Reduced order dynamic aeroelastic model development and integration with nonlinear simulation

Winther, B. A.; Goggin, Patrick; Dykman, John R.

doi:10.2514/6.1998-1897

Cited by 14 publications

(7 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although the 9th-order model is a simplication of the actual non-linear aeroelastic model used to generate the data, such simple linear models based on experimental data can accurately predict the main resonance modes and damping of the structure in the presense of small perturbations. 22 The measured step response (simulated with the F/A-18AAW aeroelastic model) and simulated step response (with the 9th order closed-loop model) show excellent agreement in Figure 3(a). The estimation of the poorly-damped resonance modes is also confirmed by the Bode plot of the 9th order modelR i (q) depicted in Figure 3(b).…”

Section: Application To F/a-18aaw Model Aeroelastic Modelmentioning

confidence: 68%

Step Based Experiment Design and System Identification for AeroElastic Dynamic Modeling

Callafon

Miller

Zeng

et al. 2009

AIAA Atmospheric Flight Mechanics Conference

View full text Add to dashboard Cite

This paper gives an overview of the main ideas behind a new estimation technique that is able to deliver a consistent estimate of (poorly-damped) structural resonance modes induced by aeroservoelastic interaction without user intervention. Instead of using broadband sinusoidal or sine-sweep excitation signals, the new estimation method proposed in this paper uses simple step-based excitation signals that can be applied to any of the control surfaces to formulate a model of the aerodynamically-induced structural vibration modes. Simple step-based excitation signals address the issue of broadband excitation while allowing a reasonable (short-time) excitation signal on the control surfaces. The proposed estimation method has been tailored to find consistent model estimates on the basis of step-based input signals. The estimation method uses a modified version of the well-known realization algorithm that is extended to arbitrary input signals to formulate a discrete-time model directly on the basis of step-based experiments. Since the numerical implementation only requires a singular value decomposition and a standard least-squares estimation, robust numerical algorithms can be put in place to formulate a dynamical model with little or no user intervention. The procedure is illustrated on time-domain data obtained from a high fidelity model of an F/A-18 to find a low-order model that captures the structural parameters, such as damping and location of the first main resonance modes.

show abstract

Section: Application To F/a-18aaw Model Aeroelastic Modelmentioning

confidence: 68%

Step Based Experiment Design and System Identification for AeroElastic Dynamic Modeling

Callafon

Miller

Zeng

et al. 2009

AIAA Atmospheric Flight Mechanics Conference

View full text Add to dashboard Cite

show abstract

“…As the wing penetrates the plane, the scale of the ow increases, requiring a lift force equal to the downward velocity multiplied by the local increase of the additional apparent mass. 1 At any cross ow plane, the apparent mass of the uid is given by db.x/ dx D U 2 ®½2¼ x tan . "/ 2…”

Section: Methodsmentioning

confidence: 99%

Analytical Representation of Aerodynamic Forces in Laplace Domain

Winther

2003

Journal of Aircraft

View full text Add to dashboard Cite

395values reach very low-pressure values with peak locations moving downstream. High-Lift Con gurationFor the high-lift con guration the aps were considered to be de ected 23 deg downward (landingmode).Solutionswere obtained at ® D 1, 4, and 7.5 deg on a mesh having a total of 1,467,309 cells and 242,984 points, with 72,615 points on the aircraft surface. The surface grid is shown in Fig. 4. Figure 5 shows the pressure contours on the aircraft at ® D 7:5 deg. C p distributions at various angles of attack are given at the 32% spanwise location in Fig. 6. It is clear that the freestream angle of attack has no signi cant effect on the pressure distribution on the surface of the aps. Because at this relatively high ap de ection angle, ow over the aps is insensitive to any change in the angle of attack.Becauseno experimentaldata on the aircraftcon gurationstudied were availablein the open literature,the CFD results were compared with the results of Ref. 6 using the Advanced Aircraft Analysis Software. This software is based on the lifting-line theory and semiempirical formulations, which are commonly used in preliminary aircraft design. 7 As shown in Fig. 7, both of the methods show linear variations for all aerodynamiccoef cients with angle of attack, having nearlythe same slopesfor C M but slightlydifferentslopesfor C L . Computational DetailsThe aircraft solutions were obtained with approximately 1000 iterations for a residual reduction of about 1.5 orders of magnitude. Cruise con guration solutions of the aircraft took about one week for each angle of attack and required 1103 MB of memory on an HP-UX workstation. For the high-lift con guration solutions took about two weeks for each angle of attack and required 1275 MB of memory. It was observed that the memory requirement of the code was directly proportionalto the number of cells in the computational domain.In this work the Courant-Friedrichs-Lewy (CFL) number was varied from 1 to a nal value of 100. However, when convergence problems were encounteredthe upper limit for the CFL number was reduced to 50 or 25. The explicit Runge-Kutta and point implicit schemes had convergence problems in all of the cases. For this reason a fully implicit scheme was used in all of the computations, although it required more computer memory. ConclusionsInviscid, subsonic ow solutions for a medium-range cargo aircraft were obtained on unstructuredgrids at cruise and high-lift congurations.From geometry modeling to the ow solution,this study was performed using the CFD-GEOM-V5 grid-generationcode and the CFD-FASTRAN-V2.2 ow solver code.The results showed that the lift coef cients varied linearly for the cruiseand landingcon gurations.The pitching-momentcoef cients showed the characteristics of a stable aircraft. Comparison of the CFD and the empirical solutions indicated a reasonable agreement for the lift and moment coef cients. NomenclatureA k = system matrix for model k C = general matrix M or P D = determinant of dynamic system E i = aerodynamic coef cients due to pitch h = vert...

show abstract

“…where n is the total number of KN steps taken using the improved method. Equation 28 indicates that only 1 augmented FEM system needs to be solved for each KN step. This feature demonstrates that the proposed method can be seen as a particular modified Newton method with a constant iteration matrix over the single KN step.…”

Section: Figurementioning

confidence: 99%

“…The idea of reduced-order modeling technique has been studied intensively in the past across various disciplines following different approaches, ie, the methods for the structural static analysis, [11][12][13][14][15][16] the structural vibration, [17][18][19] the structural dynamics, [20][21][22][23] the fluid mechanics, [24][25][26][27] the aeroelasticity, 28,29 and the optimization. 30,31 In this paper, our work focuses on the structural static analysis, and 2 families of reduced-order modeling techniques implemented in the FE method context can be recognized in this field.…”

mentioning

confidence: 99%

A modified Newton‐type Koiter‐Newton method for tracing the geometrically nonlinear response of structures

Liang

Abdalla

Qin

2017

Numerical Meth Engineering

View full text Add to dashboard Cite

Summary The Koiter‐Newton (KN) method is a combination of local multimode polynomial approximations inspired by Koiter's initial postbuckling theory and global corrections using the standard Newton method. In the original formulation, the local polynomial approximation, called a reduced‐order model, is used to make significantly more accurate predictions compared to the standard linear prediction used in conjunction with Newton method. The correction to the exact equilibrium path relied exclusively on Newton‐Raphson method using the full model. In this paper, we proposed a modified Newton‐type KN method to trace the geometrically nonlinear response of structures. The developed predictor‐corrector strategy is applied to each predicted solution of the reduced‐order model. The reduced‐order model can be used also in the correction phase, and the exact full nonlinear model is applied only to calculate force residuals. Remainder terms in both the displacement expansion and the reduced‐order model are well considered and constantly updated during correction. The same augmented finite element model system is used for both the construction of the reduced‐order model and the iterations for correction. Hence, the developed method can be seen as a particular modified Newton method with a constant iteration matrix over the single KN step. This significantly reduces the computational cost of the method. As a side product, the method has better error control, leading to more robust step size adaptation strategies. Numerical results demonstrate the effectiveness of the method in treating nonlinear buckling problems.

show abstract

Reduced order dynamic aeroelastic model development and integration with nonlinear simulation

Cited by 14 publications

References 10 publications

Step Based Experiment Design and System Identification for AeroElastic Dynamic Modeling

Step Based Experiment Design and System Identification for AeroElastic Dynamic Modeling

Analytical Representation of Aerodynamic Forces in Laplace Domain

A modified Newton‐type Koiter‐Newton method for tracing the geometrically nonlinear response of structures

Contact Info

Product

Resources

About