0
Research Papers

Efficient Parallel Simulation of Large Flexible Body Systems With Multiple Contacts

[+] Author and Article Information
Naresh Khude, Dan Negrut

Department of Mechanical Engineering,
University of Wisconsin–Madison,
Madison, WI 53706

Ilinca Stanciulescu

Department of Civil and Environmental Engineering,
Rice University,
Houston, TX 77005

Contributed by the Design Engineering Division of ASME for publication in the Journal of Computational and Nonlinear Dynamics. Manuscript received February 11, 2012; final manuscript received January 30, 2013; published online March 26, 2013. Assoc. Editor: Arend L. Schwab.

J. Comput. Nonlinear Dynam 8(4), 041003 (Mar 26, 2013) (11 pages) Paper No: CND-12-1030; doi: 10.1115/1.4023915 History: Received February 11, 2012; Revised January 30, 2013

This contribution outlines a computational framework for the analysis of flexible multibody dynamics contact problems. The framework combines a flexible body formalism, specifically, the absolute nodal coordinate formulation (ANCF), with a discrete continuous contact force model to address many-body dynamics problems, i.e., problems with hundreds of thousands of rigid and deformable bodies. Since the computational effort associated with these problems is significant, the analytical framework is implemented to leverage the computational power available on today's commodity graphical processing unit (GPU) cards. The framework developed is validated against commercial and research finite element software. The robustness and efficiency of this approach is demonstrated through numerical simulations. The resulting simulation capability is shown to result in 2 orders of magnitude shorter simulation times for systems with a large number of flexible beams that might typically be encountered in hair or polymer simulations.

Copyright © 2013 by ASME
Your Session has timed out. Please sign back in to continue.

References

Figures

Grahic Jump Location
Fig. 1

Schematic of CCFM contact

Grahic Jump Location
Fig. 2

Spherical decomposition along the centerline of a deformed beam. The decomposition of the beam geometry is subsequently used for fast collision detection purposes.

Grahic Jump Location
Fig. 3

Time looping stage of the ANCF+CCFM approach

Grahic Jump Location
Fig. 4

X coordinate of pendulum-tip (noncontact case)

Grahic Jump Location
Fig. 5

Y coordinate of pendulum-tip (noncontact case)

Grahic Jump Location
Fig. 6

X coordinate of pendulum-tip (contact case)

Grahic Jump Location
Fig. 7

Y coordinate of pendulum-tip (contact case)

Grahic Jump Location
Fig. 8

X displacement of pendulum-tip (contact case)

Grahic Jump Location
Fig. 9

Y displacement of pendulum-tip (contact case)

Grahic Jump Location
Fig. 10

X displacement of pendulum-tip (ANCF, ABAQUS, and FEAP comparison)

Grahic Jump Location
Fig. 11

Y displacement of pendulum-tip (ANCF, ABAQUS, and FEAP comparison)

Grahic Jump Location
Fig. 12

Z displacement of pendulum-tip (ANCF, ABAQUS, and FEAP comparison)

Grahic Jump Location
Fig. 13

Snapshots from simulation of model-2

Grahic Jump Location
Fig. 14

Snapshots from simulation of model-3

Grahic Jump Location
Fig. 15

Snapshot from parametric study of model-3

Grahic Jump Location
Fig. 16

X displacement of a pendulum-tip (E = 2e8)

Grahic Jump Location
Fig. 17

Y displacement of a pendulum-tip (E = 2e8)

Grahic Jump Location
Fig. 18

X displacement of a pendulum-tip (E = 2e9)

Grahic Jump Location
Fig. 19

Y displacement of a pendulum-tip (E = 2e9)

Grahic Jump Location
Fig. 20

ANCF, sequential implementation: scaling analysis performed for various numbers of model-1 beams (contact is neglected)

Grahic Jump Location
Fig. 21

The GPU is specialized for compute-intensive, highly data parallel computation due to its graphics rendering origin. This is manifested in microprocessor designs that have a very large number of arithmetic logical units.

Grahic Jump Location
Fig. 22

Overall speedup when comparing the CPU off-the-shelf sequential algorithm to the GPU parallel algorithm. The maximum speedup achieved was approximately 180 times.

Grahic Jump Location
Fig. 23

ANCF scaling analysis reports processing time for the CPU and GPU implementations for varying numbers of beams

Grahic Jump Location
Fig. 24

100 ANCF beam elements dropped on the ground [51]. The beams interact through Hertzian contact. A scaling analysis was run using this model with up to 1 × 106 beams interacting through frictional contact.

Grahic Jump Location
Fig. 25

ANCF+CCFM scaling analysis reports simulation time for the GPU implementation of flexible beams in contact. Continuous line is a curve fit that approximates the dotted line, obtained through scaling analysis. Note that the scaling is almost linear in the number of elements, the coefficient of the quadratic term being very small, of the order 10−9.

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