1. Advanced FEM and Meshfree Class
Workshop
2016
Detroit
W. Hu*, C.T. Wu, Y. Guo, B. Ren and Y. Wu
Livermore Software Technology Corporation
2. Preparations for the Workshop
Suggestions:
1. Always create a new directory for a new job run.
2. Keep all the directories in the same workshop.
3. Help each other.
Executables:
1. Beta: referred to double precision Beta version of LS-DYNA.
2. R90: referred to double precision R9.0 version of LS-DYNA.
EFG and SPG particle plotting in d3plot:
1. Click on Model and Part Appearance from the right-hand side options.
2. In Appearance, check Sphere and Shrn, then click on target parts. (Mesh turns to grids)
3. Choose Settings General settings on top menu bar.
4. In General settings, check SPH/Particle and set appropriate values of Radius and Divs.
5. In General settings, choose Smooth as Style and check Fixed Radius, then Apply.
2
4. Example 1: Punch Test
Filename: kernel_lag.k kernel_eul.k
Description:
A block of elastic material modeled by 1000 8-noded elements is
subjected to a prescribed displacement on the top surface.
You are expected to do the following studies by Beta / R90:
1. Create a new directory under Example 1 and copy kernel_lag.k
to this directory.
2. Obtain the results using Lagrangian kernel.
3. Create another new directory and copy kernel_eul.k to this
directory.
4. Obtain the results using Eulerian kernel.
6. Compare the deformation and reaction force of two kernels, and
understand the difference.
7. Try different support size.
Objective: Understand the difference between
Lagrangian and Eulerian kernels.
d
Lagrangian
Eulerian
4
6. Example 1: Punch Test (Cont.)
Set DX=DY=DZ=1.1 (*SECTION_SOLID_SPG)
Eulerian kernel Lagrangian kernel
• Recommended DX,DY,DZ in SPG method: 1.4~1.8
• Smaller support may leads to unstable result and tensile instability
• Lager support costs more computational time and memory
TSSFAC (*CONTROL_TIMESTEP)
SMSTE (*SECTION_SOLID_SPG)
6
7. Example 2: Foam Compression
Filename: foam_fem4.k foam_efg4.k foam_sgp.k
Description:
A block of low-density foam modeled by 6000 4-noded elements is
subjected to a rigid footing.
You are expected to do the following studies by Beta / R90:
1. Create three new directories under Example 2 and copy three
input files into them, respectively.
2. Obtain the results using three formulations: FEM tetrahedron
element, EFG 4-noded with semi-Lagrangian kernel, and SPG
with switching between Lagrangian and Eulerian kernels.
3. Perform EFG particle plotting if necessary.
4. Compare the resultant contact forces.
Objective: Improve the element distortion problem
in the solid analysis using various formulations.
d
7
8. Example 2: Foam Compression (Cont.)
FEM
EFG SPG
Contact force
8
10. Example 3: Taylor Bar Impact
Filename: tbar_efg8.k tbar_spg.k tbar_sph.k
Description:
Perform analyses using: 1. Stabilized EFG with 8-noded elements
2. SPG with Lagrangian/Eulerian kernel
3. SPH (Eulerian kernel)
You are expected to do the following studies by Beta / R90:
1. Create four new directories under Example 3 and copy three
input files into them, respectively.
2. Obtain the results using four formulations.
3. Perform EFG particle plotting if necessary.
4. Plot the effective plastic strain contour.
5. Compare the deformed shapes and change of time step size.
Objective: Improve the element distortion problem in the
solid analysis using various formulations.
v0
10
12. Example 4: Rubber Bushing
Filename: rubberb_efg8.k rubberb_efg4.k rubberb_mefem.k
Description:
Perform analyses using: 1. Stabilized EFG with 8-noded elements
2. 4-noded EFG with pressure smoothing
3. MEFEM with 5-noded elements
You are expected to do the following studies by R90:
1. Create three new directories under Example 4 and copy three
input files into them, respectively.
2. Obtain the results using three formulations.
3. Perform EFG particle plotting if necessary.
4. Compare the deformed shapes and contact force responses.
5. Change settings of stabilized EFG and check the results
Objective: Study the rubber behavior using various
formulation.
12
13. Example 4: Rubber Bushing (Cont.)
EFG
stabilized
8-noded
EFG
pressure
smoothing
4-noded
MEFEM
Contact force
13
14. Example 5: Shell Inflation
Filename: shell_inf.k
Description:
A shallow cap that contains 48 shell elements is subjected to a pressure
loading.
You are expected to do the following studies by Beta / R90:
1. Create a new directory for Example 5 and copy input file to this
directory.
2. Obtain the finite element results (FEM#1and/or #2 and/or #16).
3. Create a EFG directory and copy input file to this directory.
4. Modify the *SECTION card to be the EFG formulation (#41)
with suggested parameters (1.1, 1.1, , , 3) and submit the job.
5. Compare the tip displacement in the z-direction.
Objective: Improve the element distortion problem in the
shell analysis.
Pressure
FEM 1
EFG 41
Tip disp.
14
15. Example 6: Shell Tube Crushing
Filename: shell_tubecrush.k
Description:
Crushing test in an aluminum composite tube.
You are expected to do the following studies by Beta / R90:
1. Create a new directory for Example 6 and copy input file to this
directory.
2. Obtain the finite element results (FEM#1and/or #2 and/or #16).
3. Create a EFG directory and copy input file to this directory.
4. Modify the *SECTION card to be the EFG formulation (#41)
with suggested parameters (1.1, 1.1, , , -1,2) for all three parts
and add the *CONTROL_EFG cards with EFGPACK
(IMLM=2), then submit the job.
5. Compare the crushing force in the x-direction.
Objective: Improve the element distortion problem in the
crushing tube analysis.
FEM 16
EFG 41
Crushing force
15
17. Example 1: Taylor Bar Impact – Task 1
Filename: tbar_adp.k
You are expected to do the following studies by Beta / R90:
Objective: Improve the accuracy of adaptive remapping.
V
1. Submit the job using EFG without adaptivity.
2. Create a new directory and copy the input file to this directory.
3. Modify the file for the adaptive analysis as follows:
4. Submit the job.
5. Plot the effective plastic strain contour and internal energy (ASCII glstat).
6. Change IVT (in Card 2 of *CONTROL_REMESHING_EFG) as IVT=-2,2,-1,1;
7. Re-run the jobs and compare the results.
8. Set IPS=1 (in Card 3 of *SECTION_SOLID_EFG) and IVT=1
9. Re-run the job and compare the results.
Add *CONTROL_ADAPTIVE as
*CONTROL_ADAPTIVE
5.0e-6, 1.0, 7, 4, , 1.0E+20
, 1
Add *CONTROL_REMESHING as
*CONTROL_REMESHING_EFG
0.0003, 0.001
-2
Change *PART as
*PART
1, 1, 1, 0, 0, 0, 2, 0
17
18. Example 1: Taylor Bar Impact – Task 1 (Cont.)
IPS=1
IVT=1
IVT=2
IVT=-2
IVT=1
IVT=-1 18
19. Example 1: Taylor Bar Impact – Task 2
Filename: tbar_adp.k
You are expected to do the following studies by Beta / R90:
Objective: Impose constraints in adaptive analysis.
V
1. Modify the input file as follows:
2. Submit the job.
3. Plot the effective plastic strain contour and internal energy.
Remove all
*CONSTRAINED_GLOBAL
key cards
Add *CONSTRAINED_LOCAL for all three constraints as
*CONSTRAINED_LOCAL
1, 7, 1, 0.0, 0.0, 0.0, 1
…
Add *DEFINE_COORDINATE_SYSTEM for *CONSTRAINED_LOCAL as
*DEFINE_COORDINATE_SYSTEM
1, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0
1.0, 1.0, 0.0
19
20. Example 2: Sphere Forging
Filename: sphereforging_adp.k
You are expected to do the following studies by Beta / R90:
Objective: Study the interactive adaptivity.
1. Submit the job without interactive adaptivity.
2. Create a new directory and copy the input to this directory.
3. Modify the file using interactive adaptivity as follows:
4. Submit the jobs and monitor the number of adaptive steps.
5. Plot the effective plastic strain, internal energy and contact force (ASCII rcforc).
Remove
*CONTROL_REMESHING
key card
Add *CONTROL_REMESHING_EFG as
*CONTROL_REMESHING_EFG
0.0025, 0.02
, 2, 1
0.5, 5.0, 0.8
0.0025, 0.02
, 3, 1
0.5, 5.0, 0.8
0.0025, 0.02
, 1, 0
0.5, 5.0, 0.8
OR OR
20
21. Example 2: Sphere Forging (Cont.)
Contact force
Internal energy
ADPFREQ
21
22. Example 3: Simple Forging
Filename: simforging_adp.k
You are expected to do the following studies by Beta / R90:
Objective: Study local adaptivity and monotonic meshing.
1. Submit the job without local adaptivity.
2. Create a new directory and copy the input to this directory.
3. In order to have local refinement, change ADPENE (in Card 2 of *CONTROL_ADAPTIVE)
as ADPENE=1.0, 2.0, 3.0, 4.0.
5. Submit the jobs and compare the adaptive meshes.
Initial mesh: HEX
22
24. Objective: Study thermal-mechanical coupling and spring back.
Example 4: Cylinder Hot Forging
Filename: cylinder_hot_adp.k cylinder_hot_spbk.k
You are expected to do the following studies by Beta / R90:
1. Go through important keywords in this example as follows:
2. Submit the job using cylinder_hot_adp.k.
3. Create a new directory;
Copy *dynain and cylinder_hot_spbk.k to this directory.
4. Submit the job using cylinder_hot_spbk.k in the new directory.
5. Plot the temperature and stress contour using EFG particles.
*INTERFACE_SPRINGBACK_LSDYNA
*CONTROL_SOLUTION
*CONTROL_THERMAL_SOLVER
*CONTROL_THERMAL_TIMESTEP
*MAT_THERMAL_ISOTROPIC
*INITIAL_TEMPERATURE_SET
Temperature
von Mises Stress
von Mises Stress (spring back)
24
26. Example 1: Simple Forging
Filename: simforging_adp.k
You are expected to do the following studies by Beta / R90:
Objective: Study local adaptivity and monotonic meshing.
1. Submit the job with local adaptivity and no monotonic meshing.
2. Create a new directory and copy the input to this directory.
3. Modify the input file using monotonic meshing as follows:
4. Submit the job and compare the adaptive meshes.
Remove
*CONTROL_REMESHING
key card
Add *CONTROL_REMESHING_EFG as
*CONTROL_REMESHING_EFG
0.5, 3.0
, 1, , , 1
Initial mesh: HEX
26
27. Example 1: Simple Forging (Cont.)
ADPENE=3.0
No monotonic
ADPENE=3.0
Monotonic
28. Example 2: Cutting
Filename: cutting_adp.k
You are expected to do the following studies by Beta / R90:
Objective: Apply implicit EFG adaptivity to cutting problem.
1. Go through important keywords in this example as follows:
2. Submit the job.
3. Plot the effective plastic strain contour.
4. Plot the contact force (binout file: rcforc).
*SECTION_SOLID_EFG
*CONTROL_ADAPTIVE
*CONTROL_REMESHING_EFG
*CONTROL_IMPLICIT_GENERAL
*CONTROL_IMPLICIT_AUTO
*CONTROL_IMPLICIT_SOLUTION
*CONTACT_FORMING_SURFACE_TO_SURFACE_MORTAR
Contact force
28
29. Objective: Study 3D adaptivity with orbital remesher.
Example 3: Orbital Forming
Filename: orbital_adp.k
You are expected to do the following studies by Beta / R90:
1. Go through important keywords in this example as follows:
2. Submit the job.
3. Plot the effective plastic strain contour.
*CONTROL_REMESHING
0.3, 0.6, , , , , 5, 4.0
*DEFINE_COORDINATE_SYSTEM
5, …
…
*PART
1, 1, 1, 0, 0, 0, 3
29
30. Objective: Study user’s adaptivity control files .
Example 4: Multiple Adaptive Forming
Filename: multiparts_adp.k, adapt.fc1, adapt.fc2
You are expected to do the following studies by Beta / R90:
1. Go through important keywords as follows:
2. Submit the job.
3. Change control parameters in control files and test.
*CONTROL_ADAPTIVE
…
…
…
,,,,,,, 1 (IADPFC)
30
adapt.fc1 (control file)
1 0.10000000E-03 0
adapt.fc2 (control file)
2
3 0.0 0.006 0.05 0.15
5 0.0 0.025 0.10 0.25
31. Objective: Study user’s adaptivity control files to
perform material cut by CAD tool.
Example 5: Material Separation by CAD
Filename: matsep_adp.k, adapt.fc1, adapt.fc2
You are expected to do the following studies by Beta / R90:
1. Go through important keywords as follows:
2. Submit the job.
3. Use LS-Prepost to generate new mesh by removing a thin
layer of elements when LS-DYNA requests manual editing
on adaptive mesh “user.mesh”.
*CONTROL_ADAPTIVE
…
…
…
,,,,,,, 1 (IADPFC)
31
adapt.fc1 (control file)
1 0.01 3
adapt.fc2 (control file)
1
3 0.0 0.0101 0.02 0.20
33. Example 1: Ball Impacting Plate
Filename:
ballimp_spg.k
Description:
Material damage and separation in the plate using SPG.
You are expected to do the following studies by Beta:
Objective: Study the material failure and separation
in a ductile material.
1. Submit the job.
2. Perform EFG particle plotting.
3. Plot the effective plastic strain contour.
4. Test both Lagrangian and Eulerian kernels.
5. Test phenomenological material damage.
Side view
Top view
v0
33
34. Example 1: Ball Impacting Plate (Cont.)
Kernel=0
Lagrangian
No material failure
Kernel=1
Eulerian
Numerical failure
34
35. Example 2: Metal Shearing
Filename:
metalshear_spg.k
Description:
Material damage and separation in cutting process.
You are expected to do the following studies by Beta:
Objective: Study ductile material separation using SPG.
1. Submit the job.
2. Perform EFG particle plotting.
3. Plot the effective plastic strain contour.
4 Try different SPG and damage parameters.
35
36. Example 3: Kalthoff Plate Cracking
Objective: Study the cracks propagation in 2D brittle material using X-FEM.
36
Maximum tensile stress criterion
Filename: impact50.k
Description:
Crack is formed and propagated due to the impact.
You are expected to do the following studies by Beta / R90:
1. Submit the job.
2. Plot HistVar #1 (damage output of XFEM).
*DATABASE_EXTENT_BINARY
0 1
37. Example 4: Internal Pressure in Container
Objective: Study the cracks propagation in 3D shell (semi-brittle) using X-FEM.
37
Effective plastic strain criterion
Filename: container.k
Description:
Crack is formed and propagated due to internal pressure.
You are expected to do the following studies by Beta / R90:
1. Submit the job.
2. Plot HistVar #1 (damage output of XFEM).
*DATABASE_EXTENT_BINARY
0 1
38. Example 5: Twisted Cylinder
Objective: Study the cracks propagation in 3D shell (ductile) using X-FEM.
38
Filename: cylshelltwist.k
Description:
Crack is formed and propagated due to twisting.
You are expected to do the following studies by Beta / R90:
1. Submit the job.
2. Plot HistVar #1 (damage output of XFEM).
*DATABASE_EXTENT_BINARY
0 1
Effective plastic strain criterion
39. Example 6: Mode I Crack Propagation
Filename: peri_plate.k
Description:
A plate with a circular whole is stretched.
Cracks are formed and propagated automatically in the plate.
You are expected to do the following studies by Beta:
1. Submit the job.
2. Plot effective plastic strain. “EPS” in Peridynamics represents
the damage value.
3. Change DR to 0.6 or 1.0. Check the corresponding “The max.
peri neighbors” and “The min. peri neighbors”.
Objective: Study the cracks initiation and
propagation using Peridynamics.
39
Modify a FEM model to a “discete” mesh readable to Peridynamics
LS-Prepost: Element Tool ->Detach
40. Example 7: Simulation of Fragmentation
Filename: peri_glassimpact.k
Description:
A glass plate is impacted by a rigid ball. Complex damage patterns are
formed in the plate automatically.
You are expected to do the following studies by Beta:
1. Submit the job.
2. Plote effective plastic strain. “EPS” in Peridynamics represents
the damage value.
3. Change G in MAT_ELASTIC_Peri to 4.0 or 16.0. Check “The
max. peri critical bond stretch” and “The min. peri critical bond
stretch”.
4. Compare the final damage pattern.
Objective: Study initiation and propagation of multiple
cracks in brittle material using Peridyanmics.
40
Damage pattern
top view
bottom view