Parameters as function of wingspan

[Agustin Buffa]

Hi,

I would like to have the data of cd across span, like result manager shows. In which archive is saved the data?

Thanks,

Agustin

[Brandon Litherland]

Spanwise data is in the *.LOD file

[Tim Swait]

I'm looking at the .lod file. What are Cx, Cy and Cz? Or why aren't they exactly equal to Cd,Cs and Cl respectively? What's the cooridinate system here is lift not in the vertical (z) axis? For my file Cs seems to always exactly equal Cy, Cz and Cl roughly correspond, but Cd and Cx mismatch by a fair amount.

[Brandon Litherland]

Tim,

X,Y, and Z are in the body frame aligned with the XYZ of your model.  L, D, and S are in the ground/freestream frame where L is opposite gravity and D opposes forward, level flight.  S and Y are equal because they both point directly right.  Angle of attack change the relationship between Lift/Drag and Z/X.  For example, at higher alpha you'll have a significant portion of lift pointing in the negative X direction along with positive Z.

[Tim Swait]

Of course! Thanks. I guess that Cs doesn't equal Cy when there's a sideslip angle then?

For each row, I'm assuming that the force and moment coefficients are taken as acting through/about the point defined by Xavg, Yavg, Zavg? And to convert into force I use the I use the S value for each row as my reference area? Do I also use the value in the 'Chord' column as the reference length for each row for the moment coefficients? I.e. each row is self contained, I don't use any of the reference areas or lengths that are set in the analysis (and listed in the header of the file)? Is this correct?

[Brandon Litherland]

Rob posted about this before in detail here: https://groups.google.com/g/openvsp/c/SZDNiXALN68/m/BM2lku8sDAAJ

Let us know if this doesn't quite get you what you need.

[Agustin Buffa]

Hi Brandon,

Solved!

Thank you very much,

Agustin.

[Tim Swait]

Thanks for that, I think I get the idea for the force coefficients. For my purpose, I'm trying to get a value and location for the resultant force at each spanwise location, and I'm interested in the frame of reference of the aircraft, not the global. So if I want to calculate the force component in the z direction then I multiply  cz by q*chord*length. I'm assuming that I can get "length" from the Yavg values; the length for the nth point is the distance from halfway between the nth and the (n-1)th Yavg to halfway between the nth and the (n+1)th Yavg. Is that all correct?

However, if I want to find the point through which this force acts then I need to also consider the moment coefficients. I'm assuming that the moment coefficients are using the point (Xavg, Yavg, Zavg) as their reference point, is this correct? What is used as the reference length and reference area in calculating these cm values? Does it use the overall aircraft reference chord as the reference length or the chord value at each station? Is area converted into a per unit length length value as for the force coefficients. Are the 3 axes using the same values?

[Brandon Litherland]

You've got the basic idea.  Each of the Y stations are the center of the spanwise panels displayed in VSP e.g., all of your NumU sections. You _could_ do some relatively simple calculations to compute the local section width and use that for your length value but if you break thinks up into enough trapezoidal sections, it's close enough to just go with deltas (except in the case of VERY weird chord distributions).  Note that you will end up with total NumU - 1 segments so however you want to handle that missing chunk of wing is up to you.

The moments are dependent on their respective lever arms.  There are a few posts in the Group that cover this but I'll list them here again.  Moments are referenced to the defined location as you describe.  "cmy" would use the local chord for each station, "cmy*c/cref" is a chord-normalized value.

CMx = Mx/(q*Sref*bref)

CMy = My/(q*Sref*cref)

CMz = Mz/(q*Sref*bref)

[Tim Swait]

Thank you, that's very helpful, I think I get it now.

[Tim Swait]

Hi Brandon, are you sure that each value for moment (cmx, cmy, cmz)is referenced from the point (Xavg, Yavg, Zavg) as I put in my first post? Now I've played around with it more, I'm pretty sure that they are all referenced from the same overall reference point. I have been playing with a simple plank wing to make sure I understand what's going on, it's 100m span x 1m chord. If I change the Moment Reference position for the aircraft then the cm values all change. Also if you look at the cmx values along the semi-span then these only make sense if they are relative to the aircraft reference position, not to their individual reference positions; the lift from the section of wing at the tip is creating a very large moment to the aircraft centre, but shouldn't have any different moment to any other relative to the section centre.

[Brandon Litherland]

That was my error in reading your post.  I was in a rush and interpreted your avg location as the reference location.  In your VSPAERO setup, the Moment Reference Position region sets the location where all moments are calculated from.  The XYZavg points are (I think) the centroid of the spanwise section.  The moments are about the Xref, Yref, and Zref values in the solver.

[Tim Swait]

Yes, this seems to be making sense. XYZavg points do indeed seem to be the centroids of each spanwise section.

Also it seems that cmx, cmy and cmz are relative to the global coordinate system, not the aircraft local xyz system (rotated for alpha). I'm trying to find the point through which the force can be taken to act for each spanwise section. I'm calculating the x position with ('cmy*c/cref')/'cl' and the y position with ('cmx*c/cref')/'cl'. This seems to give sensible results, the x values are at around 30ish% chord back from the LE and the Y position is fairly well matched to the Yavg value. However if I use cz instead of cl in the above calculations then I get very unbelievable looking values (sometimes putting the point of action of the force outside of the spanwise section that it's referring to! This is a little confusing, I had expected cmx, cmy and cmz to be the moments around the x,y,z directions but it looks like they're actually the moments around the cd, cs and cl directions.

[Tim Swait]

Actually, I do get sort of reasonable values with using cz instead of cl too, but neither looks completely convincing! If I use cz in the calculation, then the y positions of the lift centre match very well with the Yavg values, so look very believable. However the x values near the tip look too far back. So if I calculate the x position with ('cmy*c/cref')/'cz' and the y position with ('cmx*c/cref')/'cz', then plotted it looks like this in planform, the red line is trailing edge of wing, purple line is leading edge, the top point on each vertical line is the position I've calculated as my centre of each panel:

As said, the centres all look good in y, they're lining up close to the middle of each panel. the x positions look good near the root. The aerofoil section is UI1720, so it has a very forward camber and also it's reflexed, so I would expect the lift centre to be fairly far forward, as it is. However near the tip it doesn't look so believable. The wing has a lot of twist and no reflex at the tips, so I'm not surprised that the centres move back, but that last point is so far back that it's off the trailing edge. I don't see that that's physically possible, or is it?

Am I doing the right thing with my calculation here? Also I came up with using the normalised (*c/cref') version of the moment but the not normalised version of the force a bit by trial and error as it was the only way I got remotely sensible numbers. If I use normalised for both (or if I use not normalised for both) then I get completely crazy numbers, way off the wing!