CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   FLUENT (https://www.cfd-online.com/Forums/fluent/)
-   -   Simulation of cylinder aligned in the flow direction using k-e model (https://www.cfd-online.com/Forums/fluent/103161-simulation-cylinder-aligned-flow-direction-using-k-e-model.html)

 Eliasjal June 12, 2012 13:06

Simulation of cylinder aligned in the flow direction using k-e model

1 Attachment(s)
Hello everyone,

First of all , i' still a beginner to fluent :) and what I'm trying to simulate is the boundary layer flow over a cylinder aligned in the flow direction. Please see the attached picture that shows the boundary conditions. I used a periodic boundary between the inlet and the outlet of the domain whereas the mass flow rate was taken. Then I assumed the wall has no slip and outflow pressure for the outer domain.

I need some help from you guys to resolve the following questions:

1- Is the BCs that i used are corrects ?? I wonder whether or not do i have to use the field pressure instead of outflow pressure condition ??
2- How to define the number of layers that will be used for a wall zone. I need to plot Y+ vs u+ so I need to define the wall sublayer that will be used for y+ analyses ??
3- Do you guys have and reference link, paper or materials that shows a similar simulation that i can use for comparing my results ?
4- In case of a spinning wall, do i need to use a dynamic mesh or just using the cell zone conditions ?? and if so then what is the proper choice, the frame motion or the mesh motion ???

thank you very much for your help
Regards,

EJ

 delaneyluke June 13, 2012 04:27

First of your external domain has to be way bigger if you want to capture the flow over your cylinder. Browse the net for external domain sizes for your type of problem.

1. Use a velocity inlet and pressure outlet. For outer walls use the far field pressure BC
2. Use any Y+ calculator available to find out first layer height
3. Not at the moment
4. In case of a spinning wall, you can give the angular velocity to the wall, you don't need to do a dynamic mesh analysis

P.S You could work with a sector instead of the entire cylinder

Regards
Luke

 Eliasjal June 13, 2012 08:00

Hey Luke,

First of all thank you much for your response.

If there is any specific relation that define how much the size of the domain must be ?? or how large the external domain (i.e the outer cylinder diameter ) should be for my case ??

Actually I tried to avoid using a rectangular box just to eliminate some edge effect at the corners of the domain which happens in the rectangular box domain.