Hi @TheodorosGalanos,
In the windFile, what do you exactly do with all the wind factors for 36 directions?
Following is mentioned in the paper
These WF were used in combination with the TMY data to derive hourly wind velocities withing the area of interest.
Can you please be specific?
Do you take average of the WF for all the directions and then multiply the TMY wind velocities with that number and then write that number in the rows of the windFile?
No. You are using these WF for different directions. So when wind is coming from angle 0, the WF for that direction is used and then value of that WF * TMY wind velocity is written to the wind file. Is that correct?
Hi Devang, yes WF are used for each different wind direction.
What happens is simple. You take a wind velocity plane of your area of interest. Then inside Paraview (ideally) you create a calculator filter and simply divide the wind velocity in the plane by the inlet wind velocity of your case. That gives you a WF for each point in the area of interest.
We then take these WFs and we multiply them with the wind velocity value, only for that direction, in the weather file. So the WF for 90 degrees is multiplied with WV values during all hours with 90 degrees wind in the weather file. Doing that for each direction you run the case for generates yearly wind value schedules.
I really should upload those files in hydrashare now the method and components are more stable.
Hello @TheodorosGalanos
I’ve tried to follow along with the work you and @chris have already done and elaborated. Just need some help about how to calculate the Velocity and WF in Paraview. I tried to import the .foam file into Paraview, exported into a .csv file, yet just got the u0, u1, and u2 components for velocity, differently from the one @chris already shared before “WindFactor_1.75m0” . So standing from me not a good user of Paraview, could you provide me with the equations in Paraview to calculate the wind speed at 1.1m (as recently recommended) and then get the WF. Thanks in advance.
Cheers,
Yasser
So I’m sorry for disturbing, yet I guess I’ve found that over CFD-online. You use the equation U=sqrt(U_X^2+U_Y^2+U_Z^2) to get the magnitude (Velocity), then devide by the initial wind speed input for the Butterfly_Wind Vector component. I tried that and I think it worked with me to create two new arrays with the Velocity and WindFactors. Am I right?
I’m a little late to this conversation Just wondering if there is a way to get the wind factor (WF) values without having to manually open Paraview. I have not been using BF for a while, and I’m not sure if any new components were developed to streamline the WF calculations inside Grasshopper. Or maybe, a new command-line process that was developed?
Hi @TheodorosGalanos
While watching your (very informative) tutorial on Performance Network, I was wondering if there is any underlying study/paper/etc. regarding the Wind Factor method? It feels somehow unintuitive to assume a static dependency between incoming wind speed and the affected eddy areas. Or do you consider the inaccuracies to be negligible?
In the tutorial, you multiply the Wind Factor with the EPW Wind Speed at 1.1m height. Would it also work to take the windspeed at standard height (10m), assuming that also the Wind Factor was calculated by dividing through the windspeed at 10m? Or would that make it more inaccurate?
Anyways, thank you for creating such an awesome tool!
It’s very insightful to work with BF.
Wind factors refer to amplification factors if I’m not mistaken, might be an easier term to search From my understanding, those stay quite stable across different wind velocities (within a reasonable range, which is where most average winds are anyways). Concerning the different heights, factors can change across them but that’s alright since your simulation can give you wind velocity values at any height within your domain.
I would want you to help me clarify the Wind Factors workflow, because I am getting different results from my python workflow than from Eddy3D.
The wind factor is calculated by:
Dividing the wind velocity at the comfort point by the scaled (see formula below) inlet velocity. I would get a matrix of size (wind_dir, analysis_pts).
From the weather file get the wind direction at every hour, and approximate to match simulated directions. Eg. turn 42 deg to 45 deg. (no interpolation in this case)
Having the wind dir adjusted, look up for the correct wind factors for that direction, and multiply the factors by the scaled (?) EPW wind speed at the same height of the comfort point, and repeat for each point at each hour of the year.
I end up with a matrix of size (8760 * analysis pts).
In another question, I am using the following formula for the Logarithmic wind profile, is this right?
def scaleWind (Uinit, Zinit = 10, Zend = 1.5, z0 = 1.0, d = 0):
if Zend > z0: Uend = (Uinit * ((math.log(Zend - d /z0) / (math.log(Zinit - d /z0)))))
else: Uend = 0
return Uend
It has been more than a year since the last comment. I am currently running a UTCI simulation and for that I am using the actual hourly wind speeds for the whole year. In my case the terrain roughness and wind exposures are already considered within the simulation setup. For this case, is still necessary to apply the wind profile reduction factor? have you done this before?
Just wondering if there is a way to get the wind factor (WF) values without having to manually open Paraview. I have not been using BF for a while, and I’m not sure if any new components were developed to streamline the WF calculations inside Grasshopper. Or maybe, a new command-line process that was developed?