VSPAERO help
391 views
Skip to first unread message
Emmaline McKingley
Oct 21, 2015, 2:53:12 AM10/21/15
to OpenVSP
Hi,
I'm a very new user of OpenVSP.
I'm a very new user of OpenVSP.
I'm working on replicating the Fouga CM. 170 Magister in the program and would like to be able to analyse it in VSPAERO.
Since the Magister's shape is a bit complicated I might have used slightly unconventional methods for replicating its shape (as I said, I am a new user and therefore do not always understand the implications of the choices I made while constructing the model).
When I try analyzing with VSPAERO, it always gets stuck with the only execute option being "Kill Solver". I don't even get any error messages, no matter how long I leave it that's how it stays.
I tried analyzing individual geometries to see if maybe the problem was coming from one specific one, however, I found that when I do this, VSPAERO only works on the wing geometries by themselves (or wings together).
I understand that VSPAERO can't process certain shapes. however from my understanding it should work for a fuselage. However it did not work for mine (same thing happened as described above with the "kill Solver").
I've seen from other posts what could be causing problems and have tried fixing them for mine (such as cross sections with the same x location), with no luck.
I have hit a wall and don't know how to proceed. Even if it means remodeling certain parts specifically for VSPAERO or only being able to analyze certain parts, I would greatly appreciate any help or suggestions on how to make this work.
Since the Magister's shape is a bit complicated I might have used slightly unconventional methods for replicating its shape (as I said, I am a new user and therefore do not always understand the implications of the choices I made while constructing the model).
When I try analyzing with VSPAERO, it always gets stuck with the only execute option being "Kill Solver". I don't even get any error messages, no matter how long I leave it that's how it stays.
I tried analyzing individual geometries to see if maybe the problem was coming from one specific one, however, I found that when I do this, VSPAERO only works on the wing geometries by themselves (or wings together).
I understand that VSPAERO can't process certain shapes. however from my understanding it should work for a fuselage. However it did not work for mine (same thing happened as described above with the "kill Solver").
I've seen from other posts what could be causing problems and have tried fixing them for mine (such as cross sections with the same x location), with no luck.
I have hit a wall and don't know how to proceed. Even if it means remodeling certain parts specifically for VSPAERO or only being able to analyze certain parts, I would greatly appreciate any help or suggestions on how to make this work.
I have attached the file that I've been working on. And I am using the latest version of OpenVSP (3.3.0) on PC.
Thank you,
Emmaline
Emmaline
Rob McDonald
Oct 21, 2015, 12:03:44 PM10/21/15
to ope...@googlegroups.com
Emmaline,
That is a really nice model for your first attempt at VSP. Well done.
For most purposes, you've approached this model the way I would have
-- separate components for the canopy, engines, tip-tanks. Wing
component that goes all the way to the centerline, etc.
This approach gives you the best flexibility when you go to
investigate a design change (what if the canopy was shortened to
1-person, easy to do).
You specified more cross sections for your components than I would
have. For example, your fuselage is made of 12 cross sections (point,
a bunch of ellipses, and an ending point). In VSP, less is more. I
would have started with just three (point, ellipse, point), with the
ellipse at the maximum cross section location. Then, I would have
used the skinning parameters to try to get it as close as I could. If
I could not get it close enough for my satisfaction, then I would add
an additional cross section near the area of maximum error. I would
continue this process until I was happy with the shape. Often, a body
will have some more critical features -- a sharp corner, an inflection
point in curvature, etc. I will generally pick those up with a fixed
cross section right from the start.
You can change the name of each component at the top of the Geom
Browser -- that way you can know which Fuse is the real Fuse and which
Wing is really the V-Tail.
Unfortunately, working with a thin surface aerodynamics code (like
VSPAERO) can be a bit tricky. VSP makes no attempt to modify or trim
the geometry components for you, so you have to make sure things are
ready to go from the start.
One thing VSPAERO doesn't like is when two sets of panels exactly
overlap. Unfortunatley, this happens a bunch in your model -- the
horizontal panels where the engine, fuse, and wings come together and
also the vertical panels where the fuse and canopy come together.
The best way to approach this depends on what kind of analysis you're doing.
For example, if you're doing pitch analysis (alpha sweep, induced
drag, pitch derivatives), then you'll have a symmetrical flow
throughout, and I would not expect the canopy to have a significant
impact on the solution. So, I would use VSP's Set capability to
create a VSPAERO_Pitch set and exclude the canopy from that set.
On the other hand, if you're doing asymmetrical yaw analysis, I would
go back (against my above advice) and build a single-component
fuse/canopy that matched the silhouette of the aircraft -- I wouldn't
worry much about the width or details of the cross section.
To fix the wing/fuselage junction, I would probably drop the engines
from the model and then either shift the wing up/down, or model the
root wing section starting at the side of fuselage junction.
I hope this helps you get started,
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.
That is a really nice model for your first attempt at VSP. Well done.
For most purposes, you've approached this model the way I would have
-- separate components for the canopy, engines, tip-tanks. Wing
component that goes all the way to the centerline, etc.
This approach gives you the best flexibility when you go to
investigate a design change (what if the canopy was shortened to
1-person, easy to do).
You specified more cross sections for your components than I would
have. For example, your fuselage is made of 12 cross sections (point,
a bunch of ellipses, and an ending point). In VSP, less is more. I
would have started with just three (point, ellipse, point), with the
ellipse at the maximum cross section location. Then, I would have
used the skinning parameters to try to get it as close as I could. If
I could not get it close enough for my satisfaction, then I would add
an additional cross section near the area of maximum error. I would
continue this process until I was happy with the shape. Often, a body
will have some more critical features -- a sharp corner, an inflection
point in curvature, etc. I will generally pick those up with a fixed
cross section right from the start.
You can change the name of each component at the top of the Geom
Browser -- that way you can know which Fuse is the real Fuse and which
Wing is really the V-Tail.
Unfortunately, working with a thin surface aerodynamics code (like
VSPAERO) can be a bit tricky. VSP makes no attempt to modify or trim
the geometry components for you, so you have to make sure things are
ready to go from the start.
One thing VSPAERO doesn't like is when two sets of panels exactly
overlap. Unfortunatley, this happens a bunch in your model -- the
horizontal panels where the engine, fuse, and wings come together and
also the vertical panels where the fuse and canopy come together.
The best way to approach this depends on what kind of analysis you're doing.
For example, if you're doing pitch analysis (alpha sweep, induced
drag, pitch derivatives), then you'll have a symmetrical flow
throughout, and I would not expect the canopy to have a significant
impact on the solution. So, I would use VSP's Set capability to
create a VSPAERO_Pitch set and exclude the canopy from that set.
On the other hand, if you're doing asymmetrical yaw analysis, I would
go back (against my above advice) and build a single-component
fuse/canopy that matched the silhouette of the aircraft -- I wouldn't
worry much about the width or details of the cross section.
To fix the wing/fuselage junction, I would probably drop the engines
from the model and then either shift the wing up/down, or model the
root wing section starting at the side of fuselage junction.
I hope this helps you get started,
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.
Reply all
Reply to author
Forward
0 new messages