SOWFA, HubRad and TipRad problems

[Bruno.Nguyen]

Hello everyone,
I am currently using SOWFA to study tidal turbine wakes and, while trying to calibrate my model, I stumbled upon a problem with the HubRad scalar.
Basically, I have a hub radius of 0.046 m and a blade tip radius of 0.35 m so I expect the body force to go from 0.046 m to 0.35 m along the “fake” radius of the blade.
However, when I look at the velocity on the plane where the body forces are injected, I see that the velocity is only decreased from approximately twice the hub radius.
I tried to play around the values of the hub and tip radius but I did not manage to fix this problem.
However, I looked into the code and I found something which seemed peculiar to me in the horizontalAxisWindTurbinesALM.C: (between line 554 and 580)

// Add other point distribution types here, such as cosine, tanh.

    // Now calculate the actuator section center points for each blade
    // of each turbine in the array.  All blades points will be calculated
    // at zero azimuth (blade pointing up), and then rotated to its correct
    // position before doing a global rotation to the initial azimuth of
    // the rotor.  Also calculate the radius of each point (not including coning).
    bladePoints.append(List<List<vector> >(NumBl[j], List<vector>(
	[i],vector::zero)));
    bladeRadius.append(List<List<scalar> >(NumBl[j], List<scalar>(numBladePoints[i],0.0)));
    for(int k = 0; k < NumBl[j]; k++)
    {
        vector root = rotorApex[i];
        scalar beta = PreCone[j][k] - ShftTilt[j];
        root.x() = root.x() + HubRad[j]*Foam::sin(beta);
        root.z() = root.z() + HubRad[j]*Foam::cos(beta);
        scalar dist = HubRad[j];
        for(int m = 0; m < numBladePoints[i]; m++)
        {
           dist = dist + 0.5*db[i][m];
           bladePoints[i][k][m].x() = root.x() + dist*Foam::sin(beta);
           bladePoints[i][k][m].y() = root.y();
           bladePoints[i][k][m].z() = root.z() + dist*Foam::cos(beta);
           bladeRadius[i][k][m] = dist;
           totBladePoints++;
           dist = dist + 0.5*db[i][m];
        }

This routine computes the actuator center points along the “fake” blade. I do not understand why the x and z components of the bladePoints vector gets twice the “HubRad” value. Indeed, HubRad is contained into root.x() ( respectively root.z() ) as well as in dist.
Hence, if I understood correctly, the body forces would not be injected between HubRad and TipRad but rather between 2 HubRad and TipRad+HubRad, which seems wrong to me.

I am not sure of my understanding of the code, thus I would like to know your input about it.
Thanks in advance,

Regards,

Bruno.