Guidance on Boundary layer meshing and Misc.

Provides a system for patient-specific cardiovascular modeling and simulation.
User avatar
David Parker
Posts: 1719
Joined: Tue Aug 23, 2005 2:43 pm

Re: Guidance on Boundary layer meshing and Misc.

Post by David Parker » Thu Apr 09, 2020 10:58 am

Hi Nishanth,

I was able to extract branches from a remeshed AAA model using VMTK. This assigns a group ID to sets of faces that (approximately) belong to a vessel branch. It is not perfect but could probably be used to assign faces to the model. See https://github.com/ktbolt/cardiovascula ... -extractor.

A better option might be to use Radius-Based Meshing. This is used for models that have vessels with a large difference in size. It first computes centerlines of the model and then uses the max inscribed sphere radii data to adaptively mesh the model.

I tried to do this for the AAA model but it is just too big (4,680,046 polygons). I remeshed to a smaller size and tried to do the radius-based meshing but it failed.

A couple of things you can try

1) Remesh the model to reduce the number of polygons. I've tried this for a couple of edge sizes but the meshing always fails.

2) Remove the long tortuous celiac sub-branch.

3) Use a better mesher. We also support the MeshSim mesher but you will need to purchase a license for it.

Cheers,
Dave

User avatar
Nishanth Surianarayanan
Posts: 39
Joined: Thu May 30, 2019 6:21 pm

Re: Guidance on Boundary layer meshing and Misc.

Post by Nishanth Surianarayanan » Fri Apr 10, 2020 8:15 pm

Hi Dave,

Thanks for your reply! I have downloaded the new version of SimVascular.

Even I have tried the radius-based meshing several times, it takes a lot of time to complete. Probably SV has difficulty extracting centrelines.

a) Is it possible to extract the centrelines using VMTK and later import it into SimVascular to perform radius-based meshing?
b) Can the extract centreline feature in SV Modeling be used for radius-based meshing?

Thanks for looking into the vmtk option. I will try that.

c) Is the regional refinement feature a functionality of Tetgen or VMTK or developed by Prof. Marsden group?

I am unable to figure out why regional refinement is causing me this issue (it doesn't matter if boundary layer is present or not). I want to use a fine mesh for my simulations as an edge size of 0.25 would be very coarse for my simulations.

d) Does the order in which you add regional refinement at different places matter? Like, regional refinement spheres should be added in an order from inlet to outlet. I am thinking it might be because of some intricacy like that.

e) What is the order in which SV performs meshing when using local, regional and boundary layer meshing? Does it generally do meshing like global mesh size first, local size second and regional refinement third and then boundary layer mesh at the end?

Thanks,
Nishanth

User avatar
David Parker
Posts: 1719
Joined: Tue Aug 23, 2005 2:43 pm

Re: Guidance on Boundary layer meshing and Misc.

Post by David Parker » Mon Apr 13, 2020 11:34 am

Hi Nishanth,

a) Is it possible to extract the centrelines using VMTK and later import it into SimVascular to perform radius-based meshing?

SV is not able to read in centerlines, always computes them.

b) Can the extract centreline feature in SV Modeling be used for radius-based meshing?

The SV Meshing tool should really use the centerlines from the Modeling tool but the centerlines computed
there are not managed well, they are not really part of a model, don't update when the model is updated, etc.

c) Is the regional refinement feature a functionality of Tetgen or VMTK or developed by Prof. Marsden group?

Regional refinement is part of the TetGen sizing function. SV sets a data array with the size value and
TetGen uses that to perform anisotropic meshing.

I am unable to figure out why regional refinement is causing me this issue (it doesn't matter if boundary layer is present or not). I want to use a fine mesh for my simulations as an edge size of 0.25 would be very coarse for my simulations.


It seems that this is a problem in TetGen. The element distortion is along the lines of regions with different refinement IDs that define the
regions to refine. See image.

d) Does the order in which you add regional refinement at different places matter? Like, regional refinement spheres should be added in an order from inlet to outlet. I am thinking it might be because of some intricacy like that.

This does not appear to be the case.

e) What is the order in which SV performs meshing when using local, regional and boundary layer meshing? Does it generally do meshing like global mesh size first, local size second and regional refinement third and then boundary layer mesh at the end?

All meshing using any local or global refinement is done within TetGen using values set in SV. The boundary layer meshing is done after TetGen
is finished, it uses VMTK.

I think the best option is to reduce the number of polygons in your model. This will allow you to adjust the mesh size to be smaller.

Cheers,
Dave


Screen Shot 2020-04-13 at 11.28.20 AM.jpg
Surface colored by refinement IDs with mesh overlayed.
Screen Shot 2020-04-13 at 11.28.20 AM.jpg (153.27 KiB) Viewed 1246 times

User avatar
Nishanth Surianarayanan
Posts: 39
Joined: Thu May 30, 2019 6:21 pm

Re: Guidance on Boundary layer meshing and Misc.

Post by Nishanth Surianarayanan » Sun Apr 26, 2020 3:00 pm

Hi Dave,

Thanks a lot for your response, I am learning a lot. I am sorry to bother you with a few additional questions:

1) Is remeshing with a larger edge size and Decimate option the same (reduces the triangle count)? By polygon count, you are referring to triangle count, right?

2) By "Skip walls" in the convert results section, what parameters are not calculated?

3) I was reading about the time integration rho infinity in the user manual. It says the numerical dissipation is reduced as the value increases towards 1.
3a) Does this parameter control the numerical dissipation resulting from the use of SUPG scheme (because of the stabilization)?
3b) I am applying Direct Numerical Simulation for transitional flows in aneurysm. Is it advisable to increase this value from 0.5? What value do you recommend utmost?
The manual says the value of 1 can be dangerous if the temporal and spatial discretizations are not adequate.

4) What is the meaning of the TimeDeriv parameter in the converted results?

5) Under Linear solver, there are several types of svLS- NS, GMRES, CG. The manual suggests GMRES for deformable simulation. Is NS type (default) best for pulsatile rigid simulations?

Thank you for your support! I really appreciate it.

Best,
Nishanth

User avatar
David Parker
Posts: 1719
Joined: Tue Aug 23, 2005 2:43 pm

Re: Guidance on Boundary layer meshing and Misc.

Post by David Parker » Mon Apr 27, 2020 12:28 pm

Hi Nishanth,

Not a bother at all!

1) Is remeshing with a larger edge size and Decimate option the same (reduces the triangle count)? By polygon count, you are referring to triangle count, right?

Decimate reduces the number of polygons used to represent a model, which is the surface used for meshing. Decimation will allow you to use
a larger edge size without creating a non-manifold edges error.

2) By "Skip walls" in the convert results section, what parameters are not calculated?

This selects computing results (e.g. pressure, flow) for only cap faces, i.e. inlet and outlet faces.

3) I was reading about the time integration rho infinity in the user manual. It says the numerical dissipation is reduced as the value increases towards 1.

a) Does this parameter control the numerical dissipation resulting from the use of SUPG scheme (because of the stabilization)?

This parameter controls high frequency damping in the transient equations. It is not related to the SUPG stabilization.

b) I am applying Direct Numerical Simulation for transitional flows in aneurysm. Is it advisable to increase this value from 0.5?
What value do you recommend utmost?

The manual says the value of 1 can be dangerous if the temporal and spatial discretization are not adequate.

It is probably best not to change the default parameters unless you are experiencing convergence problems. Like all solver parameters
you will need to experiment and see if you get converged solutions.

4) What is the meaning of the TimeDeriv parameter in the converted results?

That is the time derivative of the solution to the Navier Stokes equations, acceleration.

5) Under Linear solver, there are several types of svLS- NS, GMRES, CG. The manual suggests GMRES for deformable simulation. Is NS type (default) best for pulsatile rigid simulations?

I'm unable to find any information about the NS solver but I suppose that if it is the default solver then it is best for rigid simulations.


Cheers,
Dave

User avatar
Nishanth Surianarayanan
Posts: 39
Joined: Thu May 30, 2019 6:21 pm

Re: Guidance on Boundary layer meshing and Misc.

Post by Nishanth Surianarayanan » Thu May 07, 2020 8:30 pm

Hi Dave,

Thank you so much for your reply.

1) I was exploring the SV 1D solver using the Windows version of SimVascular, but I am unable to run the simulation. It says as shown below:

A solver input file has bee generated with warnings.

Status: OK
Mesh: num_nodes=29 num_elements=2308 num_segs=28
Log:
[generate-1d-mesh] INFO - Parse arguments ...
[generate-1d-mesh] INFO - Centerlines input file: E:/AAA lumen/Simulations1d/Try/centerlines.vtp
[generate-1d-mesh] INFO - Compute centerlines: False
[generate-1d-mesh] INFO - Density: 1.060000
[generate-1d-mesh] INFO - Element size: 0.050000
[generate-1d-mesh] INFO - Inflow input file: E:/AAA lumen/Simulations1d/Try/flow_restcond.flow
[generate-1d-mesh] INFO - Linear material parameter Er/h: 1127819.550000
[generate-1d-mesh] INFO - Linear material parameter pressure: 142214577.000000
[generate-1d-mesh] INFO - Material model: LINEAR
[generate-1d-mesh] INFO - Minimum number of finite elements per segment: 10
[generate-1d-mesh] INFO - Model name: AAA
[generate-1d-mesh] INFO - Number of time steps: 5000
[generate-1d-mesh] INFO - Outflow bc file: E:/AAA lumen/Simulations1d/Try/resistance.dat
[generate-1d-mesh] INFO - Outflow bc type: resistance
[generate-1d-mesh] INFO - Outlet face names file: 'E:/AAA lumen/Simulations1d/Try/outlet_face_names.dat'.
[generate-1d-mesh] INFO - Save data frequency: 10
[generate-1d-mesh] INFO - Solver output file: solver.in
[generate-1d-mesh] INFO - Simulation time step: 0.001000
[generate-1d-mesh] INFO - Uniform boundary conditions: False
[generate-1d-mesh] INFO - Units: cm
[generate-1d-mesh] INFO - Viscosity: 0.040000
[generate-1d-mesh] INFO - Wall properties are uniform.
[generate-1d-mesh] INFO - Write mesh file: False
[generate-1d-mesh] INFO - Write solver file: True
[generate-1d-mesh] INFO - Read centerlines from the file: E:/AAA lumen/Simulations1d/Try/centerlines.vtp
[generate-1d-mesh] INFO - Number of points: 12858
[generate-1d-mesh] INFO - Number of cells: 166
[generate-1d-mesh] INFO - Number of arrays: 4
[generate-1d-mesh] INFO - Generate the 1D mesh ...
[generate-1d-mesh] INFO - Read model outlet face names from: E:/AAA lumen/Simulations1d/Try/outlet_face_names.dat
[generate-1d-mesh] INFO - Number of model outlet faces names: 16
[generate-1d-mesh] INFO - Outlet face names: ['right_int_iliac', 'left_ext_iliac', 'left_int_iliac', 'right_ext_iliac', 'right_renal', 'IMA', 'SMA1', 'SMA2', 'hepatic', 'splenic', 'left_renal', 'L2_lumbar', 'L4_left_lumbar', 'L4_right_lumbar', 'L3_right_lumbar', 'L3_left_lumbar']
[generate-1d-mesh] INFO - Number of cells: 166
[generate-1d-mesh] INFO - Number of paths: 16
[generate-1d-mesh] INFO - Number of groups: 40
[generate-1d-mesh] INFO - Set variable outflow boundary conditions ...
[generate-1d-mesh] INFO - Read resistance BCs ...
[generate-1d-mesh] INFO - Face right_int_iliac value 68205.7537509692
[generate-1d-mesh] INFO - Face left_ext_iliac value 29148.0905808365
[generate-1d-mesh] INFO - Face left_int_iliac value 68161.0761844659
[generate-1d-mesh] INFO - Face right_ext_iliac value 29139.6293943678
[generate-1d-mesh] INFO - Face right_renal value 18630.3534055032
[generate-1d-mesh] INFO - Face IMA value 65951.5060636744
[generate-1d-mesh] INFO - Face SMA1 value 36067.0933876420
[generate-1d-mesh] INFO - Face SMA2 value 36794.8214039061
[generate-1d-mesh] INFO - Face hepatic value 25894.2832206710
[generate-1d-mesh] INFO - Face splenic value 25897.1125007651
[generate-1d-mesh] INFO - Face left_renal value 18606.5870885130
[generate-1d-mesh] INFO - Face L2_lumbar value 694159.453302961
[generate-1d-mesh] INFO - Face L4_left_lumbar value 693144.435418359
[generate-1d-mesh] INFO - Face L4_right_lumbar value 688502.218994594
[generate-1d-mesh] INFO - Face L3_right_lumbar value 695064.190289997
[generate-1d-mesh] INFO - Face L3_left_lumbar value 677621.346886912
[generate-1d-mesh] INFO - Read inflow BC ...
[generate-1d-mesh] WARNING - There are more than 2 child segments for groupid 2
[generate-1d-mesh] WARNING - There are more than 2 child segments for groupid 13
[generate-1d-mesh] WARNING - Ain < Aout for group id 3, face name 'right_int_iliac'
[generate-1d-mesh] WARNING - This will cause problems for the 1D solver so setting Ain = Aout
[generate-1d-mesh] WARNING - Ain < Aout for group id 35, face name 'L4_left_lumbar'
[generate-1d-mesh] WARNING - This will cause problems for the 1D solver so setting Ain = Aout
[generate-1d-mesh] INFO - Write solver file.
[generate-1d-mesh] INFO - Write solver segment section ...
[generate-1d-mesh] INFO - Uniform BC: False
[generate-1d-mesh] INFO - Outflow BC: resistance
[generate-1d-mesh] INFO - Generated 28 segments, 29 nodes and 2308 elements.
[generate-1d-mesh] INFO - Parse arguments ...
[generate-1d-mesh] INFO - Centerlines input file: E:/AAA lumen/Simulations1d/Try/centerlines.vtp
[generate-1d-mesh] INFO - Compute centerlines: False
[generate-1d-mesh] INFO - Density: 1.060000
[generate-1d-mesh] INFO - Element size: 0.050000
[generate-1d-mesh] INFO - Inflow input file: E:/AAA lumen/Simulations1d/Try/flow_restcond.flow
[generate-1d-mesh] INFO - Linear material parameter Er/h: 1127819.550000
[generate-1d-mesh] INFO - Linear material parameter pressure: 142214577.000000
[generate-1d-mesh] INFO - Material model: LINEAR
[generate-1d-mesh] INFO - Minimum number of finite elements per segment: 10
[generate-1d-mesh] INFO - Model name: AAA
[generate-1d-mesh] INFO - Number of time steps: 5000
[generate-1d-mesh] INFO - Outflow bc file: E:/AAA lumen/Simulations1d/Try/resistance.dat
[generate-1d-mesh] INFO - Outflow bc type: resistance
[generate-1d-mesh] INFO - Outlet face names file: 'E:/AAA lumen/Simulations1d/Try/outlet_face_names.dat'.
[generate-1d-mesh] INFO - Save data frequency: 10
[generate-1d-mesh] INFO - Solver output file: solver.in
[generate-1d-mesh] INFO - Simulation time step: 0.001000
[generate-1d-mesh] INFO - Uniform boundary conditions: False
[generate-1d-mesh] INFO - Units: cm
[generate-1d-mesh] INFO - Viscosity: 0.040000
[generate-1d-mesh] INFO - Wall properties are uniform.
[generate-1d-mesh] INFO - Write mesh file: False
[generate-1d-mesh] INFO - Write solver file: True
[generate-1d-mesh] INFO - Read centerlines from the file: E:/AAA lumen/Simulations1d/Try/centerlines.vtp
[generate-1d-mesh] INFO - Number of points: 12858
[generate-1d-mesh] INFO - Number of cells: 166
[generate-1d-mesh] INFO - Number of arrays: 4
[generate-1d-mesh] INFO - Generate the 1D mesh ...
[generate-1d-mesh] INFO - Read model outlet face names from: E:/AAA lumen/Simulations1d/Try/outlet_face_names.dat
[generate-1d-mesh] INFO - Number of model outlet faces names: 16
[generate-1d-mesh] INFO - Outlet face names: ['right_int_iliac', 'left_ext_iliac', 'left_int_iliac', 'right_ext_iliac', 'right_renal', 'IMA', 'SMA1', 'SMA2', 'hepatic', 'splenic', 'left_renal', 'L2_lumbar', 'L4_left_lumbar', 'L4_right_lumbar', 'L3_right_lumbar', 'L3_left_lumbar']
[generate-1d-mesh] INFO - Number of cells: 166
[generate-1d-mesh] INFO - Number of paths: 16
[generate-1d-mesh] INFO - Number of groups: 40
[generate-1d-mesh] INFO - Set variable outflow boundary conditions ...
[generate-1d-mesh] INFO - Read resistance BCs ...
[generate-1d-mesh] INFO - Face right_int_iliac value 68205.7537509692
[generate-1d-mesh] INFO - Face left_ext_iliac value 29148.0905808365
[generate-1d-mesh] INFO - Face left_int_iliac value 68161.0761844659
[generate-1d-mesh] INFO - Face right_ext_iliac value 29139.6293943678
[generate-1d-mesh] INFO - Face right_renal value 18630.3534055032
[generate-1d-mesh] INFO - Face IMA value 65951.5060636744
[generate-1d-mesh] INFO - Face SMA1 value 36067.0933876420
[generate-1d-mesh] INFO - Face SMA2 value 36794.8214039061
[generate-1d-mesh] INFO - Face hepatic value 25894.2832206710
[generate-1d-mesh] INFO - Face splenic value 25897.1125007651
[generate-1d-mesh] INFO - Face left_renal value 18606.5870885130
[generate-1d-mesh] INFO - Face L2_lumbar value 694159.453302961
[generate-1d-mesh] INFO - Face L4_left_lumbar value 693144.435418359
[generate-1d-mesh] INFO - Face L4_right_lumbar value 688502.218994594
[generate-1d-mesh] INFO - Face L3_right_lumbar value 695064.190289997
[generate-1d-mesh] INFO - Face L3_left_lumbar value 677621.346886912
[generate-1d-mesh] INFO - Read inflow BC ...
[generate-1d-mesh] WARNING - There are more than 2 child segments for groupid 2
[generate-1d-mesh] WARNING - There are more than 2 child segments for groupid 13
[generate-1d-mesh] WARNING - Ain < Aout for group id 3, face name 'right_int_iliac'
[generate-1d-mesh] WARNING - This will cause problems for the 1D solver so setting Ain = Aout
[generate-1d-mesh] WARNING - Ain < Aout for group id 35, face name 'L4_left_lumbar'
[generate-1d-mesh] WARNING - This will cause problems for the 1D solver so setting Ain = Aout
[generate-1d-mesh] INFO - Write solver file.
[generate-1d-mesh] INFO - Write solver segment section ...
[generate-1d-mesh] INFO - Uniform BC: False
[generate-1d-mesh] INFO - Outflow BC: resistance
[generate-1d-mesh] INFO - Generated 28 segments, 29 nodes and 2308 elements.
-1d-mesh] INFO - Parse arguments ...
[generate-1d-mesh] INFO - Centerlines input file: E:/AAA lumen/Simulations1d/Try/centerlines.vtp
[generate-1d-mesh] INFO - Compute centerlines: False
[generate-1d-mesh] INFO - Density: 1.060000
[generate-1d-mesh] INFO - Element size: 0.050000
[generate-1d-mesh] INFO - Inflow input file: E:/AAA lumen/Simulations1d/Try/flow_restcond.flow
[generate-1d-mesh] INFO - Linear material parameter Er/h: 1127819.550000
[generate-1d-mesh] INFO - Linear material parameter pressure: 142214577.000000
[generate-1d-mesh] INFO - Material model: LINEAR
[generate-1d-mesh] INFO - Minimum number of finite elements per segment: 10
[generate-1d-mesh] INFO - Model name: AAA
[generate-1d-mesh] INFO - Number of time steps: 5000
[generate-1d-mesh] INFO - Outflow bc file: E:/AAA lumen/Simulations1d/Try/resistance.dat
[generate-1d-mesh] INFO - Outflow bc type: resistance
[generate-1d-mesh] INFO - Outlet face names file: 'E:/AAA lumen/Simulations1d/Try/outlet_face_names.dat'.
[generate-1d-mesh] INFO - Save data frequency: 10
[generate-1d-mesh] INFO - Solver output file: solver.in
[generate-1d-mesh] INFO - Simulation time step: 0.001000
[generate-1d-mesh] INFO - Uniform boundary conditions: False
[generate-1d-mesh] INFO - Units: cm
[generate-1d-mesh] INFO - Viscosity: 0.040000
[generate-1d-mesh] INFO - Wall properties are uniform.
[generate-1d-mesh] INFO - Write mesh file: False
[generate-1d-mesh] INFO - Write solver file: True
[generate-1d-mesh] INFO - Read centerlines from the file: E:/AAA lumen/Simulations1d/Try/centerlines.vtp
[generate-1d-mesh] INFO - Number of points: 12858
[generate-1d-mesh] INFO - Number of cells: 166
[generate-1d-mesh] INFO - Number of arrays: 4
[generate-1d-mesh] INFO - Generate the 1D mesh ...
[generate-1d-mesh] INFO - Read model outlet face names from: E:/AAA lumen/Simulations1d/Try/outlet_face_names.dat
[generate-1d-mesh] INFO - Number of model outlet faces names: 16
[generate-1d-mesh] INFO - Outlet face names: ['right_int_iliac', 'left_ext_iliac', 'left_int_iliac', 'right_ext_iliac', 'right_renal', 'IMA', 'SMA1', 'SMA2', 'hepatic', 'splenic', 'left_renal', 'L2_lumbar', 'L4_left_lumbar', 'L4_right_lumbar', 'L3_right_lumbar', 'L3_left_lumbar']
[generate-1d-mesh] INFO - Number of cells: 166
[generate-1d-mesh] INFO - Number of paths: 16
[generate-1d-mesh] INFO - Number of groups: 40
[generate-1d-mesh] INFO - Set variable outflow boundary conditions ...
[generate-1d-mesh] INFO - Read resistance BCs ...
[generate-1d-mesh] INFO - Face right_int_iliac value 68205.7537509692
[generate-1d-mesh] INFO - Face left_ext_iliac value 29148.0905808365
[generate-1d-mesh] INFO - Face left_int_iliac value 68161.0761844659
[generate-1d-mesh] INFO - Face right_ext_iliac value 29139.6293943678
[generate-1d-mesh] INFO - Face right_renal value 18630.3534055032
[generate-1d-mesh] INFO - Face IMA value 65951.5060636744
[generate-1d-mesh] INFO - Face SMA1 value 36067.0933876420
[generate-1d-mesh] INFO - Face SMA2 value 36794.8214039061
[generate-1d-mesh] INFO - Face hepatic value 25894.2832206710
[generate-1d-mesh] INFO - Face splenic value 25897.1125007651
[generate-1d-mesh] INFO - Face left_renal value 18606.5870885130
[generate-1d-mesh] INFO - Face L2_lumbar value 694159.453302961
[generate-1d-mesh] INFO - Face L4_left_lumbar value 693144.435418359
[generate-1d-mesh] INFO - Face L4_right_lumbar value 688502.218994594
[generate-1d-mesh] INFO - Face L3_right_lumbar value 695064.190289997
[generate-1d-mesh] INFO - Face L3_left_lumbar value 677621.346886912
[generate-1d-mesh] INFO - Read inflow BC ...
[generate-1d-mesh] WARNING - There are more than 2 child segments for groupid 2
[generate-1d-mesh] WARNING - There are more than 2 child segments for groupid 13
[generate-1d-mesh] WARNING - Ain < Aout for group id 3, face name 'right_int_iliac'
[generate-1d-mesh] WARNING - This will cause problems for the 1D solver so setting Ain = Aout
[generate-1d-mesh] WARNING - Ain < Aout for group id 35, face name 'L4_left_lumbar'
[generate-1d-mesh] WARNING - This will cause problems for the 1D solver so setting Ain = Aout
[generate-1d-mesh] INFO - Write solver file.
[generate-1d-mesh] INFO - Write solver segment section ...
[generate-1d-mesh] INFO - Uniform BC: False
[generate-1d-mesh] INFO - Outflow BC: resistance
[generate-1d-mesh] INFO - Generated 28 segments, 29 nodes and 2308 elements.
e-1d-mesh] INFO - Parse arguments ...
[generate-1d-mesh] INFO - Centerlines input file: E:/AAA lumen/Simulations1d/Try/centerlines.vtp
[generate-1d-mesh] INFO - Compute centerlines: False
[generate-1d-mesh] INFO - Density: 1.060000
[generate-1d-mesh] INFO - Element size: 0.050000
[generate-1d-mesh] INFO - Inflow input file: E:/AAA lumen/Simulations1d/Try/flow_restcond.flow
[generate-1d-mesh] INFO - Linear material parameter Er/h: 1127819.550000
[generate-1d-mesh] INFO - Linear material parameter pressure: 142214577.000000
[generate-1d-mesh] INFO - Material model: LINEAR
[generate-1d-mesh] INFO - Minimum number of finite elements per segment: 10
[generate-1d-mesh] INFO - Model name: AAA
[generate-1d-mesh] INFO - Number of time steps: 5000
[generate-1d-mesh] INFO - Outflow bc file: E:/AAA lumen/Simulations1d/Try/resistance.dat
[generate-1d-mesh] INFO - Outflow bc type: resistance
[generate-1d-mesh] INFO - Outlet face names file: 'E:/AAA lumen/Simulations1d/Try/outlet_face_names.dat'.
[generate-1d-mesh] INFO - Save data frequency: 10
[generate-1d-mesh] INFO - Solver output file: solver.in
[generate-1d-mesh] INFO - Simulation time step: 0.001000
[generate-1d-mesh] INFO - Uniform boundary conditions: False
[generate-1d-mesh] INFO - Units: cm
[generate-1d-mesh] INFO - Viscosity: 0.040000
[generate-1d-mesh] INFO - Wall properties are uniform.
[generate-1d-mesh] INFO - Write mesh file: False
[generate-1d-mesh] INFO - Write solver file: True
[generate-1d-mesh] INFO - Read centerlines from the file: E:/AAA lumen/Simulations1d/Try/centerlines.vtp
[generate-1d-mesh] INFO - Number of points: 12858
[generate-1d-mesh] INFO - Number of cells: 166
[generate-1d-mesh] INFO - Number of arrays: 4
[generate-1d-mesh] INFO - Generate the 1D mesh ...
[generate-1d-mesh] INFO - Read model outlet face names from: E:/AAA lumen/Simulations1d/Try/outlet_face_names.dat
[generate-1d-mesh] INFO - Number of model outlet faces names: 16
[generate-1d-mesh] INFO - Outlet face names: ['right_int_iliac', 'left_ext_iliac', 'left_int_iliac', 'right_ext_iliac', 'right_renal', 'IMA', 'SMA1', 'SMA2', 'hepatic', 'splenic', 'left_renal', 'L2_lumbar', 'L4_left_lumbar', 'L4_right_lumbar', 'L3_right_lumbar', 'L3_left_lumbar']
[generate-1d-mesh] INFO - Number of cells: 166
[generate-1d-mesh] INFO - Number of paths: 16
[generate-1d-mesh] INFO - Number of groups: 40
[generate-1d-mesh] INFO - Set variable outflow boundary conditions ...
[generate-1d-mesh] INFO - Read resistance BCs ...
[generate-1d-mesh] INFO - Face right_int_iliac value 68205.7537509692
[generate-1d-mesh] INFO - Face left_ext_iliac value 29148.0905808365
[generate-1d-mesh] INFO - Face left_int_iliac value 68161.0761844659
[generate-1d-mesh] INFO - Face right_ext_iliac value 29139.6293943678
[generate-1d-mesh] INFO - Face right_renal value 18630.3534055032
[generate-1d-mesh] INFO - Face IMA value 65951.5060636744
[generate-1d-mesh] INFO - Face SMA1 value 36067.0933876420
[generate-1d-mesh] INFO - Face SMA2 value 36794.8214039061
[generate-1d-mesh] INFO - Face hepatic value 25894.2832206710
[generate-1d-mesh] INFO - Face splenic value 25897.1125007651
[generate-1d-mesh] INFO - Face left_renal value 18606.5870885130
[generate-1d-mesh] INFO - Face L2_lumbar value 694159.453302961
[generate-1d-mesh] INFO - Face L4_left_lumbar value 693144.435418359
[generate-1d-mesh] INFO - Face L4_right_lumbar value 688502.218994594
[generate-1d-mesh] INFO - Face L3_right_lumbar value 695064.190289997
[generate-1d-mesh] INFO - Face L3_left_lumbar value 677621.346886912
[generate-1d-mesh] INFO - Read inflow BC ...
[generate-1d-mesh] WARNING - There are more than 2 child segments for groupid 2
[generate-1d-mesh] WARNING - There are more than 2 child segments for groupid 13
[generate-1d-mesh] WARNING - Ain < Aout for group id 3, face name 'right_int_iliac'
[generate-1d-mesh] WARNING - This will cause problems for the 1D solver so setting Ain = Aout
[generate-1d-mesh] WARNING - Ain < Aout for group id 35, face name 'L4_left_lumbar'
[generate-1d-mesh] WARNING - This will cause problems for the 1D solver so setting Ain = Aout
[generate-1d-mesh] INFO - Write solver file.
[generate-1d-mesh] INFO - Write solver segment section ...
[generate-1d-mesh] INFO - Uniform BC: False
[generate-1d-mesh] INFO - Outflow BC: resistance
[generate-1d-mesh] INFO - Generated 28 segments, 29 nodes and 2308 elements.
e-1d-mesh] INFO - Parse arguments ...
[generate-1d-mesh] INFO - Centerlines input file: E:/AAA lumen/Simulations1d/Try/centerlines.vtp
[generate-1d-mesh] INFO - Compute centerlines: False
[generate-1d-mesh] INFO - Density: 1.060000
[generate-1d-mesh] INFO - Element size: 0.050000
[generate-1d-mesh] INFO - Inflow input file: E:/AAA lumen/Simulations1d/Try/flow_restcond.flow
[generate-1d-mesh] INFO - Linear material parameter Er/h: 1127819.550000
[generate-1d-mesh] INFO - Linear material parameter pressure: 142214577.000000
[generate-1d-mesh] INFO - Material model: LINEAR
[generate-1d-mesh] INFO - Minimum number of finite elements per segment: 10
[generate-1d-mesh] INFO - Model name: AAA
[generate-1d-mesh] INFO - Number of time steps: 5000
[generate-1d-mesh] INFO - Outflow bc file: E:/AAA lumen/Simulations1d/Try/resistance.dat
[generate-1d-mesh] INFO - Outflow bc type: resistance
[generate-1d-mesh] INFO - Outlet face names file: 'E:/AAA lumen/Simulations1d/Try/outlet_face_names.dat'.
[generate-1d-mesh] INFO - Save data frequency: 10
[generate-1d-mesh] INFO - Solver output file: solver.in
[generate-1d-mesh] INFO - Simulation time step: 0.001000
[generate-1d-mesh] INFO - Uniform boundary conditions: False
[generate-1d-mesh] INFO - Units: cm
[generate-1d-mesh] INFO - Viscosity: 0.040000
[generate-1d-mesh] INFO - Wall properties are uniform.
[generate-1d-mesh] INFO - Write mesh file: False
[generate-1d-mesh] INFO - Write solver file: True
[generate-1d-mesh] INFO - Read centerlines from the file: E:/AAA lumen/Simulations1d/Try/centerlines.vtp
[generate-1d-mesh] INFO - Number of points: 12858
[generate-1d-mesh] INFO - Number of cells: 166
[generate-1d-mesh] INFO - Number of arrays: 4
[generate-1d-mesh] INFO - Generate the 1D mesh ...
[generate-1d-mesh] INFO - Read model outlet face names from: E:/AAA lumen/Simulations1d/Try/outlet_face_names.dat
[generate-1d-mesh] INFO - Number of model outlet faces names: 16
[generate-1d-mesh] INFO - Outlet face names: ['right_int_iliac', 'left_ext_iliac', 'left_int_iliac', 'right_ext_iliac', 'right_renal', 'IMA', 'SMA1', 'SMA2', 'hepatic', 'splenic', 'left_renal', 'L2_lumbar', 'L4_left_lumbar', 'L4_right_lumbar', 'L3_right_lumbar', 'L3_left_lumbar']
[generate-1d-mesh] INFO - Number of cells: 166
[generate-1d-mesh] INFO - Number of paths: 16
[generate-1d-mesh] INFO - Number of groups: 40
[generate-1d-mesh] INFO - Set variable outflow boundary conditions ...
[generate-1d-mesh] INFO - Read resistance BCs ...
[generate-1d-mesh] INFO - Face right_int_iliac value 68205.7537509692
[generate-1d-mesh] INFO - Face left_ext_iliac value 29148.0905808365
[generate-1d-mesh] INFO - Face left_int_iliac value 68161.0761844659
[generate-1d-mesh] INFO - Face right_ext_iliac value 29139.6293943678
[generate-1d-mesh] INFO - Face right_renal value 18630.3534055032
[generate-1d-mesh] INFO - Face IMA value 65951.5060636744
[generate-1d-mesh] INFO - Face SMA1 value 36067.0933876420
[generate-1d-mesh] INFO - Face SMA2 value 36794.8214039061
[generate-1d-mesh] INFO - Face hepatic value 25894.2832206710
[generate-1d-mesh] INFO - Face splenic value 25897.1125007651
[generate-1d-mesh] INFO - Face left_renal value 18606.5870885130
[generate-1d-mesh] INFO - Face L2_lumbar value 694159.453302961
[generate-1d-mesh] INFO - Face L4_left_lumbar value 693144.435418359
[generate-1d-mesh] INFO - Face L4_right_lumbar value 688502.218994594
[generate-1d-mesh] INFO - Face L3_right_lumbar value 695064.190289997
[generate-1d-mesh] INFO - Face L3_left_lumbar value 677621.346886912
[generate-1d-mesh] INFO - Read inflow BC ...
[generate-1d-mesh] WARNING - There are more than 2 child segments for groupid 2
[generate-1d-mesh] WARNING - There are more than 2 child segments for groupid 13
[generate-1d-mesh] WARNING - Ain < Aout for group id 3, face name 'right_int_iliac'
[generate-1d-mesh] WARNING - This will cause problems for the 1D solver so setting Ain = Aout
[generate-1d-mesh] WARNING - Ain < Aout for group id 35, face name 'L4_left_lumbar'
[generate-1d-mesh] WARNING - This will cause problems for the 1D solver so setting Ain = Aout

[generate-1d-mesh] INFO - Write solver file.
[generate-1d-mesh] INFO - Write solver segment section ...
[generate-1d-mesh] INFO - Uniform BC: False
[generate-1d-mesh] INFO - Outflow BC: resistance
[generate-1d-mesh] INFO - Generated 28 segments, 29 nodes and 2308 elements.
SV 1D doubt.PNG
SV 1D doubt.PNG (40.59 KiB) Viewed 1191 times
SV 1D doubt1.PNG
SV 1D doubt1.PNG (15.87 KiB) Viewed 1191 times
I assume the problem is because the solver path was not specified in my Preferences as shown below:
SV 1D doubt2.PNG
SV 1D doubt2.PNG (22.76 KiB) Viewed 1191 times
What is the path for SV 1D solver? The windows version of SimVascular comes with the 1D solver, right? I am unable to find the path of the solver files.

2) I am facing some warnings as shown in bold letters above. Is this an issue?

3) Do you have any suggestions on how to specify the material properties for abdominal aorta?
If I use the linear material model, I need to specify Eh/r. Since the geometry has vessels of varying radius, what value of r do I assume when calculating Eh/r?

4) In 3D flow simulations, what is the units of the non-linear residual displayed in column 3 when running simulations?

Thanks again for your patience.

Best,
Nishanth

User avatar
David Parker
Posts: 1719
Joined: Tue Aug 23, 2005 2:43 pm

Re: Guidance on Boundary layer meshing and Misc.

Post by David Parker » Fri May 08, 2020 3:55 pm

Hi Nishanth,

1) I was exploring the SV 1D solver using the Windows version of SimVascular, but I am unable to run the simulation. It says as shown below:

The 1D solver executable is located in the C:\Program Files\SimVascular\SimVascular\2020-04-06\ directory
so there should be no problem finding it.

The exit code -1073740791 (0xc0000409) means that the 1D solver crashed. This was probably caused by the
warnings you saw.

2) I am facing some warnings as shown in bold letters above. Is this an issue?

The WARNING - There are more than 2 child segments for groupid message seems to mean that there are
three vessels segments meeting at a node, something must be wrong with the centerlines computed for your
model.

If you can put your project someplace I can download it I'll have a look.

3) Do you have any suggestions on how to specify the material properties for abdominal aorta?
If I use the linear material model, I need to specify Eh/r. Since the geometry has vessels of varying radius, what value of r do I assume when calculating Eh/r?

The Eh/r is symbolic for Eh/r0(x), where x varies along the vessel centerline. r0(x) is determined from the
segment areas.

4) In 3D flow simulations, what is the units of the non-linear residual displayed in column 3 when running simulations?

The non-linear residual represents the error in the velocity and pressure fields.


Cheers,
Dave

User avatar
Nishanth Surianarayanan
Posts: 39
Joined: Thu May 30, 2019 6:21 pm

Re: Guidance on Boundary layer meshing and Misc.

Post by Nishanth Surianarayanan » Mon May 11, 2020 11:44 am

Hi Dave,

Thanks for your reply and help!
Here is the link to my project:
https://drive.google.com/drive/folders/ ... sp=sharing

1) Is it necessary to manually link the path to the 1D solver (since I don't see the path added in the Preferences settings)? Or since the windows version comes along with the 1D solver, I don't have to do it?

2) As per your explanation, different segments will have different Eh/r values. But in linear material property, I can only give a single value for Eh/r. Which segment area should I consider (since different segments have different areas)?

Thanks,
Nishanth

User avatar
David Parker
Posts: 1719
Joined: Tue Aug 23, 2005 2:43 pm

Re: Guidance on Boundary layer meshing and Misc.

Post by David Parker » Wed May 13, 2020 3:28 pm

Hi Nishanth,

With the help of Marin Pfaller I've identified the main reason that your 1D simulations are failing; the flow values in your flow_restcond.flow file are negative, the values should be positive (unlike the convention used in svsolver). Martin was able to run your job fine although he thought your Eh/r values to be quite large.

There is currently no method in SV to set the material properties for different segments. You will need to edit the solver.in file and define material properties for each segment. For eaxmple, you can set a different material MAT2 for a segment like this

Code: Select all

SEGMENT Group39_Seg27 27 2.6666649725682534 53 7 28 0.045171481779235925 0.01092590748219857 0.0 MAT2 NONE 0.0 0 0 RESISTANCE RESISTANCE_15

MATERIAL MAT2 LINEAR 1.06 0.04 14577.0 1.0 1127819.55 0.0 0.0
Cheers,
Dave

User avatar
Nishanth Surianarayanan
Posts: 39
Joined: Thu May 30, 2019 6:21 pm

Re: Guidance on Boundary layer meshing and Misc.

Post by Nishanth Surianarayanan » Thu May 14, 2020 2:14 am

Hi Dave,

Thank you for your kind help. I am able to run the simulation. I had a few additional questions:
1) After converting results, I see the results such as pressures and flows are given for different segment names. But how do I identify which segment corresponds to which segment_name?
2) To specify the start and stop time when converting results, what do you mean by time here? By time, does it refer to the time given by time=(no. of timesteps)* (time step size).
So if I am running my simulation for 5 seconds= (5000 timesteps)*(0.001 step size), if I specify start as 4 and stop as 5, SV1d says 'time out of range'.
3) How many iterations do you recommend me to run for 1D simulations?
4) Why is it rigid wall 1D simulation option not available?
5) What is the order of Eh/r value, generally? Will material model affect the results significantly?

Best,
Nishanth

POST REPLY