solving large complex problems: 'efficient and smart solutions … · 2012-10-26 · 1 © 2011...

55
© 2011 ANSYS, Inc. June 7, 2012 1 Solving Large Complex Problems: “Efficient and Smart Solutions for Large Models” Presented at the 2012 Confidence by Design Detroit, June 5

Upload: others

Post on 10-Aug-2020

7 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 1

Solving Large Complex Problems: “Efficient and Smart Solutions for Large Models”

Presented at the 2012 Confidence by Design

Detroit, June 5

Page 2: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 2

ANSYS Structural Mechanics Solutions offers several techniques

Page 3: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 3

Current trends in simulation show an increased need for

the computation of large models

Page 4: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 4

Larger model size for higher accuracy

Page 5: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 5

Full assemblies for higher realism

Page 6: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 6

Better design knowledge from design variations

Page 7: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 7

Continuous variations of the parameters provides in-depth information

Page 8: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 8

Even more performance is required!

( ) + + =

x n

Page 9: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 9

High Performance Computing is a key component of the

ANSYS offering.

Page 10: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 10

High Performance Computing as an Enabler Insight you can’t get any other way

It’s all about getting better insight into product behavior quicker!

HPC enables high-fidelity • Include details - for reliable results

• “Getting it right the first time”

• CONFIDENCE by DESIGN!

HPC enables design exploration & optimization • Consider multiple design ideas

• Optimize the design

• Ensure performance across range of conditions

Page 11: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 11

High Performance Computing A Software Development Imperative

• Clock Speed – Leveling off

• Core Counts – Growing

• Exploding (GPUs)

• Future performance depends on highly scalable parallel software

Source: http://www.lanl.gov/news/index.php/fuseaction/1663.article/d/20085/id/13277

Page 12: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 12

Size of the model – how large is “large”?

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

1 2 3 4

Number of CPUs

Sp

eed

Up 1000 elements

8000 elements64000 elements512000 elements

Page 13: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 13

A simple and productive licensing scheme

ANSYS HPC Pack

ANSYS HPC Workgroup

Page 14: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 14

Which part of the simulation is faster?

Page 15: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 15

Not all steps of the simulation are parallel

Model solution

Page 16: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 16

How should I read speed-up curves? This is the solver part – excellent scaling!

This is YOUR time (elapsed)

Page 17: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 17

The right combination of algorithms and

hardware leads to maximum efficiency

Page 18: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 18

Shared Memory Parallel vs Distributed Memory Parallel

Page 19: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 19

Challenges and solutions for the distributed method

Challenges Solutions

Efficient and relevant decomposition

Partitioning methods, Solver

Load Balancing

Partitioning methods, Solver

Speed Hardware(Processors, Interconnects), Solver

Maximum Problem Size Hardware (RAM), Solver

I/O to communicate between cores.

Hardware (Interconnects), MPI, Solver

I/O to write results and overflow files during solution.

Hardware (Disks, Interconnects), MPI, Solver

Page 20: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 20

Sparse or iterative solvers?

Solver type Distributed/Shared Memory

SPARSE (direct)

DMP/SMP

PCG (Iterative)

DMP/SMP

LANB (direct, modal) SMP

LANPCG (iterative, modal) DMP/SMP

SNODE SMP

Page 21: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 21

Get it in-core!

DSPOPTION: Sets memory option for the distributed sparse solver BCSOPTION: Sets memory option for the sparse solver •INCORE - obtain enough memory to run with the entire factorized matrix in memory. avoid doing any I/O. •OPTIMAL - allocate enough work space to keep the largest front matrix in memory during factorization. As a result, this memory mode often achieves an optimal balance between I/O and memory usage. •MINIMUM - allocate the least amount of memory needed to run through the solver

Page 22: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 22

Check the PCG level!

Page 23: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 23

Balancing the load: a key to efficiency

Page 24: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 24

A consequence for contact users

Page 25: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 25

What could it look like on your model?

6 Mio Degrees of Freedom Plasticity, Contact Bolt pretension 4 load steps

1 HPC Pack

Page 26: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 26

The Right Software Architecture for HPC

Workstation-Based or

Remote Access

Job Submission Host

ANSYS Workbench

User Environment/Graphics

Compute Cluster

Mobile User Thin Client

Off-the-shelf High-Performance Computing ANSYS Remote Solve Manager: Workbench-based job submission with full portfolio support for Platform LSF, PBS Pro, and Microsoft Job Scheduler Bundled third-party Message-Passing software with optimized performance (Intel MPI, Platform MPI) on gigE, 10gigE, or Infiniband cluster fabric.

Multi-core / Multi-node

Solver Execution

Graphics Server

Page 27: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 27

Our HPC Partnerships

ANSYS maintains close technical collaboration with the leaders in HPC

This mutual commitment ensures that you get the most possible value from your overall HPC investment

Some current examples:

• Optimized performance on multicore processors from Intel, with R&D focused on Intel’s Many Integrated Core (MIC)

• Over 60% performance boost for the latest Intel® Xeon® E5-2600 processor (Sandy Bridge) family compared to previous Intel (Westmere) generation

• GPU computing accelerates ANSYS Mechanical today, with very active R&D engagement with NVIDIA across full portfolio

• ANSYS and IBM – Optimized cluster and storage architectures for ANSYS

Page 28: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 28

What do your peers say?

“By optimizing our solver selection and workstation configuration, and including GPU acceleration, we’ve been able to dramatically reduce turnaround time — from over two days to just an hour. This enables the use of simulation to examine multiple design ideas and gain more value out of our investment in simulation.”

- Berhanu Zerayohannes, Senior Mechanical Engineer, NVIDIA

Application: Deflection and bending of 3-D glasses Software: ANSYS Mechanical

HPC Solution: From 60 hours per simulation to 47 minutes (77x speedup)

Business Ability to ensure robust performance of the 3-D Solution: glasses via examining multiple design ideas

Co

pyr

igh

t 20

11

NV

IDIA

Co

rpo

rati

on

. A

ll ri

ghts

res

erve

d.

Page 29: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 29

Taking advantage of new hardware solutions: GPU

1.52

1.16

1.70

1.20

2.24

1.44

0

1

2

3

V13cg-1(JCG,

1100k)

V13sp-1(sparse,

430k)

V13sp-2(sparse,

500k)

V13sp-3(sparse,2400k)

V13sp-4(sparse,1000k)

V13sp-5(sparse,2100k)

Distributed ANSYS 14.0 Total Simulation Speedups

4 CPU cores

4 CPU cores + 1 GPU

Page 30: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 30

Speed-up from Multi-Node GPU technology

Solder Joint Benchmark - 4M DOF, Creep Strain Analysis

Linux cluster : Each node contains 8 Intel Xeon 5600-series cores, 96 GB RAM, NVIDIA Tesla M2070, InfiniBand

Mold

PCB

Solder balls

Res

ult

s C

ou

rtes

y o

f M

icro

Co

nsu

lt E

ngi

nee

rin

g, G

mb

H

Page 31: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 31

ANSYS has a strong commitment to High Performance Computing.

Page 32: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 32

High Performance Computing is dynamic, technology dependent.

1980

2010 2020

Page 33: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 33

Reduction techniques help

reduce the CPU time and can help

compute design variations more

efficiently

Page 34: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 34

Submodeling is the solution when only a portion of the model matters.

Page 35: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 35

The coarse model provides accurate deformations but inaccurate stresses

Page 36: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 36

The refined model(s) will provide accurate stresses

Page 37: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 37

From the coarse model to the submodel through results mapping

Displacements are mapped to the common boundary

Page 38: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 38

Solving two models can be faster than solving a very detailed one

Page 39: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 39

General Procedure

1. Create and analyze the initial model

2. Create the submodel

3. Perform cut boundary interpolation

4. Analyze the submodel

5. Verify that the distance between the cut boundaries and the stress concentration is adequate

Page 40: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 40

Good practice: Verify the cut-boundary distance

Compare path plots at that location between the initial model and submodel.

Use the query option.

List the results.

Etc.

Page 41: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 41

Submodeling with ANSYS Workbench

Page 42: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 42

Page 43: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 43

Submodeling works for a variety of topologies and

nonlinear models as well

Page 44: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 44

From a solid model to a solid model

Stress contour – full model

Stress contour – Submodel

Page 45: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 45

From a shell model to a solid model

Initial geometry Defeatured shell model - deformations

Solid submodel - stresses

Page 46: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 46

Nonlinearities can also be included

Plastic strain

Coarse model

Submodel (line) vs refined full model (cross)

Page 47: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 47

Substructuring or CMS allows for collaborative work or long transient simulations.

Page 48: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 48

Static (Guyan)

Guyan Reduction procedure

inertia forces are negligible compared to elastic forces

Net result: the reduced stiffness matrix is exact, whereas the reduced mass and damping matrices are approximate

Note: Choosing master DOF is an important step in a reduced analysis, impacting accuracy of results

Page 49: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 49

Component Mode Synthesis

CMS is a type of substructuring which performs a modal analysis of a structure based on independent modal analyses of its parts

The synthesis involves making the components work together as a single structure by satisfying inter-component compatibility and equilibrium constraints

Master DOF are required only at interface nodes

Page 50: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 50

CMS with ANSYS Workbench

Page 51: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 51

APDL macros embedded in the simulation tree for generation, use and expansion pass

Page 52: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 52

Results are available through standard operations

Page 53: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 53

Reduce solution time for harmonic and transient analyses

Page 54: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 54

Comparing the accuracy of a CMS analysis to a standard one

Page 55: Solving Large Complex Problems: 'Efficient and Smart Solutions … · 2012-10-26 · 1 © 2011 ANSYS, Inc. June 7, 2012 . Solving Large Complex Problems: “Efficient and Smart Solutions

© 2011 ANSYS, Inc. June 7, 2012 55

Learn about the methods available in ANSYS Structural Mechanics Solutions.