Can someone help please, I'm confused...

I'm trying to create a simple spreadsheet to give wind speeds at a range of heights, from one input wind speed at a reference height.

I want to do it using the log law, where u is a function of friction velocity. Friction velocity is fucntion of the wall shear stress. And the wall shear stress is a function of du/dy. I don't understand how to use this du/dy? I have 1 value for flow velocity (u) at height (y). I know du/dy is not simply u/y, but don't understand what else I can do!

Would greatly appreciate any help. Thank you.

how many u velocity you have?

Since u = 0 at the wall (where y = 0) then to a first approximation du/dy ~ u/y, where these values are taken very close to the wall. Depending on how far you are away from the wall this approximation will either be a good representation of du/dy, or a really bad approximation of du/dy. If you are assuming turbulent flow, and your reference velocity is in what would be the free-stream, then you would be better off using a power-law profile (see Schlichting or some other basic fluid mechanics reference). Is there a reason why you want to use the log-law profile?

You need something extra: the rougness on the floor/bottom. In the atmosferic boundary layer world it's often called 'z0' and it corresponds to a kind of aerodynamic rougness length (it can be converted to Schlichtings sand rougness: z0=ks*exp(-kappa*B)). For a meteo mast on a plain grass field the 'z0' is 3 cm (standard WMO).

Take the log-law equation:

u+ = 1/kappa ln( (z+zd)/z0 )

with u+ = u/u* (u* the friction velocity) zd = displacement thickness, in case of a meteo mast set it to zero (it doesnot matter much in that case; contrary to the influence of kappa +- 0.42)

your measurement is at z = 10 m, you have a velocity u, assume a value for z0 (3cm) => u* follows => whole profile is known. there is no reason to use a power-law profile.

Thanks ag. Ok, good explanation. The reason I want to use the log law is because I have read in a few sources that it is more accurate than the power law, for example from wikipedia, the "alpha component of the power law is an empirically derived value which depends on the stability of the atmosphere...in places where trees or structures impede the near-surface wind, the use of a constant 1/7 value for alpha may yield quite erronous estimates, and the log wind profile is preferred...even under neutral stability conditions, an exponent of 0.11 is more appropriate"

It is for a feasibility study for a wind turbine project, where I have a figures for the wind speed at 10m, 25m, and 45m AGL, from the BWEA. I would just use these to create a curve, but I'm pretty sure that they are just using the power law to give values for 2 out of the 3.

Greatly looking forward to your replies!

Hi Hendrik

ah, but u* = sqrt(wall shear stress / density), where the wall shear stress is a function of du/dy, hence du/dy is my problem?

Thank you for the time so far...please see my reply to Ag as the reason for doing this...Cheers

Are you able to use something like this?

http://i78.photobucket.com/albums/j9...ityProfile.jpg

You will need to know the roughness height (zo) as was mentioned earlier.

I believe the reference to Stull, 2000 is titled meteorology for scientists and engineers

http://www.amazon.com/Meteorology-Sc.../dp/0534372147

Yes and no... you need to start the curve/profile somehow, call it 'z0', du/dy or wall shear stress; they are linked together. A rough(er) wall has a higher shear stress, lower du/dy and a higher 'z0'. In the z0-form of the log-law you use the friction velocity u*, that's all. (ignoring the displacement thickness).

So you have data on 10, 25 and 45 m; a large empty field (grass, water) in front. Start curve fitting and find the best 'z0'. You're done.

(if you cannot estimate the surrounding of a mast, then the only option you have is to work out the rms of the fluctuating velocities; it's a measure of the roughness (in front!) of the mast. then you assume that the shear is constant in the lower part of the boundary layer etc... estimating 'z0' is easier)

[adelbaew]

A computational model for a horizontally uniform atmospheric boundary layer accounting for various options of atmospheric stability is developed on the basis of Monin-Obukhov’s similarity theory. Numerical modeling of the atmospheric boundary layer is carried out using the Fluent package. For a turbulent closure, the standard k–ε turbulence model with modified turbulence closure constants and an additional source term in the equation for k are applied. Cases of neutral, stable and unstable stratification are considered. The computational model allows preserving horizontal uniformity of vertical profiles of velocity, pressure, density, temperature and atmospheric turbulent viscosity which are set as boundary conditions at the domain’s inlet. The obtained model can be applied to modeling of propagation of hazardous gas clouds in the atmosphere over the terrestrial surfaces including surfaces filled with buildings.

http://www.fundamental-research.ru/ru/article/view?id=35083


http://www.cfd-online.com/Forums/flu...tml#post580164

[Mohamed Refaat]

Hi adelbaew
I'm simulating ABL under stable and unstable conditions.
as you mentioned, the inlet flow profiles are calculated and interpereted based on the Monin-Obukhov theory, but
I have a problem in the developing of turbulent kinetic energy K, the production is too high as a result k increases with stream-wise direction. however, the velocity is approximately developed.

So, please can help me to understand what is happening in k ?

Thanks in advance

[adelbaew]

Hi Mohamed Refaat

Originally Posted by Mohamed Refaat

Hi adelbaew
I'm simulating ABL under stable and unstable conditions.
as you mentioned, the inlet flow profiles are calculated and interpereted based on the Monin-Obukhov theory, but
I have a problem in the developing of turbulent kinetic energy K, the production is too high as a result k increases with stream-wise direction. however, the velocity is approximately developed.

So, please can help me to understand what is happening in k ?

Thanks in advance

Please look at my article -

A.I. Kuptsov,
Environmental monitoring . CFD - technology. UDF - functions.
Vestnik Kazanskogo Tekhnologicheskogo Universiteta, 20 (2015), 203 - 207.

Are you all right programmed and
note the source term for the equation k.




Also pay attention to the constant turbulence

Cμ = 0,0333; σk = 1,00; σε = 1,3;
C1ε = 1,176; C2ε = 1,92.


http://www.fundamental-research.ru/r.../view?id=35083

[Mohamed Refaat]

thanks for consideration adelbaew.
I'm using the same inlet profiles of your paper and the model constants also,but the problem is that the production of k is very high. I don't know why?????

[adelbaew]

hi Mohamed Refaat

Originally Posted by Mohamed Refaat

thanks for consideration adelbaew.
I'm using the same inlet profiles of your paper and the model constants also,but the problem is that the production of k is very high. I don't know why?????

if you used - source term for Turbulence kinetic energy?
UDF there

and On the output ( right )
border area used the boundary condition
«Pressure Outlet», demanding tasks vertical pressure profile .



In this paper we use the condition «Pressure Outlet», but the pressure is set uniform in height when the output boundary of the pressure profile is set uneven, through the use of «Boundary Profile» option (Edge profile). This profile is obtained by iteration. First, given a uniform profile, and the calculation is carried out until convergence. Profile is then written from the middle area where the pressure profile established more or less, and is used as a boundary condition at the output. Then, the boundary condition is recorded again from the middle of the field and used as the boundary condition at the output. This iterative process is repeated until, until the difference between iterations is imperceptible.


http://www.fundamental-research.ru/ru/article/view?id=35083


If this does not work - you can send your files to my e-mail with UDF and profile file - my e-mail - artpb@yandex.ru

[adelbaew]

Hi Mohamed Refaat

Originally Posted by Mohamed Refaat

thanks for consideration adelbaew.
I'm using the same inlet profiles of your paper and the model constants also,but the problem is that the production of k is very high. I don't know why?????

[adelbaew]

Have you used the source term for the turbulent kinetic energy (TKE) in UDF?

http://www.cfd-online.com/Forums/flu...y-profile.html

[adelbaew]

The difference from previous studies is that in these papers outflow boundary condition is used or «Outflow», which does not require the setting of the vertical profile of the pressure or «Pressure Outlet» with assignment of a homogeneous vertical profile height or with a profile based on the hydrostatic equation. In this paper we use the condition «Pressure Outlet», but the pressure is set uniform in height when the output boundary of the pressure profile is set uneven, through the use of «Boundary Profile» option (Edge profile). This profile is obtained by iteration. First, given a uniform profile, and the calculation is carried out until convergence. Profile is then written from the middle area where the pressure profile established more or less, and is used as a boundary condition at the output. Then, the boundary condition is recorded again from the middle of the field and used as the boundary condition at the output. This iterative process is repeated until, until the difference between iterations is imperceptible.

http://www.fundamental-research.ru/r.../view?id=35083

[adelbaew]

They send me all of your calculation files (ansys fluent or fluent) by e-mail

artpb @ yandex . ru

[Mohamed Refaat]

Thanks again for your help.

Exactly as you said, I'm using outlet pressure BC with vertical distribution of pressure, obtained as you mentioned. Also the use of the source term does not change the results.
please note that, I'm working in ABL as a first step to include wind turbine later in order to study the effect of the ABL on the performance of the turbine.
there is no problem in Neutral cases, but the problem I'm asking you appeared only when simulating Stable cases, I mean when the energy equation is required to be solved.

you get the message?