축소시스템 기반 비행체 날개 최적화 연구

김현기(Kim, Hyun-Gi),최인호(Choi, In-Ho) 한국산학기술학회 2012 한국산학기술학회논문지 Vol.13 No.10

본 연구에서는 축소모델을 기반으로 비행체 날개를 최적화하는 기법을 제안한다. 잘 구축된 축소모델은 고 유치 문제나 동적 해석 시 정확한 해석결과를 제공하며, 최적화 과정에서 필요한 민감도 계산에서도 정확한 결과를 제공할 수 있다. 이러한 축소모델은 모드기반으로 구축되는 축소차수모델(Reduce Order Model)과 자유도기반으로 구 축되는 축소시스템(Reduced System)으로 구분되는데, 본 연구에서 사용하는 자유도 기반 축소시스템은 구조물의 거동 에 지배적인 자유도를 적절히 선정하는 것이 중요하므로, 이를 위하여 기존 연구에서 신뢰성이 검증된 2단계 축소방 법을 사용하였고, IRS(Improved Reduced System)에 의해 최종시스템을 구축하였다. 수치예제에서 최적화 과정에서 계 산되는 등가응력, 고유치 및 설계민감도는 모두 축소시스템 기반으로 구해지며, 축소시스템을 통해 구속조건을 잘 만 족하면서 목적함수에 대한 최적 결과를 얻을 수 있음을 보인다. The present study proposes the optimization of wing structure base on reduced model which assures the solution accuracy and computational efficiency. Well-constructed reduced model assures the accurate result in the eigenvalue problem, dynamic analysis or sensitivity of design optimization. Reduced system is classified into the reduce-order model based on structural modes and the reduced system based on degrees of freedom. Because this study uses the reduced system based on degrees of freedom, it is important to select the dominant degrees of freedom properly. For this work, robust selection method, two-level selection scheme, is employed and IRS(Improved Reduced System) is applied to construct the final reduced system. In the optimization process based on the reduced system, all of the equivalent stress, eigenvalue and design sensitivities are calculated from the reduced system. Through a numerical example, it is shown that the present optimization methodology based on the reduction method can provide an optimal results for objective function satisfying constraint condition.

Proper Orthogonal Decomposition을 이용한 천음속 날개/동체 모델의 최적 설계

박경현(Kyung-Hyun Park),전상욱(Sang-Ook Jun),조맹효(Maeng-Hyo Cho),이동호(Dong-Ho Lee) 한국항공우주학회 2010 韓國航空宇宙學會誌 Vol.38 No.5

본 연구에서는 천음속 날개/동체 모델에 대한 축소모델(Reduced Order Model; ROM)의 정확성을 검증하고, Proper Orthogonal Decomposition(POD)을 이용한 최적설계를 통해 그 효율성을 검토하였다. full order 공력해석을 통한 Snapshot을 추출하기 위해 삼차원 오일러 방정식을 이용하였으며, 이들 snapshot들을 통해 날개/동체 모델 주위 유동장의 거동을 모사하는 POD의 기저벡터를 계산 하였다. 이러한 과정을 거쳐 구축된 축소모델은 6개의 Case 들로 검증하였으며, 그 결과 ROM을 이용해 관심영역에 대한 유동장의 예측을 할 수 있다는 것을 확인하였다. 그리고 ROM을 통한 날개/동체 모델의 최적설계를 수행 하였으며, 그 결과는 반응면모델(Response Surface Model; RSM)을 이용한 최적설계 결과와 비교 하였다. 이를 통해 ROM을 바탕으로한 최적설계가RSM을 이용한 것보다 효율적임을 확인하였다. This paper presents a validation of the accuracy of a reduced order model(ROM) and the efficiency of the design optimization using a Proper Orthogonal Decomposition(POD) to transonic wing/fuselage system. Three dimensional Euler equations are solved to extrude snapshot data of the full order aerodynamic analysis, and then a set of POD basis vectors reproducing the behavior of flow around the wing/fuselage system is calculated from these snapshots. In this study, reduced order model constructed through this procedure is applied to several validation cases, and then it is confirmed that the ROM has the capability of the prediction of flow field in the space of interest. Additionally, after the design optimization of the wing/fuselage system with the ROM is performed, results of the ROM are compared with results of the design optimization using response surface model(RSM). From these, it can be confirmed that the design optimization with the ROM is more efficient than RSM.

Constructing a digital twin for estimating the response and load of a piping system subjected to seismic and arbitrary loads

Shinyong Kwag,Dongchang Kim,Gungyu Kim,Seunghyun Eem 국제구조공학회 2023 Smart Structures and Systems, An International Jou Vol.31 No.3

In recent years, technological developments have rapidly increased the number of complex structures and equipment in the industrial. Accordingly, the prognostics and health monitoring (PHM) technology has become significant. The safety assessment of industrial sites requires data obtained by installing a number of sensors in the structure. Therefore, digital twin technology, which forms the core of the Fourth Industrial Revolution, is attracting attention in the safety field. The research on digital twin technology of structures subjected to seismic loads has been conducted recently. Hence, this study proposes a digital twin system that estimates the responses and arbitrary load in real time by utilizing the minimum sensor to a pipe that receives a seismic and arbitrary load. To construct the digital twin system, a finite-element model was created considering the dynamic characteristics of the pipe system, and then updating the finite-element model. In addition, the calculation speed was improved using a finite-element model that applied the reduced-order modeling (ROM) technology to achieve real-time performance. The constructed digital twin system successfully and rapidly estimated the load and the point where the sensor was not attached. The accuracy of the constructed digital twin system was verified by comparing the response of the digital twin model with that derived by using the load estimated from the digital twin model as input in the finite-element model.

전투기 날개/동체모델의 공력-구조 연계해석 축소모델을 위한 POD Basis Weight 예측기법 연구

박경현,최선,김의영,박훈일,조맹효,이동호 한국항공우주학회 2012 한국항공우주학회 학술발표회 논문집 Vol.2012 No.11

본 연구에서는 전투기 날개/동체모델의 공력-구조 연계 축소모델을 위한 POD (Proper Orthogonal Decomposition) basis weighting 기법의 연구를 수행하였다. POD basis weighting 기법으로는 기저벡터의 일차결합, 인공신경망, 크리깅 기법을 이용하였으며, 이들 기법의 효과적인 비교를 위하여 천음속 영역에서 날개/동체모델의 후퇴각, 스팬길이, 스파 및 리브의 두께에 대하여 총 10개의 Snapshot을 추출하였다. Snapshot은 삼차원 오일러 방정식과 Aminpour의 4절점 쉘요소를 이용하여 추출하였으며, 이들 Snapshot들을 통해 날개/동체 모델 주위 유동장을 대표하는 POD 기저벡터를 계산하였다. 각각의 POD basis weighting 기법을 통하여 계산된 POD basis는 full order mapping을 통하여 날개/동체 주변 유동장 및 구조변형을 예측하였으며, 각 기법을 통하여 예측된 결과를 비교분석 하였다. 이를 통해 크리깅 기법을 이용한 POD 해석기법이 가장 높은 신뢰도를 지닌 것을 확인할 수 있었다. In this paper, the POD basis weight estimation method for the reduced order model of aero-structural analysis of a fighter wing/fuselage system is studied. The linear combination, the artificial neural network, and the Kriging method are adopted to predict POD basis weight coefficient. To effectively compare each method, snapshots are selected from the combinations of the upper and lower bounds of each variable (sweep angle, span length, and spar&rib thickness of wing/fuselage system) in the transonic regime. Three dimensional Euler equations are solved to extrude snapshot data of the full order aerodynamic analysis, and then a set of POD basis vectors reproducing the behavior of flow around the wing/fuselage system is calculated from these snapshots through the each POD basis weight estimation method. Then, the comparison of each POD basis weighting method is performed. From these, it is confirmed that the Kriging method based POD analysis give highest robust prediction result of the three.

보일러 연소특성확인을 위한 시험연소로에서의 동적 차수감소모델에 대한 연구

남윤민(Yunmin Nam),이창민(Changmin Lee),한가람(Karam Han),김범신(Bumshin Kim) 한국연소학회 2021 KOSCOSYMPOSIUM논문집 Vol.2021 No.5

Unlike the static reduced order model based on the correlation of in/output variables at a specific point in time, the dynamic reduced order model is a model that considers changes in the correlation of in/output data over time. This is proper to use in particle combustion reactions or chemical reactions. In this study, by using a pilot scale coal combustion furnace that is easy to control in/output variables, A dynamic reduced order model for boiler combustion model was developed.

Proper Orthogonal Decomposition을 이용한 고고도 장기체공 항공기 날개의 정적 공탄성 해석에 관한 연구

이상아,박경현,김정화,전상욱,이동호 한국항공우주학회 2011 한국항공우주학회 학술발표회 논문집 Vol.2011 No.4

본 연구에서는 고고도 장기체공 항공기 날개 모델에 대해 공력 축소모델(Reduced Order Model;ROM)의 정확성을 검증하고 Proper Orthogonal Decomposition(POD)를 이용한 정적공탄성 해석을 수행하였다. Full order 공력해석을 통한 Snapshot을 추출하기 위해 3D Euler Solver를 이용하였으며, 이들 Snapshot들을 통해 날개 주위 유동장의 거동을 모사하는 POD의 기저벡터를 계산 하였다. 계산된 기저벡터는 POD의 공력해석 검증과 정적 공탄성 해석을 위한 공력-구조 연계해석에 적용 하였다. 해석대상이 된 날개는 구조적 대변형이 발생하는 유연날개 구조를 가지므로 비선형성이 고려된 정적 공탄성 해석을 수행하였으며 해석 결과는 full order 3D Euler Solver와 유한요소법을 이용한 결과와 비교하였다. 이를 통해 POD를 이용한 정적 공탄성 해석의 정확성 및 효율성을 확인하였다. In this Study, static aeroelasticity analysis using Proper orthogonal decomposition(POD) is performed for HALE aircraft wing. Three dimensional Euler equation is solved to exclude snapshot data of the full order aerodynamic analysis, and then a set of POD basis vectors reproducing the flow field around the wing is calculated from these snapshots. The POD basis is applied to validation of POD and fluid-structure interaction analysis for static aeroelasticity. Because the wing analysed in this study has characteristic of flexibility of large deflection, the study has performed with nonlinear Finite Element Method(FEM) analysis. Results are compared with full order analysis and accuracy and efficiency of POD are confirmed.

Study on the system reduction under the condition of dynamic load

Hyun-gi Kim,조맹효 대한기계학회 2013 JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY Vol.27 No.1

This study presents the methodology of the reduced system for dynamic analysis considering the selection criterion of the primary degrees of freedom based on the relation between natural frequency and external loading frequency. A well-constructed reduced system can provide the accurate representation of the dynamic behavior of a structure under arbitrary dynamic loads. If dynamic response based on a reduced system can be calculated accurately, the proposed method can save the computing time remarkably in problems which require repeated calculation such as optimization procedure or time integration. In addition, the proposed method shows the reliable dynamic response in multi-domain structure. Numerical examples demonstrate the reliability of the dynamic analysis obtained from a reduced system and responses of the reduced system are compared with those of a global system.

BFS 유동장의 축약 모델링 오차 분석

이진익(J.I.  Lee),이은석(E.S. Lee),이광섭(K.S. Lee),전병을(B.E. Jun) 한국전산유체공학회 2011 한국전산유체공학회 학술대회논문집 Vol.2011 No.11

Error analysis in the reduction of BFS(Backward Facing Phase) flow field is dealt with in this paper. The POD(Proper Orthogonal Decomposition) technique is adopted in reduced order modeling of flow field from a bunch of flow snapshots. It is well-known for capturing the global flow states rather than local optimum ones. In order for the modeling of the flow field on BFS, the open loop forcing data set are used in the modeling study. The reduced global states from all test snapshots are compared with the local optimal states from individual one for each test case. By analyzing not only the reconstruction errors but also the POD modes, we analyze the reduced order modeling of flow field in detail.

포인트 클라우드 기반의 격자 비의존 합성곱 신경망을 이용한 다양한 형상의 CFD 유동장 예측

황원태(Wontae Hwang),최성임(Seongim Choi) 한국항공우주학회 2023 한국항공우주학회 학술발표회 논문집 Vol.2023 No.11

차원축소모델을 이용한 지면 소음 평가를 위한 소음 반구 예측 연구

권순명(Soon Moung Kwon),강유업(Yu Eop Kang),홍윤표(Yoon Pyo Hong),이관중(Kwan Jung Yee) 한국항공우주학회 2023 한국항공우주학회 학술발표회 논문집 Vol.2023 No.4