Arman Pazouki Simulation-Based Engineering Laboratory Department of - - PowerPoint PPT Presentation

MULTIPHYSICS SIMULATION USING GPU

Arman Pazouki Simulation-Based Engineering Laboratory Department of Mechanical Engineering University of Wisconsin - Madison

A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

Acknowledgements

• Prof. Dan Negrut
• Dr. Radu Serban
• Hammad Mazhar
• Andrew Seidl
• Colleagues from Simulation Based Engineering Laboratory, University of Wisconsin-

Madison

• National Science Foundation
• NVIDIA

2 A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

Motivation and Background

3 A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

Motivation

4 A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

Fluid Rigid Bodies Flexible Bodies

5 6/30/2014 A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems

Fluid Simulation: SPH

• Continuity:
• Momentum (Navier-Stokes):
• Lagrangian Kinematics:

6

• J. Monaghan, Smoothed particle hydrodynamics, Reports on Progress in Physics 68 (1) (2005) 1703-1759.
• M. Liu, G. Liu, Smoothed particle hydrodynamics (SPH): An overview and recent developments, Archives of Computational Methods in

Engineering 17 (1) (2010) 25-76.

A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

• Weakly Compressible model
• XSPH
• Shepard Filtering

a b

ab

r

W h

S

Neighbor Search and GPU programming

7

Case Study: How GPU can affect an algorithmic design

A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

VS

Algorithm 1

8

Core property: saving contacts list Parallel threads: Bins Advantages

• One calculation per intersection
• Possibility of re-using contacts list
• Arbitrary shapes

A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

Successful for rigid body dynamics: O(1e7) Failed for SPH: O(1e6)

Algorithm 2

9

Core property: find intersection whenever needed Advantage.

• More process, less memory
• Fixed size spheres

Parallel threads: Particles

• Reduces memory access
• Improves cache hits.

A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

Screen shot from particles demo, NVIDIA CUDA (CUDA Samples)

Rigid and Flexible Bodies Dynamics

10

q q q q q q q q q q q q

i y z y x w z z w x y w z z x y x y x

                                      G 

• 3D rigid body dynamics

1

T i i  

q q

d 1 , d 2

T i i i

t  q G  d d

i i

t  X V

i i i i i

d dt         ω J T J ω 

d d

i i i

t M  V F

Dynamics Kinematics

• E. Haug, Computer aided kinematics and dynamics of mechanical systems, Allyn and Bacon Boston, 1989.
• A. Shabana, Dynamics of Multibody Systems, Cambridge University Press, 2005.

A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

• Gradient-deficient ANCF beam element

11 11 s

( )                       

  

Me + Q Q Q Q = e e Q = S g Q = S F

e g a T T e g T a T a

EA dx+ EI dx A dx x

Dynamics

3 12

; ( ) = ( ) ; ( ) = ( ) shapefunction matrix



               r r e = r e S e v e S e r r S

L L x R R x

x, x x, x

Kinematics

Fluid-Solid Interaction

11

Boundary Condition Enforcing (BCE) markers for no-slip condition

• Rigidly attached to the solid body (hence their velocities are those of the corresponding material points on the solid)
• Hydrodynamic properties from the fluid

Rigid bodies/walls Flexible beams

A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

Parallelization

12 A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

• thrust::reduce_by_key to reduce surface reaction forces and torques on to nodal

values

• Custom kernels to update solid objects
• Fine grain parallelization
• Position
• Rotation
• Velocity
• Angular velocity
• …

Example Simulations

13 A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

Flow in porous media

• Example applications
• Oil Recovery
• Biology
• Diffusion of macro-

molecules within tissues

• Blood flow through

muscles

14 A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

Simulation of dense suspensions

• “Dense” suspension
• Finite size particles (rigid bodies)

interaction

• Drafting, Kissing, and Tumbling (DKT)
• Short range interactions
• Lubrication and collision
• Flow characteristics
• Particle Reynolds number ≤ 1.0
• Channel Reynolds number: 66
• Channel Dimension: (1.1, 1.0, 1.0) m
• Volumetric concentration: 40%
• Computational aspects
• 23,000 rigid ellipsoids: (1.5, 1.5, 2.0) cm
• 2,000,000 SPH markers.
• Simulation performed on a single GPU,

NVIDIA GTX 480

• 3.2 seconds of dynamics
• 72 hrs to complete.

15 A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

Animation shows the channel mid-section

Interacting rigid and flexible objects in channel flow

16 A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

Performance

17 A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

Scaling analysis on NVIDIA GeForce GTX 680

18

• Rigid body dynamics
• Flexible body dynamics
• Fluid flow

A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

Scaling analysis (all together, table)

19 A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

Validation

20 A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

Particle migration in 2D and 3D Poiseuille flow

• Transient Poiseuille flow

21

• Sphere in pipe flow,
• Effect of rigid body rotation
• Cylinder in channel flow

APS- Pittsburgh

• T. Inamuro, K. Maeba, F. Ogino, IJMF 26 (12) (2000) 1981-2004.
• J. Morris, P. Fox, Y. Zhu, JCP 136 (1) (1997) 214-226.
• J. Schonberg, E. Hinch, JFM 203 (1) (1989) 517.
• D. Oliver, Nature (194) (1962) 1269-1271.
• B. H. Yang, J. Wang, D. D. Joseph, H. H. Hu, T.-W. Pan, and R. Glowinski, JFM 540 (2005) 109.

A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

22

Increasing distance from inlet

Radial distribution of particles in suspension [1/2]

0.07 ( )( )( ) [0,0.69] 60, 0.027%         a R a av l L R R Re

• G. Segre, A. Silberberg, Nature (189) (1961) 2.

A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

• 192, 10 hour long simulation
• 14 seconds real time
• Bootstrapping method,95% confidence interval

Applications

23 A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

Hanging flexible beam in viscose fluid

24 A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

• Flexible cantilever in contained fluid
• Track position of beam tip

Flow cytometry using microfluidic techniques

• Fluorescence and laser-beam cell sorting
• Limited particle size,
• Unknown effect of external field on cell viability
• Purification of 3D micro-tissues and cell aggregates
• Finite size particles,

25 A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

50 m   a 25..500 m   a

Work in progress

26 A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015

Tha hank nk Yo You!

Chrono::FSI (Project Chrono: https://github.com/projectchrono) Simulation Based Engineering Lab (SBEL) University of Wisconsin-Madison Email: pazouki@wisc.edu

27 A Lagrangian-Lagrangian Approach For the Simulation of FSI Problems 3/25/2015