Learning Forum

[Karthik K R]

How can i select an optimum domain size for fluid flow problems( to eliminate wall effects) ?

[Rob]

Can you post a sketch of what you're modelling?  There are a few guidelines, but essentially they originate from the wind tunnel textbooks. 

[Karthik K R]

Its a scaled under water vehicle

[Karthik Remella]

Hello Karthik,

You will have to succinctly describe your problem if you are trying to find help on this forum. We will not be able to answer general questions here. Please provide as many details as possible when posting your questions. A sketch of your geometry and the application you're trying to simulate is mandatory when seeking help on your issue. It is also important to describe your physical model. You should help us understand what your knowns are and what you intend to understand from the simulation.

Thank you.

Best Regards,

Karthik

[seeta gunti]

Hello Karthik,

 

As Karthik mentioned, it is important to describe your problem accurately when you are looking help on particular problem. 

 Domain size is always depends on what you want to simulate, which models you want to use and what are the boundary conditions that you are using etc. It is difficult to help you with out clear objective of your problem statement.

For example, if it an external aerodynamics, domain sizes are so big compared to internal flows. If I understand, You want to simulate the internal flow of a vehicle similar to flow over a bluff body. If your problem is similar to that, then we consider your vehicle width as "W" then upstream of the vehicle is 5W, downstream minimum 15W, lateral 6W and spanwise 6W will be considered to avoid wall effects. I hope this will help you.

Regards,

Seeta Gunti

[Karthik K R]

My problem is an external flow around a streamline body like the movement of a large fish under water

[Karthik Remella]

Hello,

For an external flow problem, you want to keep your flow domain as large as possible. As a rule of thumb, if you are running a simulation on an airfoil, you might want to have at least 5 chord lengths in all direction (10 preferably). You might want to have larger length in the downstream direction. The main idea is to have boundaries at a considerable distance so they do not see the effect of the wall and you can apply farfield boundary condition on your model. 

I hope this helps.

Best Regards,

Karthik 

[Karthik K R]

Is there any journals or text books available regarding the same..?

[Karthik Remella]

Hello,

I found this on the web. This is short course (step-by-step) on how to generate the mesh in an old tool called Gambit. You need not follow the steps to create the mesh, but it certainly explains the geometry.

https://www.southampton.ac.uk/~zxie/SESS6021/Cornell_airfoil/courses/step1.htm

You could club this with the introductory tutorial on flow over an airfoil example for Cornell. 

https://confluence.cornell.edu/display/SIMULATION/FLUENT+-+Flow+over+an+Airfoil

I hope this helps answer your question.

Thank you.

Best Regards,

Karthik

Guidelines for Posting on the Community

[raul.raghav]

As Seeta and Karthik mentioned, the distance in all the directions play a crucial role. The exact computational size required to accurately capture the physics would vary depending on the problem. But as a general guideline, maintain the upstream and radial distance at 5 chord lengths and the downstream distance between 5-15 chord lengths. You can play with these numbers and see the corresponding effect on the flow field.

Attached below is a paper which uses 1.3L in the upstream and radial directions and 5L in the downstream:

https://www.sciencedirect.com/science/article/pii/S2468013316300729