[Jamamoto takasi]

Hallo,

I use power law velocity inlet,

U/Uo=(Z/Zo)^alpha

and have some problem when calculate turbulence intensity?
thanks

[Bazinga]

By using the power law to model the atmospheric boundary layer you are just describing the mean velocity profile over height. There is no information about your turbulence intensity. You need to give more information to get help.

[Jamamoto takasi]

Originally Posted by Bazinga

By using the power law to model the atmospheric boundary layer you are just describing the mean velocity profile over height. There is no information about your turbulence intensity. You need to give more information to get help.

Hi Bazinga,

I use power law velocity inlet,

U/Ur=(Z/Zr)^alpha

Ur, Zr, alpha are known
I followed as in an article. That is simulate pollution dispersion in street canyon, around buildings. In there, just have Reynolds number = 10000, viscosity and the geometrical parameters are known: the height of buildings (h).... Please, could you please, help me how to calculate kinematic energy, dissipation rate,...

Thanks

Takasi

[Bazinga]

Do you run a RANS or an LES? What turbulence model do you use?

[Jamamoto takasi]

Originally Posted by Bazinga

Do you run a RANS or an LES? What turbulence model do you use?

Hi,

I use RANS. I run it with k-epsilon, and k-omega SST.

Takasi

[Bazinga]

For RANS and k-eps you have transport equations that calculate your turbulent kinetic energy and its dissipation. You can find the results in each time directory.

[Jamamoto takasi]

Originally Posted by Bazinga

For RANS and k-eps you have transport equations that calculate your turbulent kinetic energy and its dissipation. You can find the results in each time directory.

Hi,

I would like to ask how to calculate the kinetic energy and turbulent intensity? The formula of them?

Thanks

Takasi

[Bazinga]

For RANS it is a transport equation for k. The formula of the equation can be found in Ferziger & Peric for example. Definition of k can be found here:
https://en.wikipedia.org/wiki/Turbulence_kinetic_energy

Turbulence intensity is the quotient of fluctuating part of the velocity and the mean value.
http://www.cfd-online.com/Wiki/Turbulence_intensity

[pela145]

Hello!

This is a question for Tanaka.

I am modelling a building in a flow field with k-epsilon too, and I would like to use a power law to simulate the ABL in the inlet conditions, excatly the same as you.

How or where did you set that conditions? Is there any tutorial that I can take as example?

I used a uniform velocity profile first, and then the ABL log law that is used in the turbineSiting tutorial. I changed the parameters on the U file in the 0 folder, since the atmBoundaryLayerInletVelocity (log law) is already a function and I didn't need to change the source code or anything, I just modified the U file, as I told you.

Pelayo

[clktp]

If you don't have specific values for your ABL case, then you can go for atmBoundaryLayerInletK boundary conditions. It uses generic values for ABL cases. There are ABL boundary conditions for epsilon and nut too.

[Bazinga]

Originally Posted by pela145

Hello!

This is a question for Tanaka.

I am modelling a building in a flow field with k-epsilon too, and I would like to use a power law to simulate the ABL in the inlet conditions, excatly the same as you.

How or where did you set that conditions? Is there any tutorial that I can take as example?

I used a uniform velocity profile first, and then the ABL log law that is used in the turbineSiting tutorial. I changed the parameters on the U file in the 0 folder, since the atmBoundaryLayerInletVelocity (log law) is already a function and I didn't need to change the source code or anything, I just modified the U file, as I told you.

Pelayo

You can use the incompressible/simpleFoam/turbinesiting tutorial. It has all the data you need. Check the k, epsilon and nut file in the 0 time directory.

[pela145]

First of all, thank you to both Bazinga and Taygun.

I had already used the AtmBoundaryLayerInletVelocity, which is a Log Law for the ABL.

However, I want to use a power law, like this:

U=Uref*(Z/Zref)^alpha

Where I can edit the value of Uref, Zref and alpha.

Or something similar, I mean an utility or a BoundaryCondition that give me the Ufield from an expression than I can create and modify.

[Bazinga]

Dear pela,
to a certain extent log law and power law are the same. You can change the values for the log law: uref, zref z0 (roughness length) and u* can be changed to model the abl. Either directly in the U file or maybe in the include/ABLconditions file

[clktp]

atmBoundaryLayerInletVel uses log law, you can copy the U file and include folder in the 0.org folder of the turbineSitting tutorial and use in your case. In include/ABLConditions you can change variables. You should change the U file based on your boundary conditions of course.

[pela145]

Dear Bazinga

But it still would be a Log Law, wouldn't it? My point was precisely comparing between them. Analysing how some different meshes react to both the Log Law and the Power Law profiles. That's why I need it so strongly.

So, you think that there is no way of accomplish this task?

Pelayo

P.D: Anyway, thank you for your response.

[clktp]

Sorry, I misread what you wrote.

I guess you can use swak4Foam to generate inlet velocity for power law. I might be wrong, I remember I did something similar in the past, but I don't recall exactly.

[pela145]

Don't worry taygun!

Originally Posted by clktp

Sorry, I misread what you wrote.

I guess you can use swak4Foam to generate inlet velocity for power law. I might be wrong, I remember I did something similar in the past, but I don't recall exactly.

Yes, I think so. I have found it out today. But, that librery have two utilities quite similar, funkySetBoundaryFields and funkySetFields, and I am thinking about which of them is better. Or better said, which would be more useful. Because they both are completely new for me.

Do you know any of them?

Pelayo

[clktp]

I used both and if I'm not wrong funkySetFields is used to generate initial conditions and the other one is for boundary conditions. I don't remember how to use it exactly, but it's quite simple. There should be good tutorial for both of them. If you need any question you can ask, maybe if you show something I remember.

By the way if you wanna use power-lag low then you may need to change the k and epsilon, because they are designed for log-law. You can generate inlet conditions by using funkySetBoundary fields too for k and epsilon.

[pela145]

Yes, i have noticed that i need epsilon and k profiles too. The table that I have used for the U profile also has the values of K. However, it hasn't got the epsilon values, so i am a bit blind here. I have looked for some equations that could be helpful, in order to get another table of discrete values to set with setdiscreteFields. But i didn't find anything. All those equations that i found were hard to simulate with Excel, since they contains partial derivatives and very especific terms like Pk, the production term of Kinetic energy, which I absolutely don't know.

My question now is, do you know any simple (more or less) way of obtaining k and epsilon from U, the geometry of the case and/or simple constants (Cmu, C1, C2...)?

I know it is hard, but any info (even the negative one) will be very helpful. Because the other option that I am thinking at, is using the U power law profile and uniform values for k and epsilon. Even though I know it would be wrong. But, for the moment, it is that or nothing.

Pelayo

[clktp]

For constants you can use the constants in the turbineSitting case, the reference is written as well.
OpenFOAM/OpenFOAM-2.x.x/tutorials/incompressible/simpleFoam/turbineSiting/constant/RASProperties

You can obtain k and epsilon in a simple way from here,
http://www.cfd-online.com/Wiki/Turbu...ary_conditions