반대로 시약급 MgO를 고정된 FeO 첨가량 내에서 첨가 시 융액생성온도는 증가하였고 IFL은 감소하였다. 이는 FactSage 소프트웨어로 계산된 각 시약급 FeO/MgO 첨가량과 융점 및 점도의 상관성과 일치하였다. 이후에 FeO원으로 Mill scale을 첨가시 소결광의 융액생성온도와 IFL에서 향상된 결과를 보였으며 MgO원으로 Dolomite를 첨가했을 때는 시약급 MgO와 달리 융액생성온도, IFL, 그리고 소결강도를 나타내는 압축강도와 RDI 모두 개선되는 결과를 보였다. 이를 통해 철강 공정 중 부산물인 Mill scale을 소결 공정에 활용할 수 있는 가능성을 보았으며 MgO원으로서의 Dolomite의 염기도 증가와 조밀한 기공의 미세구조에 따라 융화거동이 추가적으로 개선될 수 있을 것이다. The current study was focused on the utilization of low-grade ores containing high Al2O3 in the iron ore sintering process because Al2O3 has been reported to be a harmful gangue component for the melting behavior of the sinter mix. The formation temperature of melts and IFL (Index of the fluidity of liquid) were improved when smaller than 40% of high Al2O3 ore was blended to the standard sinter mix. To cope with the adverse effect of Al2O3 on the melting behavior of the sinter mix, FeO and MgO were mixed together with a sinter mix containing high Al2O3 ore. Adding FeO to the sinter mix was effective for lowering the melting point and viscosity of the sinter mix according to the FactSage calculation, while the addition of MgO has an optimal ratio for minimizing them when it was mixed with FeO. The thermal analyses (DSC) and fluidity test showed similar results to the calculation when FeO and MgO reagent were used where the positive effect was maximized at 5 wt% of FeO addition. Effect of MgO reagent on melt formation temperature and IFL was negative in the entire range of MgO addition. This experimental procedure was repeated using mill scale and calcined dolomite instead of FeO and MgO reagents. Adding mill scale to sinter mix was effective for decreasing melt formation temperature especially at 7.5 wt%, while it improved IFL, compressive load and RDI. Adding calcined dolomite together with mill scale also decreased both melt formation temperature and IFL, which was different from the case of reagent MgO. It would be estimated that the development of Magenesio-spinel [(Fe,Mg)O∙Fe2O3] and mineralogical properties of mill scale and dolomite might strongly be connected to the melting behavior of sinter mix. The current research would provide the fundamental understanding for the utilization of FeO- and MgO-containing materials together to maintain sinter quality in case high Al2O3 ores are blended.

Modeling of Predictive Maintenance Systems for Laser-Welders in Continuous Galvanizing Lines (CGL) Based on Machine-Learning with Welder’s Control Data : 데이터 머신러닝 기반의 철강 아연도금라인용 레이저 용접기 고장예지 및 방지 시스템 모델

최진성 포항공과대학교 철강·에너지소재대학원 2023 국내석사

A Knowledge Retrieval Model Based on a Graph Database for Semantic Search in Equipment Purchase Order’s Specifications for Steel Plants

차호진 포항공과대학교 철강에너지소재대학원 2023 국내석사

Due to the complexity and aging history of industrial plant engineering, equipment maintenance and replacement activities in recent years have been rapidly increasing. Consequently, the plant owners are challenged to reduce the process and review time of equipment Purchase Order (PO) documents. Currently, the traditional keyword-based document search technology might generate unintentional errors and omissions, resulting in inaccurate search results when processing PO documents with the equipment suppliers. Through this research, a document knowledge retrieval model that applies knowledge graph (KG) technology to PO documents for steel plant equipment was designed and developed, so-called purchase order knowledge retrieval model (POKREM). For the development of POKREM, four domain data were defined: (1) factory hierarchy, (2) document hierarchy, (3) equipment classification hierarchy, and (4) PO data. Information for each domain was created in a graph database through the three sub-processes of being (a) defined in a hierarchical structure, (b) classified into nodes and relationships, and (c) written in triples. Ten CSV (comma-separated values) files were created and imported into the graph database as data pre-processing for creating multiple nodes. Finally, rule-based reasoning technology was applied to enhance the model’s contextual search performance. POKREM was developed using Neo4j as an open-source graph DB, and implemented by converting it into a cloud platform on the web for better system and user and interface. As a result of POKREM's performance evaluation, accuracy was 99.7%, precision 91.7%, recall 100%, and F1 score 95.7%. A validation study showed that POKREM was able to search for accurate answers to fact-related queries in most cases, although , it retrieved some incorrect answers in a few cases to reasoning-related queries. An expert survey of PO practitioners indicated that the PO document review time with POKREM was reduced by about 40%, compared to the previous manual process. The contribution of this study might upgrade engineers’ work efficiency by improving document search time and accuracy and could be expandable to other plant engineering documents such as contracts and drawings.

저탄소 가공용 BP재의 시효에 미치는 Mn량 및 석출물의 영향

황선택 포항공과대학교 철강대학원 1997 국내석사

노이즈가 심한 철강 제품에 대한 표면결함 검출 방법에 관한 연구 : A Study on Vision-based Surface Inspection Method for Highly Noised Steel Production

박창현 포항공과대학교 철강대학원 2010 국내박사

An automated vision-based surface inspection for highly noised steel production is proposed in this thesis. The ``highly noised'' means the surface of the steel product is covered with a lot of scales, which make the inspection difficult task. This thesis focuses the inspection method for steel wire rod. The optimal illumination structure is proposed to detect defects with depth by using double dark fields. The inspection methods for the most critical defects, roller-mark (periodic defect) and scratch (non-periodic defect) of wire rod are presented. The two kinds of defects are caused by defective rollers and occurs in a large quantity. It is essential for operators to receive early warnings of such defects and take appropriate action to avoid a large quantity of defective products. This thesis presents an efficient and real-time method for detecting defects in continuous steel wire rods. The detection algorithm is based on the undecimated discrete wavelet transform (UDWT). To overcome the noisy background of surface image, the algorithm exploits the translation invariant property of the UDWT, which improve the signal-to-noise ratio (SNR) of the image. The horizontal and vertical detail coefficients are binarized by double thresholding technique followed by the appropriate size filtering. The final defect candidates are obtained by morphological reconstructing operation. To identify periodic defects, the frequency spectrum of defect positions is analyzed. The spatial periodicity of detected defects is estimated from the inversely related periodic information of frequency spectrum. To get 1D frequency spectrum of defects, the algorithm converts 2D positional information of defects into 1D defect histogram. A robust algorithm to detect repeating impulse-like points in the spectrum uses a median and local maximum filters. The algorithm was proved to operate even when the false positive defects are over 15 times more than true positive defects, or 70 % of true positive defects are missed. To classify non-periodic defects, the classification using support vector machine (SVM) is proposed. The 14 features are extracted from binary and gray information of detected objects. The best model SVM classifier with radial basis function (RBF) kernel is obtained by 5-fold cross validation. The grid search for finding the two parameters for RBF kernel is performed. The support vectors are generated from train data and the final confusion matrix is obtained from test data, which show 85 % and 94 % of true positive and negative rate, respectively. The inspection system using the proposed method was applied to a real production line and showed the effectiveness of the method.