FSI Mesh Motion Trouble

cfd noob · April 6, 2015, 19:36

I am trying to carry out a 2D FSI simulation using a rigid body solver in water. I have attached the picture of the mesh. I have used no expression so far in my simulation. I have specified a spring with its equilibrium position at the Center of the hole in the center. I am running the simulation with a laminar flow at a Re of 120.I am not getting any Mesh Displacement in Y(I know this as I have set a monitor point at the circumference of the hole). Is it a problem with my flow (Is the force too small to move the mesh) or is it a problem with my setup ?

Extra Info

Flow is in the x direction and im monitoring motion in Y direction. I have used a velocity Inlet and Pressure outlet (reference pressure=0). I have set the mesh motion of all boundaries to unspecified except the interface which uses the rigid body solver.

mvoss · April 8, 2015, 05:24

Did you get any vortex shedding behind the cylinder? How did you set/model the displacement?

ghorrocks · April 8, 2015, 06:55

This is a classic case of where you should make sure you can model the simple problem accurately before you consider the complex one.

I strongly recommend you model a stationary cylinder in cross flow and make sure you can model the vortex street accurately enough for your purposes. There are extensive data available on the web for this flow so you have high accuracy solutions to compare against.

Once you can model the simple stationary cylinder model then you activate FSI and try to model the more complex combined simulation.

cfd noob · April 10, 2015, 13:59

Thank you Matthias,
Yes I get vortices, but i have not modeled the displacement . I expect cfx solver to internally solve for mesh displacement. Am i wrong in that aspect?

(Just to check if the mesh moves at all, I simulated the flow for 10 sec and gave it a large spring force to move it 2*diameter distance in that time. As i was trying this the mesh always failed due to negative element so i set the time step down to 0.01 sec and yet got results only up-to 8 sec, beyond that the mesh folds. Does this mean my mesh is faulty. Now as i was simulating this I noticed that I got a mesh displacement in Y = 0 at every time-step but some increasing value for the Total mesh displacement in Y. How do i visualize this displacement in CFX post? )

Thank you Glenn
I have already simulated the stationary cylinder case. I got laminar vortex street at Re 120. To validate stationary cylinder I wanted to match my results to the St Vs Re plot. I am not aware how to measure the shredding frequency and hence could not complete the validation. How do i measure the vortex shredding frequency in CFX post?

ghorrocks · April 12, 2015, 08:48

The best way to measure the vortex shedding frequency is to put a monitor point in the cylinder wake and monitor the pressure or velocity. Rerun the simulation and you should get a periodic trace, and measure the frequency.

mvoss · April 13, 2015, 08:32

I would recommend to do as Glenn said and model this step by step. For the displacement just model the displacement w.r.t. e.g. the time-step (via sin(f(Timestep))*1[cm] e.g.) and run the simulation without solving for the fluid to check if the displacement-formulation (convergence, tightness of displacement residuals) is capable to run more than let´s say 4 or 5 periods of motion. So you can check for the mesh to be appropriate over the whole expected movement with the expected vortex shedding.

viraj20feb · October 26, 2016, 08:08

Quote:

Originally Posted by mvoss

I would recommend to do as Glenn said and model this step by step. For the displacement just model the displacement w.r.t. e.g. the time-step (via sin(f(Timestep))*1[cm] e.g.) and run the simulation without solving for the fluid to check if the displacement-formulation (convergence, tightness of displacement residuals) is capable to run more than let´s say 4 or 5 periods of motion. So you can check for the mesh to be appropriate over the whole expected movement with the expected vortex shedding.

Hello Mr. Matthias,

I want to plot the displacement over time curve to validate a FSI solver in foam-extend-3.2. Can you please explain how to plot displacement over time? Thank you

ghorrocks · October 26, 2016, 08:29

Try the open foam forum. This is the CFX forum, we know nothing about Openfoam here.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
how to set periodic boundary conditions	Ganesh	FLUENT	15	November 18, 2020 07:09
3D Hybrid Mesh Errors	DarrenC	ANSYS Meshing & Geometry	11	August 5, 2013 07:42
[ICEM] Generating Mesh for STL Car in Windtunnel Simulation	tommymoose	ANSYS Meshing & Geometry	48	April 15, 2013 05:24
3D valve motion but no mesh motion with smoothing	slaol2	FLUENT	1	September 22, 2011 19:06
large scale mesh motion	sb	FLUENT	1	April 27, 2007 23:23

April 8, 2015, 05:24		#2
mvoss Senior Member Matthias Voß Join Date: Mar 2009 Location: Berlin, Germany Posts: 449 Rep Power: 20	Did you get any vortex shedding behind the cylinder? How did you set/model the displacement?

April 8, 2015, 06:55		#3
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	This is a classic case of where you should make sure you can model the simple problem accurately before you consider the complex one. I strongly recommend you model a stationary cylinder in cross flow and make sure you can model the vortex street accurately enough for your purposes. There are extensive data available on the web for this flow so you have high accuracy solutions to compare against. Once you can model the simple stationary cylinder model then you activate FSI and try to model the more complex combined simulation.

April 10, 2015, 13:59		#4
cfd noob New Member Join Date: Mar 2015 Posts: 2 Rep Power: 0	Thank you Matthias, Yes I get vortices, but i have not modeled the displacement . I expect cfx solver to internally solve for mesh displacement. Am i wrong in that aspect? (Just to check if the mesh moves at all, I simulated the flow for 10 sec and gave it a large spring force to move it 2*diameter distance in that time. As i was trying this the mesh always failed due to negative element so i set the time step down to 0.01 sec and yet got results only up-to 8 sec, beyond that the mesh folds. Does this mean my mesh is faulty. Now as i was simulating this I noticed that I got a mesh displacement in Y = 0 at every time-step but some increasing value for the Total mesh displacement in Y. How do i visualize this displacement in CFX post? ) Thank you Glenn I have already simulated the stationary cylinder case. I got laminar vortex street at Re 120. To validate stationary cylinder I wanted to match my results to the St Vs Re plot. I am not aware how to measure the shredding frequency and hence could not complete the validation. How do i measure the vortex shredding frequency in CFX post?

April 12, 2015, 08:48		#5
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	The best way to measure the vortex shedding frequency is to put a monitor point in the cylinder wake and monitor the pressure or velocity. Rerun the simulation and you should get a periodic trace, and measure the frequency.

April 13, 2015, 08:32		#6
mvoss Senior Member Matthias Voß Join Date: Mar 2009 Location: Berlin, Germany Posts: 449 Rep Power: 20	I would recommend to do as Glenn said and model this step by step. For the displacement just model the displacement w.r.t. e.g. the time-step (via sin(f(Timestep))*1[cm] e.g.) and run the simulation without solving for the fluid to check if the displacement-formulation (convergence, tightness of displacement residuals) is capable to run more than let´s say 4 or 5 periods of motion. So you can check for the mesh to be appropriate over the whole expected movement with the expected vortex shedding.

October 26, 2016, 08:29		#8
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	Try the open foam forum. This is the CFX forum, we know nothing about Openfoam here.