Question Regarding Rel and Abs Dihedral

[Juan]

Hello everyone, 

Forgive me if this has been asked before, I used the search bar on Google Groups to find info about this but couldn't find much. 

I was playing around a multi-section section wing and its dihedral angle and noticed something that caught my attention. I understand that the the absolute flag sets the angle wrt the absolute coordinate system whereas the relative flag is wrt to the previous wing section (or a delta-angle as mention in the Ground School videos). With that in mind, if I have a simple multi-section wing with an arbitrary dihedral angle, I should get the same geometry with either of the following characteristics, right? 

1. Absolute flag on - all wing sections with the same dihedral angle as section 1

or

2. Relative flag on - wing section 1 with desired dihedral angle and all n+1 sections with dihedral angle = 0

I noticed that if I create two wings as described above, with all other parameters equal (including airfoils), the projected span is not the same and the dihedral angle appears to be slightly different. A wing with relative dihedral on appears to have a slightly larger projected wingspan compared to the absolute dihedral wing. I'm attaching a simple .vsp3 model to show what I'm talking about. This is simply an example geometry of 2 equal wings.

Is my understanding of the way OpenVSP calculates dihedral correct? Is there any reason why the two methods described above don't yield the same geometry? and what would be a recommended practice when setting Dihedral for a multi-section wing?

Thank you!

Juan

[Rob McDonald]

Thanks for pointing that out.  Not sure why it is doing that.

From what I can tell, the geometries are actually identical.  The projected span is not actually calculated by projecting the span -- it is calculated from trig here...

https://github.com/OpenVSP/OpenVSP/blob/main/src/geom_core/WingGeom.cpp#L2393

It looks to me like the above code is not intelligent enough to handle rel vs abs properly.

The only things based on this are the calculation of the projected span and the overall aspect ratio.  Everything else should be identical.

For humans, using the relative dihedral is usually easiest and most intuitive.  In your example where the wing has a constant dihedral, you only have to adjust one place and everything updates.  If you had a wing with five panels, the inboard three at one dihedral, the outboard two at another, you only have to adjust two settings -- probably in an intuitive (relative) way.

On the other hand, if you are trying to duplicate a model that someone has built in cad -- or they have used an optimization algorithm to design the polyhedral -- those tools are likely to specify the dihedral in a global coordinate system for every section -- easy for computers.

So, I would use whatever makes sense in your head (how do you think about your problem).  Unless you are interfacing with another tool that gives you a compelling reason to try to be compatible with that tool.

Rob

--
You received this message because you are subscribed to the Google Groups "OpenVSP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openvsp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openvsp/86382df4-484e-4a83-bdb2-c2147d14d5ddn%40googlegroups.com.

[Juan]

Hi Rob,

Thank you so much for the info and recommendation. I agree that using the relative option is the most intuitive way of setting dihedral. I wanted to look at the code you sent before getting back to you to learn more about the inner workings of vsp, and maybe find something about the proj span. The proj span calculations makes sense to me. I found that on line 2588 https://github.com/OpenVSP/OpenVSP/blob/9b64795e98fe1ee489ca279d16804b9b12c663f2/src/geom_core/WingGeom.cpp#L2588 there is an if statement that returns m_Dihedral for when the relative flag is off. I don't know if that has any influence on the proj span calculation. I went back to OpenVSP and found something that I think isolates where the issue may be coming from. It's a little hard to explain but I'll do my best:

If I add a WingGeom, which by default has only 1 wing section, and I first give it an arbitrary dihedral angle and then split the wing, the projected span is the same with either Rel or Abs dihedral on. However, if I split the wing first and then assign a dihedral angle, the projected span is different depending on the Rel or Abs dihedral flag. I was not able to replicate this by using the insert wing section (projected span is different in this case regardless of which step is performed first). It seems like split and insert influence how proj span is calculated? 

Another smaller thing I noticed was that on 3.22-3.24 the Total Planform information does not update when XZ symmetry flag is turned on/off. This does happen in version 3.21 and earlier. Not sure if that was intentional but wanted to point that out too. 

I hope this info helps!

Best, 

Juan

[jvgr...@gmail.com]

Juan,

I went ahead and fixed the projected span and AR calculations for wings with dihedral specified as relative vs absolute. Thank you for reporting this issue. 

The total planform parameters not updating when symmetry is turned on or off is in fact a bug. If you force an update, such as by changing the wing tip cap, the total planform parameters will update. I'll make sure this also gets fixed for the next release. 

Thank you,

Justin 

[Juan]

Hi Justin, 

No problem! I'm happy to have been able to bring this up and I hope my input helped. Thank you for the info and updates to the projected span calculation. I appreciate all the hard work you guys put into OpenVSP; it's an amazing program.

Thank you, 

Juan

To view this discussion on the web visit https://groups.google.com/d/msgid/openvsp/a2aceeff-8197-47dc-9118-c68b7abb660cn%40googlegroups.com.