Flow around a cylinder with initial disturbance

freddiew93 · November 28, 2015, 09:11

Hi

I am studying the effects of turbulence modelling on the near-wake flow past a circular cylinder. According to literature on the subject an initial disturbance needs to be included in the model in order to break the symmetry of the flow and hence trigger the generation of vortex shedding. A standard practice is to rotate the cylinder a few turns, say 4, clockwise over N time steps at the beginning of the simulation.

My question is how would one implement that in CFX? I don't have that much experience using the software so I would appreciate some advice, preferably in layman's terms.

Cheers

freddiew93 · November 28, 2015, 12:38

Would setting the following expressions work?

nturns = 4;
nTS = 200;

vel_rot = (2*pi*nturns)/nTS *(1-step(ctstep - nTS));

and setting the cyilinder as a rotating wall with velocity vel_rot around vertical axis. Would that result in cylinder turning nturns times during the first nTS timesteps and then remaining stationary for the remainder of the simulation?

I'm having difficulties understanding differences between current and accumulated time steps.
I'd also like this expression to work universally for steady and transient simulations?
Any suggestions welcomed

Thanks

ghorrocks · November 28, 2015, 19:33

First of all - it is unlikely you need to rotate the cylinder to induce the asymmetric flow. I have never had to do it before for cylinder flows. There is enough asymmetry in the mesh and noise in the numerical algorithm that if the flow is unstable it will start all by itself.

Ignoring that and assuming you want to rotate the cylinder anyway:
I think your equation is wrong. Something like:
rotational_velocity = if(ctstep<200,10[rpm],0[rpm])
looks better (and simpler) to me. I am just using 10rpm rather than 4 turns, it is a bit arbitrary anyway.

November 28, 2015, 09:11	Flow around a cylinder with initial disturbance	#1
freddiew93 New Member Fryderyk Join Date: Nov 2015 Posts: 2 Rep Power: 0	Hi I am studying the effects of turbulence modelling on the near-wake flow past a circular cylinder. According to literature on the subject an initial disturbance needs to be included in the model in order to break the symmetry of the flow and hence trigger the generation of vortex shedding. A standard practice is to rotate the cylinder a few turns, say 4, clockwise over N time steps at the beginning of the simulation. My question is how would one implement that in CFX? I don't have that much experience using the software so I would appreciate some advice, preferably in layman's terms. Cheers

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
settlingFoam unstable?	bendel_boy	OpenFOAM Running, Solving & CFD	38	July 8, 2016 05:07
Floating point exception error	lpz_michele	OpenFOAM Running, Solving & CFD	53	October 19, 2015 02:50
Upgraded from Karmic Koala 9.10 to Lucid Lynx10.04.3	bookie56	OpenFOAM Installation	8	August 13, 2011 04:03
Differences between serial and parallel runs	carsten	OpenFOAM Bugs	11	September 12, 2008 11:16
Unknown error	sivakumar	OpenFOAM Pre-Processing	9	September 9, 2008 12:53

November 28, 2015, 12:38		#2
freddiew93 New Member Fryderyk Join Date: Nov 2015 Posts: 2 Rep Power: 0	Would setting the following expressions work? nturns = 4; nTS = 200; vel_rot = (2pinturns)/nTS *(1-step(ctstep - nTS)); and setting the cyilinder as a rotating wall with velocity vel_rot around vertical axis. Would that result in cylinder turning nturns times during the first nTS timesteps and then remaining stationary for the remainder of the simulation? I'm having difficulties understanding differences between current and accumulated time steps. I'd also like this expression to work universally for steady and transient simulations? Any suggestions welcomed Thanks

November 28, 2015, 19:33		#3
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	First of all - it is unlikely you need to rotate the cylinder to induce the asymmetric flow. I have never had to do it before for cylinder flows. There is enough asymmetry in the mesh and noise in the numerical algorithm that if the flow is unstable it will start all by itself. Ignoring that and assuming you want to rotate the cylinder anyway: I think your equation is wrong. Something like: rotational_velocity = if(ctstep<200,10[rpm],0[rpm]) looks better (and simpler) to me. I am just using 10rpm rather than 4 turns, it is a bit arbitrary anyway.