[saugatshr4]

I am relatively new to Ansys Fluent. For Reynold's number of 30000, I'm getting a drag of 0.5 (The correct drag should be around 1.2) with a time step of 0.002 seconds. I ran the simulations for 25 seconds, and it still is the same?? I'm running the K-epsilon model for airflow around a cylinder. What should I do to get the correct result??
-Decrease the time step
-Use finer Mesh
-Run the simulation for longer to obtain the correct result

[Vaelyn]

You could refine the mesh by the cylinder-gas interface. The time-step and seems fine if it is converging well.

[arnie333]

You need to describe your analysis in more detail.
Is it an external or internal flow analysis ?
What are you analyzing and where did you get the correct Cd of 1.2 ?
Did you input the correct Reference Values ?

If the flow is around an object then Re of 30 000 is turbulent flow and k-epsilon model is acceptable if Y+ lies between 30 and 300. Remember that you are not directly resolving the shear stress, but rather using a model that applies a near-wall treatment.
To accurately resolve the wall shear stress in the boundary layer, I would rather use k-omega SST with a Y+ between 0.1 and 2

[shereez234]

Originally Posted by saugatshr4

I am relatively new to Ansys Fluent. For Reynold's number of 30000, I'm getting a drag of 0.5 (The correct drag should be around 1.2) with a time step of 0.002 seconds. I ran the simulations for 25 seconds, and it still is the same?? I'm running the K-epsilon model for airflow around a cylinder. What should I do to get the correct result??
-Decrease the time step
-Use finer Mesh
-Run the simulation for longer to obtain the correct result

Like others said. So many questions comes to mind?


Geometry is 2D?
What reference area are you using?
Is your Y+ <1?
Is your viscosity and far field velocity/pressure correctly set?
Do you have a plot of the convergence?

[saugatshr4]

It is an internal wind tunnel analysis. The walls and the cylinder is initialized as the no-slip condition. It is a 2D condition. I am using the standard default reference values. The reference value is 1, and even when I calculate my Cd it is incorrect. I am getting somewhere around 1.2 but I am getting 0.5.
So, from the literature from the flow around a cylinder the Cd is 1.2. It attainable for low Reynolds number till 1000 but I seem to have a problem for higher Reynolds number.
I am trying the k-w model, as it gives better results for the near-wall region. I have the same problem with incorrect drag.
I m trying to have the +y value less than 1 but, it seems that the mesh size has to be substantially small to attain >+y condition. How do we get a +y<1? I have attached the images of the associated files. https://drive.google.com/drive/u/0/f...moHiW5vV31k_tb

[vinerm]

It is the drag force that is based on the flow-field and not its coefficient. is the ratio of the drag force to some normalizing force. And, hence, if you use a different force to normalize, you get a different value and all those values are correct as long as you also report the normalizing force used. To avoid this, standard normalization values are used. Therefore, you have to modify your reference values; Fluent uses these reference values to calculate normalizing component. You have to modify length scale, velocity scale, and density. Use free-stream values for the velocity and density, and diameter of the cylinder as length scale. is alright; lower could be better but it is not bad.