[SrNaranja]

Hi @Hugo,

thank you for your response about the simulation.
I have fixed the movement of the beam with the masses, but I have still several problems with the simulation:

1. The error you told about
[ERROR] [OscillatorConstraint(OscillatingConstraint1)] applyConstraint(mparams, matrix) not implemented.
is still remaining, but the oscillator works fine apparently. If I change the solver to a CGLinear solver, the error disappears. I don’t know if I have to change the solver to an indirect solver as you told me. I wonder what are these indirect solver because I have not found any indirect solvers in the documentation and the forum.

2. I have another issue about the beam mass. When I simulate the beam with its real mass (0.1796kg) the simulation is worse.

<UniformMass template="Rigid3d,RigidMass<3u,double>" name="MassBeam" totalMass="0.17906"/>

This scene is called beam_code_13.8 in the github repository attached below.
But when I change the mass property taking into account the SOFA example

<UniformMass vertexMass="1 1 0.01 0 0 0 0.1 0 0 0 0.1" printLog="false" />

the simulation runs ok but it is not the real mass. So how we can guarantee the natural vibrations using the real mass?
This scene is called beam_code_13.9 in the github repository attached below.

3. Finally, I would like to know why if I change the simulation to real time, the beam goes away.
In the next link you have all the codes and the sphere object.
https://github.com/jruro/SOFA_Beam.git

Thank you so much for your help and I wait your answers,
Juan.

[SrNaranja]

Hi @Hugo and everyone!

That problem was solved, but I have another problem with the connection between the horizontal beam and the first part of the leg, called “Hip” in the code.

The problem is that when I put the PartialLinearMovementConstraint inside the Hip node, during the movement, the “Hip” is disconnected from the beam, and then when the movements finish, it reconnects.

The correct implementation of the movement should be the Hip rotating but connected with the beam.

The code of the scene is in the next Github link:

https://github.com/jruro/SOFA_Beam.git

The scene is called beam_code2_2.1.scn

Thank you so much for help and I wait your answers,

Juan.

[SrNaranja]

Hi Sahar,

sorry but I don’t know how to extract the properties you want. I only know how to extract positions, velocities, forces and energies using Monitor and ExtraMonitor.

Here you have the documentation.

I hope it helps.

Cheers,

Juan.

[SrNaranja]

Hola Sergio,

I think it is not possible to use the BoxRoi to attach several points in one code line.

As far as I know, there could be a solution using as many nodes as points you want to attach. For BillateralInteractionConstraint, if you want to attach three points of one body to three points of another body, you will need to create three different constraint nodes as follows:

      <Node name="ConstraintPoint_beam_1">
          <MechanicalObject name="points_beam" template="Vec3d" position="-0.021 0 0" />
          <RigidMapping index="56"/>
      </Node>

and, of course, outside the body node, you will need three BilateralInteractionConstraint instances.

I hope this will help you, let me know if you achieve or not.

Un saludo,

Juan.

[SrNaranja]

Hi @Hugo,
thank you for your response about the simulation.
I have fixed the movement of the beam with the masses, but I have still several problems with the simulation:

1. The error you told about

[ERROR] [OscillatorConstraint(OscillatingConstraint1)] applyConstraint(mparams, matrix) not implemented.

is still remaining, but the oscillator works fine apparently. If I change the solver to a CGLinear solver, the error disappears. I don’t know if I have to change the solver to an indirect solver as you told me. I wonder what are these indirect solver because I have not found any indirect solvers in the documentation and the forum.

2. I have another issue about the beam mass. When I simulate the beam with its real mass (0.1796kg) the simulation is worse.

<UniformMass template="Rigid3d,RigidMass<3u,double>", name="MassBeam", totalMass="0.17906"/>

This scene is called beam_code_13.8 in the github repository attached below.
But when I change the mass property taking into account the SOFA example

<UniformMass vertexMass="1 1 0.01 0 0 0 0.1 0 0 0 0.1" printLog="false" />

the simulation runs ok but it is not the real mass. So how we can guarantee the natural vibrations using the real mass?
This scene is called beam_code_13.9 in the github repository attached below.

3. Finally, I would like to know why if I change the simulation to real time, the beam goes away.
In the next link you have all the codes and the sphere object.
https://github.com/jruro/SOFA_Beam.git

Thank you so much for your help and I wait your answers,
Juan.

[SrNaranja]

Hi @Hugo

No problem for the delay of the reply 😉

About the points we are talking about:

1. Thank you so much, this problem is resolved.

2. The problem is when I put

<UniformMass template=”Rigid3d,RigidMass<3u,double>” name=”MassBeam” totalMass=”0.17906″/>

the behavior of the vibrations is shown in the video called beam_code_14_3__r60_0001.mp4

But when I change the mass, from a totalMass to a vertexMass:

<UniformMass vertexMass=”1 1 0.01 0 0 0 0.1 0 0 0 0.1″ printLog=”false” />

the behavoir is shown in the video called beam_code_14_3__r60_0002.mp4

This last simulation is what is expected, not the first one. I don’t know if there is a way to put the total mass of 0.1796 and then the behavior of the beam will look like the second video.

Both videos are uploaded in the Github repository attached below.

3. Thank you, I won’t use real time simulation. But I have another issue with that.
When the time step DT is set to

dt=”0.005″

the beam is bended to a maximum of 9.1cm at its furthest point from the center, as it shows the figure below.

But when I change the time step to a DT

dt=”0.0005″

the maximum bending point is at around 20cm, as it shows the figure below.

I think this is due to the linear solver, in this case I am using a direct solver. But with an indirect solver the behavior is the same. the objective is to obtain a maximum bending point of about 40cm, so I don’t know if this is possible with these linear solvers or I am trying something impossible.

In the next link you have all the scenes. Last scenes are called “beam_code_14_1” and “beam_code_14_2”
https://github.com/jruro/SOFA_Beam.git

Thank you so much for help and I wait your answers,

Juan.

[SrNaranja]

Hi @Hugo

No problem for the delay of the reply 😉

About the points we are talking about:

1. Thank you so much, this problem is resolved.

2. The problem is when I put

<UniformMass template=”Rigid3d,RigidMass<3u,double>” name=”MassBeam” totalMass=”0.17906″/>

the behavior of the vibrations is shown in the video called beam_code_14_3__r60_0001.mp4

But when I change the mass, from a totalMass to a vertexMass:

<UniformMass vertexMass=”1 1 0.01 0 0 0 0.1 0 0 0 0.1″ printLog=”false” />

the behavoir is shown in the video called beam_code_14_3__r60_0002.mp4

This last simulation is what is expected, not the first one. I don’t know if there is a way to put the total mass of 0.1796 and then the behavior of the beam will look like the second video.

Both videos are uploaded in the Github repository attached below.

3. Thank you, I won’t use real time simulation. But I have another issue with that.
When the time step DT is set to

dt=”0.005″

the beam is bended to a maximum of 9.1cm at its furthest point from the center, as it shows the figure called dt_1.png in the Github repository.

But when I change the time step to a DT

dt=”0.0005″

the maximum bending point is at around 20cm, as it shows the figure called dt_2.png in the Github repository.

I think this is due to the linear solver, in this case I am using a direct solver. But with an indirect solver the behavior is the same. the objective is to obtain a maximum bending point of about 40cm, so I don’t know if this is possible with these linear solvers or I am trying something impossible.

Here you have the link to the github repository, where you can find the latest scenes and the figures I have mentioned above.
https://github.com/jruro/SOFA_Beam.git

Thank you so much for help and I wait your answers,

Juan.

[SrNaranja]

Hi @Hugo,
thank you for your response about the simulation.
I have fixed the movement of the beam with the masses, but I have still several problems with the simulation:

1. The error you told about

[ERROR] [OscillatorConstraint(OscillatingConstraint1)] applyConstraint(mparams, matrix) not implemented.

is still remaining, but the oscillator works fine apparently. If I change the solver to a CGLinear solver, the error disappears. I don’t know if I have to change the solver to an indirect solver as you told me. I wonder what are these indirect solver because I have not found any indirect solvers in the documentation and the forum.

2. I have another issue about the beam mass. When I simulate the beam with its real mass (0.1796kg) the simulation is worse.
<UniformMass template="Rigid3d,RigidMass<3u,double>" name="MassBeam" totalMass="0.17906"/>
This scene is called beam_code_13.8 in the github repository attached below.
But when I change the mass property taking into account the SOFA example
<UniformMass vertexMass="1 1 0.01 0 0 0 0.1 0 0 0 0.1" printLog="false" />
the simulation runs ok but it is not the real mass. So how we can guarantee the natural vibrations using the real mass?
This scene is called beam_code_13.9 in the github repository attached below.

3. Finally, I would like to know why if I change the simulation to real time, the beam goes away.
In the next link you have all the codes and the sphere object.
https://github.com/jruro/SOFA_Beam.git
Thank you so much for your help and I wait your answers,
Juan.

[SrNaranja]

Hi @Hugo,

In the same link, I have attached two videos showing the corresponding simulation.

Thanks,

Juan.

[Author]

Posts