How is the dihedral is applied in OpenVSP?

[Ehiremen Nathaniel Omoarebun]

Hello, 

I would like to know how the dihedral is applied to a model in OpenVSP. Is it the angle between the line joining the quarter-chord from the root to the tip with respect to the lateral axis?

Thanks in advance,

Nathaniel

[Rob McDonald]

Though the process is more complex (it used to be this simple), you
can think of constructing a wing as starting with two airfoil sections
-- the root and tip section.

At the beginning, both are located at the origin, both are
not-rotated, not-scaled (chord=1), etc.

To build a wing, you scale, rotate, and translate the two sections
into their final states -- the wing is then constructed by connecting
the two together into a surface.

The x-axis runs from nose to tail, which is expected to be aligned
with the chordwise direction.

Dihedral is a rotation of the tip airfoil about the x-axis. Using the
'rotate airfoil to match dihedral' complicates this a bit, but you
should get the idea.

So, dihedral doesn't care about which x/c line you reference it to.
If your wing thickness tapers geometrically, the top/bottom surface
will have slightly different apparent dihedral than the value you
specified.

OpenVSP performs these transformations assuming your initial airfoil
has its leading edge at (0,0) and its trailing edge at (1,0). If this
is true for your root and tip foils, then you should see the LE and TE
fall along the nominal dihedral angle. If this is not true for your
foils, things will be off. Twist will also throw this off.


You've asked a bunch of questions like this (some in private), but you
haven't talked about what you are actually trying to do.

What actual problem are you trying to solve? Are you trying to
duplicate OpenVSP's wing? Trying to reconstruct in CAD? Something
else? There may be an easier/better way.

If you are really looking for a definitive description of exactly how
the wing is constructed, it is very complicated (especially with
blending) and you're going to be better off trying to understand the
code.

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.
> For more options, visit https://groups.google.com/d/optout.

[Ehiremen Nathaniel Omoarebun]

Thanks a lot for that. I am trying to match a complicated CAD polyhedral wing geometry with an OpenVSP model.

Nathaniel

[Rob McDonald]

Have you tried exporting a STL file form CAD and Importing that into OpenVSP?

While it isn't the final answer, it should provide a guide that helps
you adjust the OpenVSP model to match.

Knowing the exact definitions of OpenVSPs' wing definition isn't a lot
of help if you don't know exactly how the wing was constructed in CAD.
If the CAD wing has multiple sections, each with twist, taper, and
dihedral, it can be very tricky to try to measure the individual
parameters that result in its shape.

Best idea is to import a STL file, and then adjust to match. Use
FitModel if you have to.

Rob

[Ehiremen Nathaniel Omoarebun]

Thanks for your response. We have tried converting the CAD to STL and matching it with the VSP model, and have gotten a really good approximation. We were just trying to get a very accurate geometry which I do not think is feasible as we don't fully know exactly how the wing was built. With your input, I believe what we have so far is good enough. Thanks a lot for your contribution and help, and I appreciate it.   

Nathaniel

> email to openvsp+unsubscribe@googlegroups.com.

> For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to a topic in the Google Groups "OpenVSP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openvsp/WDtfmDHxLyU/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openvsp+unsubscribe@googlegroups.com.

[Rob McDonald]

If you're already that far along, then I would suggest you look into
FitModel as the next step.

Without knowing exactly how the original was built in CAD, you can't
figure out how to reconstruct it in another program. Even if you do
know, doing the math to convert one order of transformations to
another is a real pain. I don't recommend it.

FitModel allows you to match (in a least-squares sense) a OpenVSP
model to a cloud of points (can be extracted from a STL import).

It is designed to be an interactive process where you match a few
parameters at a time -- similar to how you would manually match a
model to a STL file.

There is a tutorial video here:

https://ac.arc.nasa.gov/p2ay28mwnl4/

Which references these slides:

https://nari.arc.nasa.gov/sites/default/files/attachments/Fit%20Model.pdf

Rob

On Tue, Jul 11, 2017 at 9:44 AM, Ehiremen Nathaniel Omoarebun

>> > email to openvsp+u...@googlegroups.com.

>> > For more options, visit https://groups.google.com/d/optout.
>>
>> --
>> You received this message because you are subscribed to a topic in the
>> Google Groups "OpenVSP" group.
>> To unsubscribe from this topic, visit
>> https://groups.google.com/d/topic/openvsp/WDtfmDHxLyU/unsubscribe.
>> To unsubscribe from this group and all its topics, send an email to

>> openvsp+u...@googlegroups.com.

>> For more options, visit https://groups.google.com/d/optout.
>
>
> --

> 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.