Vortex Induced Vibration

vidya.rudn · May 7, 2017, 06:54

Hello all,

I am trying to simulate vortex shedding behind a 3D cylinder in Fluent for Re = 1000. I have tried laminar, k-epsilon, k - omega SST also LES. In all cases the normalized drag is above 0.9 and lift is nearly 0.4. Even after 140 sec of flow time the lift oscillations are not steady. In RANSE it is going on increasing and in LES it is decreasing. My time step size is 0.001. The major problem is that I am not getting the expected shedding frequency at Re 1000 corresponding to Sr = 0.2. What I am expecting is 3.2 secs, but I get 10 - 14 secs in various simulations. Somebody please guide me through this. Time histories are attached here.

vidya_k epsilpon.jpg

vidya_k omega SST.jpg

vidya_LES.jpg

Thankyou.

LuckyTran · May 7, 2017, 15:45

I'm not sure why you get the wrong shedding frequency, that seems to be important and you should focus on fixing that.

But let's say the 10-14 s vortex shedding cycle is the correct one. You need to run your simulation for many shedding cycles, in your sims so far you have only a few cycles. You also need to run it for a few flow-thru times so that the vortex is advected out of the domain. It should eventually reach some statistically stationary state. If it doesn't it's a hint that you have accumulation of numerical errors.

Btw is this constant density or ideal gas? If ideal gas, make sure you are using the non-reflecting boundary conditions, otherwise the acoustic energy accumulates in the domain.

vidya.rudn · May 8, 2017, 22:01

Thank you so much for your reply.

I used water as fluid( density = 998.2 kg/m3). I have run the sims. for more than 140 secs. But it seems that since the time period I get is larger there are only few cycles. I have seen many works where shedding and steady periodic oscillation of lift and drag starting much earlier.
Recently I simulated it for 2D unstructured mesh and I am getting the proper values of shedding frequency and strouhal number. I cant identify what is the problem in 3D simulation.

shashanktiwari619 · May 11, 2017, 11:50

I have three questions
1. Why did you choose a time step of 0.001. Is the CFL condition satisfied for this time step for the simulations?
2. What is the reference area you are using?
3. How are you calculating the Strouhal number? Are you using the FFT function for calculating it.

May 7, 2017, 06:54	Vortex Induced Vibration	#1
vidya.rudn New Member vidya Join Date: Mar 2017 Posts: 2 Rep Power: 0	Hello all, I am trying to simulate vortex shedding behind a 3D cylinder in Fluent for Re = 1000. I have tried laminar, k-epsilon, k - omega SST also LES. In all cases the normalized drag is above 0.9 and lift is nearly 0.4. Even after 140 sec of flow time the lift oscillations are not steady. In RANSE it is going on increasing and in LES it is decreasing. My time step size is 0.001. The major problem is that I am not getting the expected shedding frequency at Re 1000 corresponding to Sr = 0.2. What I am expecting is 3.2 secs, but I get 10 - 14 secs in various simulations. Somebody please guide me through this. Time histories are attached here. vidya_k epsilpon.jpg vidya_k omega SST.jpg vidya_LES.jpg Thankyou.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
vortex induced vibration from dynamic mesh	natto	Main CFD Forum	4	April 18, 2015 19:42
vortex induced vibration	dorin	CFX	11	December 1, 2009 13:03
Vortex Induced Vibration modelling	Kwong	FLUENT	0	March 24, 2007 00:41
"bit" phenomenon in Vortex Induced Vibration?	JOEY	Main CFD Forum	0	August 4, 2006 08:49
Vortex Induced vibration of subsea pipeline	Mobasheri	FLUENT	5	February 9, 2006 07:07

May 7, 2017, 15:45		#2
LuckyTran Senior Member Lucky Join Date: Apr 2011 Location: Orlando, FL USA Posts: 5,674 Rep Power: 65	I'm not sure why you get the wrong shedding frequency, that seems to be important and you should focus on fixing that. But let's say the 10-14 s vortex shedding cycle is the correct one. You need to run your simulation for many shedding cycles, in your sims so far you have only a few cycles. You also need to run it for a few flow-thru times so that the vortex is advected out of the domain. It should eventually reach some statistically stationary state. If it doesn't it's a hint that you have accumulation of numerical errors. Btw is this constant density or ideal gas? If ideal gas, make sure you are using the non-reflecting boundary conditions, otherwise the acoustic energy accumulates in the domain.

May 8, 2017, 22:01		#3
vidya.rudn New Member vidya Join Date: Mar 2017 Posts: 2 Rep Power: 0	Thank you so much for your reply. I used water as fluid( density = 998.2 kg/m3). I have run the sims. for more than 140 secs. But it seems that since the time period I get is larger there are only few cycles. I have seen many works where shedding and steady periodic oscillation of lift and drag starting much earlier. Recently I simulated it for 2D unstructured mesh and I am getting the proper values of shedding frequency and strouhal number. I cant identify what is the problem in 3D simulation.

May 11, 2017, 11:50		#4
shashanktiwari619 New Member Shashank Join Date: Sep 2014 Posts: 28 Rep Power: 11	I have three questions 1. Why did you choose a time step of 0.001. Is the CFL condition satisfied for this time step for the simulations? 2. What is the reference area you are using? 3. How are you calculating the Strouhal number? Are you using the FFT function for calculating it.