[Konstantine]

Hello everyone,

I need some assistance on understanding why in the drag coefficient i am getting the correct frequency according to the Re number but the amplitude is different from studies in papers. icoFoam is the solver and for different meshes i get different amplitudes but same frequency. the amplitude should oscillate around 1.44 and it oscillates around 3.
Does anybody know about amplitude dependency on either numerical or physical factors that could help me understand why.

[mohsen cheraghi]

Originally Posted by Konstantine

Hello everyone,

I need some assistance on understanding why in the drag coefficient i am getting the correct frequency according to the Re number but the amplitude is different from studies in papers. icoFoam is the solver and for different meshes i get different amplitudes but same frequency. the amplitude should oscillate around 1.44 and it oscillates around 3.
Does anybody know about amplitude dependency on either numerical or physical factors that could help me understand why.

Hi, 1.44 and 3. It seems that in the references they report the amplitude of magnitude drag and you consider negative and positive drags and its magnitude.

[Konstantine]

thank you my friend mohsen for your response.
So if i understand correctly you suggest that the calculation adds up to more than it should do, meaning the magnitude as it is calculated by the total drag force + the negative and positive oscillations so it adds up to more than just the magnitude. You could have a point there but how is it calculated? i am searching now from the forceCoeffs.C routine to understand better what it actually calculates and what formula it uses for the coefficient.

[Konstantine]

Dear mohsen,

i found how it calculates the coefficient. it is the usual formula of force/(dynamic pressure* area). then i found that it adds pressure, viscous and porous forces to calculate the total force. so it is something about the last three that is wrong, i am thinking that maybe porous is not zero so it gives different result than usually or pressure or viscous friction is overpredicted. i am not sure yet i will continue searching.

[sherif35]

Hello,

I'm having what I assume is a similar issue, I'm getting higher average Cd than what I should be getting, however, not as much as yours.

I started looking at sectional Cd in the vertical direction and i'm finding two distinct groups of sectional Cd curves, one group is higher than the other. I described the issue in the following thread

http://www.cfd-online.com/Forums/ope...direction.html


Please take a look and let me know what do you think. Thank you.

[Konstantine]

hello my friend sherif35
thank you for your concern on the topic.
total drag= pressure drag + friction drag
so what you see there may be the pressure drag and the friction drag , those will oscillate around a non-zero value in contrast to the lift which oscillates around zero.

[sherif35]

Originally Posted by Konstantine

hello my friend sherif35
thank you for your concern on the topic.
total drag= pressure drag + friction drag
so what you see there may be the pressure drag and the friction drag , those will oscillate around a non-zero value in contrast to the lift which oscillates around zero.

Hello Konstantine,

The sectional Cd curves I show are for the total drag on the section. I'm still trying to figure out why certain sections will oscillate about certain value while the rest oscillate around another!!

[Konstantine]

i guess you talk about sections of the cylinder, sorry i did not understand the sectional forces well, it is reasonable that because of different pressures (due to vortex shedding) around the cylinder you will have different forces on the cylinder wall points, therefore it would oscillate around different values for each section but the frequency should be the same as vortex shedding is the phenomenon responsible for these pressures. if you have the same forces let's say on all sections then either you would have symmetric vortex shedding which can be the case ( i think only numerically achieved) some times but is not stable and will brake after a couple of shedding periods or symmetric laminar flow like on Re=30-50 where there is a wake bubble stable that just circulates in two symmetric regions the water behind the cylinder and does not shed any vortex.
In general i believe you are interested in the total drag and not on sectional unless you want to compare with experimental on particular points of the cylinder.

[sherif35]

Originally Posted by Konstantine

i guess you talk about sections of the cylinder, sorry i did not understand the sectional forces well, it is reasonable that because of different pressures (due to vortex shedding) around the cylinder you will have different forces on the cylinder wall points, therefore it would oscillate around different values for each section but the frequency should be the same as vortex shedding is the phenomenon responsible for these pressures. if you have the same forces let's say on all sections then either you would have symmetric vortex shedding which can be the case ( i think only numerically achieved) some times but is not stable and will brake after a couple of shedding periods or symmetric laminar flow like on Re=30-50 where there is a wake bubble stable that just circulates in two symmetric regions the water behind the cylinder and does not shed any vortex.
In general i believe you are interested in the total drag and not on sectional unless you want to compare with experimental on particular points of the cylinder.

Thank you for your quick response.
You are correct, I'm interested in the total drag, but what I don't really understand is why there are two distinct group of sectional Cd curves. If they are all different I would have assumed that this is 3-D effects. But in my case, it looks like certain sections produce certain results.

If you look at the attached picture for Re=10,000 you will see section 3&6 Cd curves are exactly the same (until the shedding starts) and the rest of the sections are higher.

I think if I understand what causes this issue for the sectional Cd, it will help me with the average drag for the whole cylinder.

So I'm wondering if you have seen similar sectional Cd issue with your simulation?

[Konstantine]

everything is about pressure, i have seen different pressures around the cylinder wall yes ( therefore the forces and the coefficients are different), this is reasonable. on Re=200 you may have 3d effects and maybe transition to turbulence which makes the phenomena more complicated. depending on how many parts you have devided the cylinder there may be parts which have the same pressure curve e.g. the front parts of the cylinder before the seperation. after the separation, the pressure will oscillate due to vortex shedding or be more chaotic due to turbulent eddies which may appear because of transition to turbulence. you calculate the total force and then the coefficient so the average coefficient will be in the middle around which it oscillates. now if you want to relate the sectional coefficients on particular parts of the cylinder with the total average, think that calculate the total force you add the forces on each node where it is calculated so eventually the sectional average Cds would add up to the total average Cd. think if you really need that have you found any similar research, it has been many times when i was stuck with ideas and i was searching without a reason. you may be correct ok i am not questioning your idea but is it worth it as a result what do you want to prove.