Two-Way Coupled FSI Problem

alijamali · November 17, 2016, 15:38

I am trying to get a sphere to deform and pass through a choke-like flow path. I am very new to both ANSYS and FSI system coupling. I have read/watched many tutorials to make this model; however, I think I have either messed up or not quite understood parts of the physics of the problem and that's why it does not converge. I have attached 4 pictures of how the sphere moves upward until the simulation stops at t = 0.4 sec with the following error:

Fluent encountered fatal error after sync pint solve: update-dynamic mesh failed. Negative cell volume detected.

Ideally, and based on the physics of my problem, I would like the sphere to either stop or deform and pass through the choke. What am I doing wrong here?

Here is a link to the simulation model

The ultimate goal is to simulate something like this video except that in my case it's a solid moving inside a fluid medium.

alijamali · November 20, 2016, 11:23

Anybody? Please.

fresty · November 22, 2016, 06:08

Negative volume error represents moving mesh issues and resulting unreal mesh elements due to self-intersection.. Have you tried simulating with a model where radius of the ball/ sphere is shorter than the channel? I would think of that working pretty fine without any significant changes in your current model mesh quality..
From top of my head, actually the only way you could simulate this with the moving/ dynamic mesh method is if the circular body is allowed to pass through the channel, the gap however could be extremely minor... on the other hand, this simulation could be divided into 2 parts: (i) ball upward movement by virtue of flow till the very last element possible (ii) ball stuck in the channel (this should be fairly simple & straightforward)..

alijamali · November 22, 2016, 08:41

Thank you fresty for your response. I would seriously consider breaking the problem down into two parts as you suggested; however, I preferred if I could see how the sphere interacts with the wall and with the fluid, deforms, and passes through the flow restriction. In fact, the whole purpose of this simulation is to observe pass and stop modes as wells as pressure profile through the throat.

fresty · November 23, 2016, 00:55

I initially assumed that the sphere was a rigid body (without watching the linked video)... once again, even if the sphere is deforming (like it is in the video), to support dynamic mesh simulation in ANSYS, it would be essential that you begin with a model having sphere size smaller than the channel (as it is in the video).. in this case, you would do well without breaking the simulation down into two parts..
the problem only arises IMO once 'zero' mesh is encountered at the corners.. once the smaller sphere problem works fine, you could parameterize the geometry and simulate accordingly..

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Simple FSI (fluid structure interaction) Problem	kilroy	Main CFD Forum	0	January 7, 2015 14:42
FSI Structural convergence problem	Blizzard99	CFX	13	December 23, 2014 03:45
Problem with FSI Boundary Condition	ashtonJ	CFX	6	June 28, 2012 04:28
fsi problem	berelian90	CFX	0	October 19, 2010 07:30
how to extend FSI 2D codes to 3D, need advises	abouziar	Main CFD Forum	1	May 30, 2008 04:08

November 20, 2016, 11:23		#2
alijamali New Member Ali Jamali Join Date: Oct 2016 Posts: 5 Rep Power: 9	Anybody? Please.

November 22, 2016, 06:08		#3
fresty Senior Member Join Date: Aug 2014 Location: UK Posts: 213 Rep Power: 13	Negative volume error represents moving mesh issues and resulting unreal mesh elements due to self-intersection.. Have you tried simulating with a model where radius of the ball/ sphere is shorter than the channel? I would think of that working pretty fine without any significant changes in your current model mesh quality.. From top of my head, actually the only way you could simulate this with the moving/ dynamic mesh method is if the circular body is allowed to pass through the channel, the gap however could be extremely minor... on the other hand, this simulation could be divided into 2 parts: (i) ball upward movement by virtue of flow till the very last element possible (ii) ball stuck in the channel (this should be fairly simple & straightforward)..

November 22, 2016, 08:41		#4
alijamali New Member Ali Jamali Join Date: Oct 2016 Posts: 5 Rep Power: 9	Thank you fresty for your response. I would seriously consider breaking the problem down into two parts as you suggested; however, I preferred if I could see how the sphere interacts with the wall and with the fluid, deforms, and passes through the flow restriction. In fact, the whole purpose of this simulation is to observe pass and stop modes as wells as pressure profile through the throat.

November 23, 2016, 00:55		#5
fresty Senior Member Join Date: Aug 2014 Location: UK Posts: 213 Rep Power: 13	I initially assumed that the sphere was a rigid body (without watching the linked video)... once again, even if the sphere is deforming (like it is in the video), to support dynamic mesh simulation in ANSYS, it would be essential that you begin with a model having sphere size smaller than the channel (as it is in the video).. in this case, you would do well without breaking the simulation down into two parts.. the problem only arises IMO once 'zero' mesh is encountered at the corners.. once the smaller sphere problem works fine, you could parameterize the geometry and simulate accordingly..