• 제목/요약/키워드: DOE(Design Of Experiments)

검색결과 223건 처리시간 0.025초

판재 굽힘 성형에서 실험계획법 및 인공신경망을 이용한 탄성회복 보정 (Springback Compensation of Sheet Metal Bending Process Based on DOE & ANN)

  • 안재홍;고대철;이찬주;김병민
    • 대한기계학회논문집A
    • /
    • 제32권11호
    • /
    • pp.990-996
    • /
    • 2008
  • Nowadays, the trend to a lightweight design accelerates the use of advanced high strength steel (AHSS) in automotive industry. Springback phenomena is a hot issue in the sheet metal forming, especially bending process using AHSS. Several analytical methods for that have been proposed in recent years. Each of method has their advantages and disadvantages. There are only a few optimal solutions which can minimize the two objectives simultaneously. In this study, an effective method optimized the multi objective value. The method by the design of experiments(DOE) and artificial neural network(ANN) was presented to compensate springback of bending parts. This method was applied to L and V bending process. The effective method could be optimized to multiple object. It was confirmed that the proposed method was more efficient than traditional manual FEA procedure and the trial and error approach for springback compensation.

정현적으로 착자된 영구자석을 갖는 마그네틱 위치센서 설계 (Design of Rotary Magnetic Position Sensor with Sinusoidally Magnetized Permanent Magnet)

  • 정승호;류세현;권병일
    • 전기학회논문지
    • /
    • 제56권3호
    • /
    • pp.506-513
    • /
    • 2007
  • This paper proposes a rotary magnetic position sensor which has a sinusoidally magnetized permanent magnet with a small number of poles. To make the sinusoidal magnetic flux density distribution from the permanent magnet, a magnetizing future is optimized by the DOE(Design of Experiments) method. The magnetization process is analyzed using the Preisach model and 2 dimensional finite element method. The magnetic flux density distribution from the magnetized permanent magnet is very similar to ideal sine wave. The simulation result of the magnetic flux density distribution is compared with the experimental one. Also the availability of the proposed rotary type magnetic position sensor is confirmed by position calculation technique.

3 축 나노 스테이지 동특성 해석 및 개선 (Analysis and Improvement of Dynamics Characteristic of 3-axis Nano Stage)

  • 김충;이강녕;이동주;이문구;최형길;이석원;박노철;박영필
    • 한국소음진동공학회:학술대회논문집
    • /
    • 한국소음진동공학회 2005년도 춘계학술대회논문집
    • /
    • pp.555-558
    • /
    • 2005
  • The precision positioning system requires robust design to obtain enough bandwidth. Therefore, The sub-resonance occurred by the disaccord of force center and mass center should be oppressed. And it is necessary to move the flexible mode to a higher frequency. In this paper, the 3-axis nano stage was proposed and dynamic characteristics was improved by design of experiments (DOE).

  • PDF

실험 계획법을 이용한 점착방지막용 플라즈마 증착 공정변수의 최적화 연구 (Optimizing the Plasma Deposition Process Parameters of Antistiction Layers Using a DOE (Design of Experiment))

  • 차남구;박창화;조민수;박진구;정준호;이응숙
    • 한국재료학회지
    • /
    • 제15권11호
    • /
    • pp.705-710
    • /
    • 2005
  • NIL (nanoimprint lithography) technique has demonstrated a high potential for wafer size definition of nanometer as well as micrometer size patterns. During the replication process by NIL, the stiction between the stamp and the polymer is one of major problems. This stiction problem is moi·e important in small sized patterns. An antistiction layer prevents this stiction ana insures a clean demolding process. In this paper, we were using a TCP (transfer coupled plasma) equipment and $C_4F_8$ as a precursor to make a Teflon-like antistiction layer. This antistiction layer was deposited on a 6 inch silicon wafer to have nanometer scale thicknesses. The thickness of deposited antistiction layer was measured by ellipsometry. To optimize the process factor such as table height (TH), substrate temperature (ST), working pressure (WP) and plasma power (PP), we were using a design of experimental (DOE) method. The table of full factorial arrays was set by the 4 factors and 2 levels. Using this table, experiments were organized to achieve 2 responses such as deposition rate and non-uniformity. It was investigated that the main effects and interaction effects between parameters. Deposition rate was in proportion to table height, working pressure and plasma power. Non-uniformity was in proportion to substrate temperature and working pressure. Using a response optimization, we were able to get the optimized deposition condition at desired deposition rate and an experimental deposition rate showed similar results.

Taguchi 실험 계획법에 의한 CH3SH 반도체 악취 가스 센서의 개발 (Development of a Semiconductor Odor Gas Sensor for the Measurement of CH3SH with Taguchi Experimental Design)

  • 김선태;최일환
    • 한국대기환경학회지
    • /
    • 제20권6호
    • /
    • pp.783-792
    • /
    • 2004
  • In this study, a thick-film semiconductor odor gas sensor for the detection of $CH_3$SH was developed using SnO$_2$ as the main substrate and was investigated in terms of its sensitivity and reaction time. In the process of manufacturing the sensor, Taguchi's design of experiment (DOE) was applied to analyze the effects of a variety of parameters, including the substrate, the additives and the fabrication conditions, systematically and effectively. Eight trials of experiments could be possible using the 27 orthogonal array for the seven factors and two levels of condition, which originally demands 128 trials of experiments without DOE. The additives of Sb$_2$O$_{5}$ and PdCl$_2$ with the H$_2$PtCl$_{6}$ ㆍ6$H_2O$ catalyst were appeared to be important factors to improve the sensitivity, and CuO, TiO$_2$, V$_2$O$_{5}$ and PdO were less important. In addition, TiO$_2$, V$_2$O$_{5}$ and PdO would improve the reaction time of a sensor, and CuO, Sb$_2$O$_{5}$, PdCl$_2$ and H$_2$PtCl$_{6}$ㆍ6$H_2O$ were negligible. Being evaluated simultaneously in terms of both sensitivity and reaction time, the sensor showed the higher performance with the addition of TiO$_2$ and PdO, but the opposite results with the addition of CuO, V$_2$O$_{5}$, Sb$_2$O$_{5}$ and PdCl$_2$. The amount of additives were superior in the case of 1% than 4%. H$_2$PtCl$_{6}$ㆍ6$H_2O$ would play an important role for the increase of sensor performance as a catalyst.nce as a catalyst.

Statistical Qualitative Analysis on Chemical Mechanical Polishing Process and Equipment Characterization

  • Hong, Sang-Jeen;Hwang, Jong-Ha;Seo, Dong-Sun
    • Transactions on Electrical and Electronic Materials
    • /
    • 제12권3호
    • /
    • pp.115-118
    • /
    • 2011
  • Process characterization of the chemical mechanical polishing (CMP) process for undensified phosphosilicate glass (PSG) film is reported using design of experiments (DOE). DOE has been addressed to experimenters to understand the relationship between input variables and responses of interest in a simple and efficient way. It is typically beneficial for determining the adequate size of experiments with multiple process variables and making statistical inferences for the responses of interests. Equipment controllable parameters to operate the machine include the down force (DF) of the wafer carrier, pressure on the backside of the wafer, table and spindle speed (SS), slurry flow rate, and pad condition. None of them is independent; thus, the interaction between parameters also needs to be indicated to improve process characterization in CMP. In this paper, we have selected the five controllable equipment parameters, such as DF, back pressure (BP), table speed (TS), SS, and slurry flow (SF), most process engineers recommend to characterize the CMP process with respect to material removal rate (RR) and film uniformity as a percentage. The polished material is undensified PSG. PSG is widely used for the plananization in multi-layered metal interconnects. We identify the main effect of DF, BP, and TS on both RR and film uniformity, as expected, by the statistical modeling and analysis on the metrology data acquired from a series of $2^{5-1}$ fractional factorial design with two center points. This revealed the film uniformity of the polished PSG film contains two and three-way interactions. Therefore, one can easily infer that the process control based on better understanding of the process is the key to success in semiconductor manufacturing, typically when the wafer size reaches 300 mm and is continuously scheduled to expand up to 450 mm in or little after 2012.

실험계획법과 최적화알고리듬을 이용한 터널지보의 파손하중 예측 (Failure Load Prediction of Tunnel Support using DOE and Optimization Algorithm)

  • 이동우;조석수
    • 한국산학기술학회논문지
    • /
    • 제13권4호
    • /
    • pp.1480-1487
    • /
    • 2012
  • 최근, 국내 탄광의 안전성이 향상되고 있으나 사고는 꾸준히 발생하고 있다. 대부분의 터널 지보는 I빔과 I빔을 연결하는 이음판 부분에서 지압을 견디지 못하고 파손이 발생한다. XX 탄광의 경우, 터널지보의 아치부가 일반적인 굽힘 거동이 아닌, 위쪽 방향으로 굽힘이 발생하고 있다. 이러한 경우는 터널지보에 수직하중 이외에 수평하중이 과대하게 작용하는 경우로 볼 수 있으며 이러한 수평하중은 지하암반의 급격한 변화에 의한 암반의 이완범위의 증가나 지하수의 누출 등으로 인한 수리학적인 요인 등의 복합적인 문제가 작용하여 나타난다. 따라서 본 연구에서는 위쪽 방향으로 굽힘을 일으키는 수평하중의 크기를 추정하기 위하여 실험계획법과 최적화 알고리듬을 적용하여 터널지보의 굽힘거동을 규명하였다.

유한요소법과 반응표면법을 이용한 경량 맨홀 커버 구조 최적화 설계 (Structural Optimization of a Light-weight Manhole Cover Using FEM and Response Surface Method)

  • 이형욱
    • 한국생산제조학회지
    • /
    • 제25권6호
    • /
    • pp.462-470
    • /
    • 2016
  • The locking load of a conventional manhole depends on the weight of its cover. Locking-type manhole structures with a special locking mechanism were recently developed to prevent accidents such as stolen cover, away cover from a frame. The weight of the manhole structure can be reduced under structural safety because the locking force of a locking-type manhole is greater than the weight of the cover. A light-weight manhole cover is developed in this study by using a finite element stress analysis and the design of experiments. Static stress analysis and fracture experiments are also conducted to analyze the states of the initial product. The optimum light-weight manhole cover considering manufacturing molds is developed and tested. Consequently, the weight was found to reduce by 16%. In addition, the fracture load increased by 38%.

철도차량 현수장치의 탈선에 대한 민감도 연구 (The Sensitivity Analysis of Derailment in Suspension Elements of Rail Vehicle)

  • 심태웅;박찬경;김기환
    • 한국철도학회:학술대회논문집
    • /
    • 한국철도학회 1999년도 추계학술대회 논문집
    • /
    • pp.566-573
    • /
    • 1999
  • This paper is the result of sensitivity analysis of derailment with respect to the selected suspension elements for the rail vehicle. Derailment phenominon has been explained by the derailment quotient. Thus, the sensitivity of derailment is suggested by a response surface model(RSM) which is a functional relationship between derailment quotient and characteristics of suspension elements. To summarize generation of RSM, we can introduce the procedure of sensitivity analysis as follows. First, to form a RSM, a experiment is performed by a dynamic analysis code, VAMPIRE according to a kind of the design of experiments(DOE). Second, RSM is constructed to a 1$\^$st/ order polynomial and then main effect fators are screened through the stepwise regression. Finally, we can see the sensitivity level through the RSM which only consists of the main effect factors and is expressed by the liner, interaction and quadratic effect terms.

  • PDF

Experimental and numerical study on energy absorption of lattice-core sandwich beam

  • Taghipoor, Hossein;Noori, Mohammad Damghani
    • Steel and Composite Structures
    • /
    • 제27권2호
    • /
    • pp.135-147
    • /
    • 2018
  • Quasi-static three-point bending tests on sandwich beams with expanded metal sheets as core were conducted. Relationships between the force and displacement at the mid-span of the sandwich beams were obtained from the experiments. Numerical simulations were carried out using ABAQUS/EXPLCIT and the results were thoroughly compared with the experimental results. A parametric analysis was performed using a Box-Behnken design (BBD) for the design of experiments (DOE) techniques and a finite element modeling. Then, the influence of the core layers number, size of the cell and, thickness of the substrates was investigated. The results showed that the increase in the size of the expanded metal cell in a reasonable range was required to improve the performance of the structure under bending collapse. It was found that core layers number and size of the cell was key factors governing the quasi-static response of the sandwich beams with lattice cores.