0
Research Papers: Fundamental Issues and Canonical Flows

Optimization Using Arbitrary Lagrangian–Eulerian Formulation of the Navier–Stokes Equations

[+] Author and Article Information
Eysteinn Helgason

Division of Fluid Dynamics,
Department of Applied Mechanics,
Chalmers University of Technology,
Gothenburg S-412 96, Sweden
e-mail: eysteinn@chalmers.se

Siniša Krajnović

Division of Fluid Dynamics,
Department of Applied Mechanics,
Chalmers University of Technology,
Gothenburg S-412 96, Sweden

1Corresponding author.

Contributed by the Fluids Engineering Division of ASME for publication in the JOURNAL OF FLUIDS ENGINEERING. Manuscript received January 23, 2014; final manuscript received February 2, 2015; published online March 9, 2015. Assoc. Editor: Zhongquan Charlie Zheng.

J. Fluids Eng 137(6), 061202 (Jun 01, 2015) (9 pages) Paper No: FE-14-1039; doi: 10.1115/1.4029724 History: Received January 23, 2014; Revised February 02, 2015; Online March 09, 2015

In this paper, we present a new shape optimization method by using sensitivities obtained from the Arbitrary Lagrangian–Eulerian (ALE) form of the Navier–Stokes equations. In the ALE description, the nodes of the computational domain may be moved with the fluid as in the Lagrangian description, held fixed in space as in the Eulerian description, or moved in some arbitrary way in between. Applying the adjoint method with respect to mesh motion allows the whole sensitivity field for the shape changes to be calculated using only two solver calls, a primal solver call and an adjoint solver call. We show that the sensitivities with respect to the mesh motion can be calculated in a postprocessing step to the primal and adjoint flow simulations. The resulting ALE sensitivities are compared to sensitivities obtained using a finite difference approach. Finally, the sensitivities are coupled to a mesh motion smoothing algorithm, and a duct is optimized with respect to the total pressure drop using the proposed method.

FIGURES IN THIS ARTICLE
<>
Copyright © 2015 by ASME
Your Session has timed out. Please sign back in to continue.

References

McNamara, A., Treuille, A., Popović, Z., and Stam, J., 2004, “Fluid Control Using the Adjoint Method,” ACM Trans. Graphics, 23(3), pp. 449–456. [CrossRef]
Achdou, Y., and Pironneau, O., 2005, Computational Methods for Option Pricing, Vol. 30, Society for Industrial Mathematics, Philadelphia. [CrossRef]
Pironneau, O., 1973, “On Optimum Profiles in Stokes Flow,” J. Fluid Mech., 59(1), pp. 117–128. [CrossRef]
Jameson, A., 1988, “Aerodynamic Design Via Control Theory,” J. Sci. Comput., 3(3), pp. 233–260. [CrossRef]
Jameson, A., Pierce, N., and Martinelli, L., 1998, “Optimum Aerodynamic Design Using the Navier-Stokes Equations,” Theor. Comput. Fluid Dyn., 10(1), pp. 213–237. [CrossRef]
OpenFOAM® The open source CFD toolbox. http://www.openfoam.com
Othmer, C., Villiers, E. D., and Weller, H., 2007, “Implementation of a Continuous Adjoint for Topology Optimization of Ducted Flows,” 18th AIAA Computational Fluid Dynamics Conference, June.
Giles, M., and Pierce, N., 2000, “An Introduction to the Adjoint Approach to Design,” Flow Turbul. Combust., 65(3–4), pp. 393–415. [CrossRef]
Nadarajah, S., and Jameson, A., 2000, “A Comparison of the Continuous and Discrete Adjoint Approach to Automatic Aerodynamic Optimization,” AIAA Paper No. 2000-0667. [CrossRef]
Peter, J., and Dwight, R., 2010, “Numerical Sensitivity Analysis for Aerodynamic Optimization: A Survey of Approaches,” Comput. Fluids, 39(3), pp. 373–391. [CrossRef]
Richter, T., and Wick, T., 2013, “Optimal Control and Parameter Estimation for Stationary Fluid-Structure Interaction Problems,” SIAM J. Sci. Comput., 35(5), pp. B1085–B1104. [CrossRef]
Bazilevs, Y., Hsu, M.-C., and Bement, M., 2013, “Adjoint-Based Control of Fluid-Structure Interaction for Computational Steering Applications,” Procedia Comput. Sci., 18(0), pp. 1989–1998. [CrossRef]
Othmer, C., 2008, “A Continuous Adjoint Formulation for the Computation of Topological and Surface Sensitivities of Ducted Flows,” Int. J. Numer. Methods Fluids, 58(8), pp. 861–877. [CrossRef]
Donea, J., Huerta, A., Ponthot, J.-P., and Rodríguez-Ferran, A., 2004, Arbitrary Lagrangian-Eulerian Methods, John Wiley & Sons, Ltd., Chichester, West Sussex, UK.
Duran, A., 2000, “A Numerical Formulation to Solve the ALE Navier-Stokes Equations Applied to the Withdrawal of Magma Chambers,” Ph.D. thesis, Universitat Politècnica de Catalunya, Departament de Matemàtica Aplicada III, Barcelona, Spain.
Soemarwoto, B., 1997, “The Variational Method for Aerodynamic Optimization Using the Navier-Stokes Equations,” Institute for Computer Applications in Science and Engineering, Report No ICASE-97-71.
Zymaris, A., Papadimitriou, D., Giannakoglou, K., and Othmer, C., 2010, “Adjoint Wall Functions: A New Concept for Use in Aerodynamic Shape Optimization,” J. Comput. Phys., 229(13), pp. 5228–5245. [CrossRef]
Zymaris, A., Papadimitriou, D., Giannakoglou, K., and Othmer, C., 2009, “Continuous Adjoint Approach to the Spalart-Allmaras Turbulence Model for Incompressible Flows,” Comput. Fluids, 38(8), pp. 1528–1538. [CrossRef]
Marta, A. C., and Shankaran, S., 2013, “On the Handling of Turbulence Equations in RANS Adjoint Solvers,” Comput. Fluids, 74, pp. 102–113. [CrossRef]
Nielsen, E. J., and Anderson, W. K., 1999, “Aerodynamic Design Optimization on Unstructured Meshes Using the Navier-Stokes Equations,” AIAA J., 37(11), pp. 1411–1419. [CrossRef]
Dwight, R. P., and Brezillon, J., 2006, “Effect of Various Approximations of the Discrete Adjoint on Gradient-Based Optimization,” AIAA Paper No. 2006-0690. [CrossRef]
Löhner, R., 2008, Applied CFD Techniques: An Introduction Based on Finite Element Methods, 2nd ed., John Wiley & Sons Ltd., Chichester, West Sussex, UK.
Amoignon, O. G., Pralits, J. O., Hanifi, A., Berggren, M., and Henningson, D. S., 2006, “Shape Optimization for Delay of Laminar-Turbulent Transition,” AIAA J., 44(5), pp. 1009–1024. [CrossRef]
Othmer, C., and Grahs, T., 2005, “Approaches to Fluid Dynamic Optimization in the Car Development Process,” International Conference on Evolutionary and Deterministic Methods for Design, Optimization and Control With Applications to Industrial and Societal Problems.
Othmer, C., Kaminski, T., and Giering, R., 2006, “Computation of Topological Sensitivities in Fluid Dynamics: Cost Function Versatility,” European Congress on Computational Methods in Applied Sciences and Engineering, ECCOMAS CFD, pp. 1–12.
Helgason, E., and Krajnović, S., 2012, “Aerodynamic Shape Optimization of a Pipe Using the Adjoint Method,” ASME International Mechanical Engineering Congress & Exposition, Nov. 9–15.

Figures

Grahic Jump Location
Fig. 1

The geometry used for the validation. The value of α is varied in both the x- and y-directions for each of the six cells marked on the figure, and the results are compared to the gradient obtained using the adjoint method.

Grahic Jump Location
Fig. 2

Gradient obtained using finite difference compared to the gradient obtained using the adjoint method. The step size varies from 1 × 10–8 to 1 × 10–4 depending on the location of the cell; (a) x-direction and (b) y-direction.

Grahic Jump Location
Fig. 3

(a) Artificial gradients are added to the domain and the resulting mesh motion is shown in (b)–(f). (b) After one loop through the mesh coupling algorithm. (c)–(f) show the mesh motion for different criteria on I. (c) I = 0.95, (d) I = 0.9, (e) I = 0.5, and (f) I = 0.1. Note that the vectors are not in the same scale.

Grahic Jump Location
Fig. 4

Flow chart of the optimization process. The dotted box contains the optimization loop used for smoothing the gradients. The cell sensitivities are used as an input to the mesh motion. When the convergence criterion is reached the mesh motion, s, is directly added to the location of the points in the mesh, denoted with x, and the loop starts again.

Grahic Jump Location
Fig. 5

The S-bend seen from the side and below. The flow direction is from left to right, as indicated by the arrows.

Grahic Jump Location
Fig. 6

The total pressure drop compared to the original design for two different criteria on I; (a) I = 0.5 and (b) I = 0.95

Grahic Jump Location
Fig. 7

(a) The original geometry seen from the side and top. The inlet is at the left and the outlet at the right. The original geometry can be compared to the final geometry obtained using two different conditions on I; (b) I = 0.5, and (c) I = 0.95.

Grahic Jump Location
Fig. 8

A cut along the center of the domain showing the velocity vectors (left) and contours of total pressure (right) for the original geometry, (a), and the final geometry obtained using two different conditions on I; (b) I = 0.5, and (c) I = 0.95

Grahic Jump Location
Fig. 9

A cut along the duct showing the impact on the structure of the mesh for the upper part of the first bend for the original geometry (a) and two different conditions on I, (b), and (c).

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In