0
Research Papers

Robust and Dynamically Consistent Model Order Reduction for Nonlinear Dynamic Systems

[+] Author and Article Information
David B. Segala

Naval Undersea Warfare Center,
1176 Howell Street,
Newport, RI 02841
e-mail: david.segala@navy.mil

David Chelidze

Department of Mechanical Engineering,
University of Rhode Island,
Kingston, RI 02881
e-mail: chelidze@egr.uri.edu

1Corresponding author.

Contributed by the Dynamic Systems Division of ASME for publication in the JOURNAL OF DYNAMIC SYSTEMS, MEASUREMENT, AND CONTROL. Manuscript received January 30, 2014; final manuscript received August 24, 2014; published online September 24, 2014. Assoc. Editor: Prashant Mehta.

J. Dyn. Sys., Meas., Control 137(2), 021011 (Sep 24, 2014) (8 pages) Paper No: DS-14-1049; doi: 10.1115/1.4028470 History: Received January 30, 2014; Revised August 24, 2014

There is a great importance for faithful reduced order models (ROMs) that are valid over a range of system parameters and initial conditions. In this paper, we demonstrate through two nonlinear dynamic models (pinned–pinned beam and thin plate) that are both randomly and periodically forced that smooth orthogonal decomposition (SOD)-based ROMs are valid over a wide operating range of system parameters and initial conditions when compared to proper orthogonal decomposition (POD)-based ROMs. Two new concepts of subspace robustness—the ROM is valid over a range of initial conditions, forcing functions, and system parameters—and dynamical consistency—the ROM embeds the nonlinear manifold—are used to show that SOD, as opposed to POD, can capture the low order dynamics of a particular system even if the system parameters or initial conditions are perturbed from the design case.

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

References

Chatterjee, A., 2000, “An Introduction to the Proper Orthogonal Decomposition,” Curr. Sci., 78(7), pp. 808–817.
Liang, Y. C., Lee, H. P., Lim, S. P., Lin, W. Z., Lee, K. H., and Wu, C. G., 2002, “Proper Orthogonal Decomposition and Its Applications—Part I: Theory,” J. Sound Vib., 252(3), pp. 527–544. [CrossRef]
Feeny, B. F., and Kappagantu, R., 1998, “On the Physical Interpretation of Proper Orthogonal Modes in Vibrations,” J. Sound Vib., 211(4), pp. 607–616. [CrossRef]
Volkwein, S., 2008, “Model Reduction Using Proper Orthogonal Decomposition, Lecture Notes,” Institute of Mathematics and Scientific Computing, University of Graz, Graz, Austria.
Kershen, G., Poncelet, F., and Golinval, J., 2007, “Physical Interpretation of Independent Component Analysis in Structural Dynamics,” Mech. Syst. Signal Process., 21(4), pp. 1567–1575.
Feeny, B., 2002, “On the Proper Orthogonal Modes and Normal Modes of a Continuous Vibration System,” J. Sound Vib., 124(1), pp. 157–160.
Shlens, J., 2005, A Tutorial on Principal Component Analysis, Institute for Nonlinear Science, University of California, San Diego, La Jolla, CA.
Berkooz, G., Holmes, P., and Lumley, J. L., 1993, “The Proper Orthogonal Decomposition in the Analysis of Turbulent Flows,” Annuals Rev. Fluid Mech., 25(1), pp. 539–575. [CrossRef]
Rega, G., and Troger, H., 2005, “Dimension Reduction of Dynamical Systems: Methods, Models, Applications,” Nonlinear Dyn., 41(1–3), pp. 1–15. [CrossRef]
Shaw, S., and Pierre, C., 1994, “Normal Modes of Vibration for Non-Linear Continuous Systems,” J. Sound Vib., 169(3), pp. 319–347. [CrossRef]
Shaw, S., and Pierre, C., 1991, “Non-Linear Normal Modes and Invariant Manifolds,” J. Sound Vib., 150(1), pp. 170–173. [CrossRef]
Shaw, S., and Pierre, C., 1993, “Normal Modes for Non-Linear Vibratory Systems,” J. Sound Vib., 164(1), pp. 85–124. [CrossRef]
Vakakis, A., 1997, “Non-Linear Normal Modes (NNMS) and Their Application in Vibration Theory: An Overview,” Mech. Syst. Signal Process., 11(1), pp. 3–22. [CrossRef]
Chelidze, D., and Liu, M., 2005, “Dynamical Systems Approach to Fatigue Damage Identification,” J. Sound Vib., 281(3–5), pp. 887–904. [CrossRef]
Chelidze, D., and Liu, M., 2008, “Reconstructing Slow-Time Dynamics From Fast-Time Measurements,” Philos. Trans. R. Soc. A, 366(1866), pp. 729–3087. [CrossRef]
Chelidze, D., and Cusumano, J. P., 2006, “Phase Space Warping: Nonlinear Time Series Analysis for Slowly Drifting Systems,” Philos. Trans. R. Soc. A, 364(1846), pp. 2495–2513. [CrossRef]
Chelidze, D., 2004, “Identifying Multidimensional Damage in a Hierarchical Dynamical System,” Nonlinear Dyn., 310(4), pp. 307–322. [CrossRef]
Chelidze, D., and Zhou, W., 2006, “Smooth Orthogonal Decomposition Based Modal Analysis,” J. Sound Vib., 292(3–5), pp. 461–473. [CrossRef]
Farooq, U., and Feeny, B., 2008, “Smooth Orthogonal Decomposition for Modal Analysis of Randomly Excited Systems,” J. Sound Vib., 316(1–5), pp. 137–146. [CrossRef]
Chelidze, D., 2013, “Smooth Robust Subspace Based Model Reduction,” ASME Paper No. DETC2013-13333. [CrossRef]
Chelidze, D., 2009, “Nonlinear Normal Mode Embedding for Model Reduction,” Euromech Colloquium 503: Nonlinear Normal Modes, Model Reduction and Localization, Frascati, Sept. 27–Oct. 2.
Segala, D., and Chelidze, D., 2013, “Robust and Dynamically Consistent Reduced Order Models,” ASME Paper No. IMECE2013-62522. [CrossRef]
Rathinam, M., and Petzold, L. R., 2003, “A New Look at Proper Orthogonal Decomposition,” J. Numer. Anal., 41(5), pp. 1893–1925. [CrossRef]
Kennel, M., Brown, R., and Abarbanel, H., 1992, “Determining Embedding Dimension for Phase-Space Reconstruction Using a Geometrical Construction,” Phys. Rev. A, 45(6), pp. 3403–3411. [CrossRef] [PubMed]

Figures

Grahic Jump Location
Fig. 1

A simply supported beam with two nonlinear springs positioned left and right of center

Grahic Jump Location
Fig. 2

A simply supported pate with a nonlinear spring positioned at the center node

Grahic Jump Location
Fig. 3

(Left) Subspace robustness and (Right) dynamical consistency for both periodically and randomly forced POD-based (∇ and , respectively) and SOD-based (◇ and ◯, respectively) ROMs for the beam

Grahic Jump Location
Fig. 4

(Left) Subspace robustness and (Right) dynamical consistency for both periodically and randomly forced POD-based (∇ and , respectively) and SOD-based (◇ and ◯, respectively) ROMs for the plate

Grahic Jump Location
Fig. 5

(Left) Phase portrait for the eight DOF's full scale trajectory with periodic forcing with amplitude f = 1.0 for the beam that we wish to reconstruct. (Right) Corresponding L2-norm between the full scale trajectory and the POD-based ROM (–) and the SOD-based ROM (- -) for the periodically forced beam.

Grahic Jump Location
Fig. 6

(Left) Phase portrait for the 21st DOF's full scale trajectory with periodic forcing with amplitude f = 100.0 for the plate that we wish to reconstruct. (Right) Corresponding L2-norm between the full scale trajectory and the POD-based ROM (–) and the SOD-based ROM (- -) for the periodically forced plate.

Grahic Jump Location
Fig. 7

(Left) Phase portrait for the eight DOF's full scale trajectory with random forcing with amplitude f = 0.6 for the beam that we wish to reconstruct. (Right) Corresponding L2-norm between the full scale trajectory and the POD-based ROM (–) and the SOD-based ROM (- -) for the randomly forced beam.

Grahic Jump Location
Fig. 8

(Left) Phase portrait for the 21st DOF's full scale trajectory with random forcing with amplitude f = 100.0 for the plate that we wish to reconstruct. (Right) Corresponding L2-norm between the full scale trajectory and the POD-based ROM (–) and the SOD-based ROM (- -) for the randomly forced plate.

Grahic Jump Location
Fig. 9

(Left) Phase portrait for the eight DOF's full scale trajectory with periodic forcing with amplitude f = 2.0 and random initial condition for the beam that we wish to reconstruct. Subspaces were constructed with system parameters f = 1.3 and zeroes as the initial condition. (Right) Corresponding L2-norm between the full scale trajectory and the POD-based ROM (–) and the SOD-based ROM (- -) for the periodically forced beam off-design configuration.

Grahic Jump Location
Fig. 10

(Left) Phase portrait for the 21st DOF's full scale trajectory with periodic forcing frequency of ω = 4π/5 and a random initial condition for the plate that we wish to reconstruct. Subspaces were constructed with system parameters ω = 3π/5 and zeroes as the initial condition. (Right) Corresponding L2-norm between the full scale trajectory and the POD-based ROM (–) and the SOD-based ROM (- -) for the periodically forced plate.

Grahic Jump Location
Fig. 11

(Left) Phase portrait for the eight DOF's full scale trajectory with zeroes as the initial condition and nonlinear spring coefficient β = 10 for the beam that we wish to reconstruct. Subspaces were constructed with system parameters zeroes as the initial condition and nonlinear spring coefficient β = 5. (Right) Corresponding L2-norm between the full scale trajectory and the POD-based ROM (–) and the SOD-based ROM (- -) for the randomly forced beam off-design configuration.

Grahic Jump Location
Fig. 12

(Left) Phase portrait for the 21st DOF's full scale trajectory with random initial condition and nonlinear spring coefficient β = 16 for the plate that we wish to reconstruct. Subspaces were constructed with system parameters zeroes as the initial condition and nonlinear spring coefficient β = 8. (Right) Corresponding L2-norm between the full scale trajectory and the POD-based ROM (–) and the SOD-based ROM (- -) for the randomly forced plate.

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