Multi resolution in SPH

pohan

Hello,

I am trying to reduce the computation time of SPH simulation by multi resolution in which there are two regions fine and coarse ones.

The diameter of SPH particle in the coarse region (r1) is double than the diameter of the fine one (r2). And the distance between the two regions is r1+r2

As far as I know, in Radioss the contact between the two SPH regions is automatic. However, in my model the connect between them is not smooth.

If anyone has any ideal about this problem?

Thank you in advance.

 

Polyvios Romanidis

Hi pohan,

What you propose is not an effective way to use SPH approach, because of the issues appear like the one you referred to.  So, generally do not use different particle diameters in SPH.

There is a situation were smoother change in particle diameter could work, but by changing the diameter by a factor of 2 causes problems.

In order to run your simulation faster I can propose you to use the SOL2SPH methodology.  In this methodology the solid elements turn to SPH particles when and where is necessary only.

You can find some exhibition of this method and the way to set up a problem like this in the Knowledge Base article contained in the following link.

https://community.altair.com/community/en/5-56x45-ss109-impact-on-an-armox-500t-plate?sys_kb_id=bd0e9a3adb6e3554cfd5f6a4e296194d&id=kb_article_view&sysparm_rank=1&sysparm_tsqueryId=c611e4041b6eced8c4dfdbd9dc4bcbcb

Polyvios

pohan

Polyvios Romanidis said:

Hi pohan,

What you propose is not an effective way to use SPH approach, because of the issues appear like the one you referred to.  So, generally do not use different particle diameters in SPH.

There is a situation were smoother change in particle diameter could work, but by changing the diameter by a factor of 2 causes problems.

In order to run your simulation faster I can propose you to use the SOL2SPH methodology.  In this methodology the solid elements turn to SPH particles when and where is necessary only.

You can find some exhibition of this method and the way to set up a problem like this in the Knowledge Base article contained in the following link.

https://community.altair.com/community/en/5-56x45-ss109-impact-on-an-armox-500t-plate?sys_kb_id=bd0e9a3adb6e3554cfd5f6a4e296194d&id=kb_article_view&sysparm_rank=1&sysparm_tsqueryId=c611e4041b6eced8c4dfdbd9dc4bcbcb

Polyvios

Hello. Thank you for your answer. I will try the Sol2solid later. You has mentioned that There is a situation were smoother change in particle diameter could work, but by changing the diameter by a factor of 2 causes problems.

Could you please give me more details about this situation?

Best regards

Polyvios Romanidis

pohan said:

Hello. Thank you for your answer. I will try the Sol2solid later. You has mentioned that There is a situation were smoother change in particle diameter could work, but by changing the diameter by a factor of 2 causes problems.

Could you please give me more details about this situation?

Best regards

Hi pohan,

Maybe if you try to reduce the diameter very smoothly you can make this model work.  But to do that you should probably use command line orders to create a model with very smooth decrease.  After all I am not sure if indeed this model will work.  What I am saying is that potentially can.

From my point of view it is preferable to start with SOL2SPH and if the simulation time is still that big to attempt change the diameter smoothly with command orders.  I am saying that because it can be very challenging and time-consuming to create a model like that in pre-processors.

Polyvios

pohan

Polyvios Romanidis said:

Hi pohan,

Maybe if you try to reduce the diameter very smoothly you can make this model work.  But to do that you should probably use command line orders to create a model with very smooth decrease.  After all I am not sure if indeed this model will work.  What I am saying is that potentially can.

From my point of view it is preferable to start with SOL2SPH and if the simulation time is still that big to attempt change the diameter smoothly with command orders.  I am saying that because it can be very challenging and time-consuming to create a model like that in pre-processors.

Polyvios

Hello,

In my opinion Sol2SPH is a good solution in the case of solid simulation. However, in my case, I want to model water, so it is not easy because of failure criterial of water.

Best regards 

Polyvios Romanidis

pohan said:

Hello,

In my opinion Sol2SPH is a good solution in the case of solid simulation. However, in my case, I want to model water, so it is not easy because of failure criterial of water.

Best regards 

Hi pohan,

In this case you can use the attached files.

In order to create the smooth diameter increase you should first create an increasing size mesh.  Then you can use any of the following macro .tcl I include in this message.

The one file creates particles in the same coordinates that nodes exist in solid mesh and the other at the center of the solid elements.  The only difference between them is the definition of connections or BCS with solid elements, where particles placement can be important.

Hope this helps.

Polyvios

pohan

Polyvios Romanidis said:

Hi pohan,

In this case you can use the attached files.

In order to create the smooth diameter increase you should first create an increasing size mesh.  Then you can use any of the following macro .tcl I include in this message.

The one file creates particles in the same coordinates that nodes exist in solid mesh and the other at the center of the solid elements.  The only difference between them is the definition of connections or BCS with solid elements, where particles placement can be important.

Hope this helps.

Polyvios

Thank you very much, I will try and fed back soon!