[pradeap]

Hi

I am working on a flow over a cylinder problem in 2D. I am in the process of validating my code over this example.

I would like to know how to calculate the vortex shedding frequency in order to obtain the strouhals number.

Thanks

[Abdul kadir]

HI Pradeap !

I am also working on "Flow induced vibration over in a heat exchanger"

& Need to know required data & information to be collected for same.

Please help me for better start in this direction...

I am new to CFD tools like FLUENT,GAMBIT,CFX..

I am student of MTech thermal Engg.

Thank you.

[swamysrikanth]

hello pradeap
to find out the St u required variation of lift coefficient vs t file, from that data after reaching the steady state (ie equal amplitude) calculate the time required for 1 cycle. inverse of that time is frequency. thereby u can calculate the the St (if u know the Umax, d).

[pradeap]

is it enough to plot the variation of lift coefficient is enough or we should take Fast fourier transform of Lift coefficients and follow the same procedure as told by u..

Thanks

[swamysrikanth]

just lift coefficient data obtained as u run the simulation is enough for calculation of St.

[vmlxb6]

I also need to find the vortex shedding frequency from a vibrating cylinder.

Is this procedure the same for vibrating cylinders or just stationary cylinders ???

Thank you.

[lava12005]

Hi,

i'm also doing Flow aroung a cylinder thing. I wonder how do we get the value of the steady state coefficient, let say steady state C_D. Or how long does it take (in nondimensional time) for a simulation usually reach steady state condition?

I did one case with Re = 40, from the papers that I read, I found that usually the steady state C_D is about 1.5++. What happened with mine is that after 10 nondimensional time (t* = tU/D), the value of the C_D has gone down to 1.3 and it keeps decreasing..

And for vortex shedding part, at what Re does it usually start? And how many nondimensional time that we will be able to see the wake oscillating?

[vmlxb6]

Check out the following image for vortex shedding at different Re:

http://img62.imageshack.us/i/reovercyl.jpg/

I tried simulating a steady state flow over a stationary cylinder at Re=200. The contours obtained from FLUENT were different than that obtained from CFX. If your using an upwind scheme, the convergence is much much better as compared to Higher order.
Why do you want to run a steady state ??? If you want to see vortex shedding, try running Transient. If your Re number is small, give an initial perturbation to the flow. For low Re number, the vortex shedding takes time to develop.

-Pof. Chaos

[Shenren_CN]

In this case, you can still plot Cl vs t, and you will generally get 2 different kinds of results:
1. one dominant frequency, meaning vortex sheds at the same frequency as the cylinder vibration (frequency lock-in)
2. two frequency, meaning vertex sheds at a different frequency from the cylinder vibration. this happens when cylinder vibration amplitude is relatively small.

Also, it depends on if you are prescribing a vibration for the cylinder, or is the cylinder vibration coupled with the flow, which is then a fully coupled aero-elastic problem. But the frequency analysis remains the same, except the CFD part will be much more involved in that case.

Originally Posted by vmlxb6

I also need to find the vortex shedding frequency from a vibrating cylinder.

Is this procedure the same for vibrating cylinders or just stationary cylinders ???

Thank you.

[Milad]

If we have a rotating aerofoil, wind turbine aerofoil, is it correct to consider the turbine rotating frequency as the vortex shedding frequency?

[leflix]

Originally Posted by pradeap

Hi

I am working on a flow over a cylinder problem in 2D. I am in the process of validating my code over this example.

I would like to know how to calculate the vortex shedding frequency in order to obtain the strouhals number.

Thanks

Hi Pradeap,

from your simulation, record the vertical velocity componnent in a given monitor point of the wake of the cylinder versus time. Let's say V(t). Apply a Fast Fourrier Transform (FFT) to this velocity signal versus time.
From the power spectrum or amplitude spectrum you will notice a peak frequency named fundamental frequency. This is the vortex shedding frequency you are looking for.

[lovecraft22]

I have been doing a lot of work in a wind tunnel on the wake of a circular cylinder and the way to follow is that suggested by leflix.

You can get the vortex shedding by velocity signals (i.e. hot wire anemometer) in the wake, by pressure fluctuations on the cylinder surface or by the forces fluctuation (remember that the fluctuation of the drag occurs at twice the frequency of the vortex shedding while the fluctuation of the side force is that of the vortex shedding).

The best way in wind tunnel is the anemometry though because the balance signals take into account the natural frequencies of the balance itself (which need to be evaluated wind-off) while the pressure signals will miss the low frequencies.

Obviously one will not have this issue in a simulation.

Once you have your signal you may use the Fourier transform, the Wavelet transform (better) or even Proper Orthogonal Decomposition (POD) to get the dominating frequency which needs to be multiplied for the cylinder diameter and divided by the fluid speed to get the Strouhal number.

[uli]

Hi all,

for validating my simulation results I am interested in experimental data of vortex shedding in the wake of a circular cylinder (other geometries are welcome as well) at 1e4 < Re < 1e5.

So far I found 1.2 to be the mean C_D value for a circular cylinder at 2e4 < Re < 1e5 (Roshko 1960).

Schewe 1983 seems to be a interesting paper, but I did not find the fulltext yet (does anyone have the paper?).

Besides that I would find it very interesting to see which configurations (controldict, fvsolution, fvschemes, grid) were found to produce reasonable results for turbulent vortex shedding in OF.

Anyone working on that topic, too?

Uli

[cesc]

Originally Posted by pradeap

Hi

I am working on a flow over a cylinder problem in 2D. I am in the process of validating my code over this example.

I would like to know how to calculate the vortex shedding frequency in order to obtain the strouhals number.

Thanks

1-In your output, CL or Cd is faqtuating, pick to pick of CL or CD is your period time. Inverse is, that is frequency.

2- capture your contour, spot when vortex is shedding from one point, and write zone number it, then the difference between two zone number that is vortex is shedding from one point, is period time.

[divyemalik14]

Hello Prof,

Can you explain how can we give initial perturbation for the vortex shedding

[shailesh jha]

Hii
I am working on flapping flat plate. I am validating my case.
Can somebody help me in finding Cd value as in my case i am getting Cd in the range of 1and 1.4 for 5 degree Angle of attack for Re = 10000 range which is not correct . What must be the region for my error ? should i use mean drag ? And how mean drag is taken in transient case.

[shirlin]

hi.
i am trying to capture the vortex shedding of a bridge deck for a Re 10^5,,when i run a unsteady simulation my lift vs time plot doesnt start with zero.why is that so,, also i tried for various meshes and the lift plot is constant for a particular mesh and fluctuates for a different mesh.(all done for same Re).could someone pls expalin.

[cfluid]

Originally Posted by leflix

Hi Pradeap,

from your simulation, record the vertical velocity componnent in a given monitor point of the wake of the cylinder versus time. Let's say V(t). Apply a Fast Fourrier Transform (FFT) to this velocity signal versus time.
From the power spectrum or amplitude spectrum you will notice a peak frequency named fundamental frequency. This is the vortex shedding frequency you are looking for.

I'm a bit confused on this. Would it make a difference between taking the FFT of the lift coefficient and taking the FFT of the velocity signal in a monitoring point. Especially in a tube bundle?

Thank you