Learning Forum

[José Mantovani]

Hello everyone!

I want know a thing: I need make a simulation of  a turbulent flow in a channel. In the paper that I was based the simulation have the dimensions of channel of the dimensions is: 2pi x 2 x pi. But I do not understand this.

Could someone explain me better? How would it be? The channel is square section.

Thanks, 

Mantovani. 

[mcoderon]

Hello Jose, 

It seems that the paper you are looking at is giving the non-dimensional size of the domain. If it specify how the quantities are non-dimensionalized you could find the reference dimension and convert the domain to its dimensional size, if you want. The reason of using pi could be related to a specific periodic phenomenon that occurs in the channel. 

Hope that helps,

Marco

[José Mantovani]

Hello Marco, thanks for answering and explaining.

In the article, there is information about geometry, just about the grid. In another article based on what I said, they used a geometry with these parameters 2pi x 2x pi. It is a DNS approach, but I do not know how to know how to assemble the geometry in ANSYS to replicate in LES.

Thanks for help,

Mantovani. 

[DrAmine]

I will say that Marco already wrote: they are using a sort of non-dimensional geometric specifications which is depicting a repeating pattern. You can contact the authors of the paper for clarity. 

[seeta gunti]

Hello Mantovani,

As Macro and Amine has mentioned, Pi is a non dimensional number which they should mention in the paper or they would have given some reference paper about this. If they did not discuss about the non-dimensional number, they should give the units of the domain right! So I suggest please go through the paper once again and check if they mentioned anywhere. If not, try to look at the reference papers. 

Regards,

Seeta

 

 

[raul.raghav]

You would definitely have information regarding the friction Reynolds number and the flow Reynolds number which would help you get an idea of the dimensions. Could you share the paper that you're referring to, so we could take a look at it?

[José Mantovani]

Hello guys!

So Rahul, the paper is found here: https://cfd.spbstu.ru/agarbaruk/doc/1999_Moser-Kim-Mansour_Direct-numerical-simulation-of-turbulent-channel-flow-up-to-Re-590.pdf

Turbulent Channel Flow with a DNS approach. I want the values in metric dimensions because I want to perform a RANS and LES approach through FLUENT. In another article that was based on that, indicated the geometry as: 2pi x 2 x pi. Due to the computational cost, I will realize the value of Rethau = 180.

Thanks for help guys and good afternoon, 

Mantovani. 

[José Mantovani]

Hello guys, someone can take a look in the paper and help me?

Mantovani. 

[DrAmine]

They are using the channel half width as standard dimension. All dimensions and dimensionsless numbers are relative to it. 

[José Mantovani]

Thanks to reply Amine.

The last problem is: in the paper don't have the channel half width dimension in metric dimensions or other dimension form. I think that I need create a geometry in order to obtain the value of Rethau...

Good Weekend,

Mantovani. 

[DrAmine]

Usually the Reynolds number is 10 times larger then the wall shear velocity based one 

[José Mantovani]

Thanks to reply Amine, alright. I will try make some tests here. 

Good Week,

Mantovani.

[DrAmine]

?