VSPAero QNAN Problem

715 views
Skip to first unread message

Nicolas Gonthier

unread,
Jun 4, 2015, 5:07:54 AM6/4/15
to ope...@googlegroups.com
Hello people,

I have some problems with VSPAero when I try to run the solver on my plane (a fuselage plus a wing), it always stop operating.

The error is :

Loop: 24828 on body surface: 6 has zero diagonal... Mach is: 0.200000 
Wake Iteration:     1 ... GMRES Iteration:     0 ... Reduction:    1.#QNAN ...  Maximum:    1.#QNAN

What can I do to settle this problem ?

All the best,

Nicolas  

Rob McDonald

unread,
Jun 4, 2015, 11:14:32 AM6/4/15
to ope...@googlegroups.com
Maybe Dave has some ideas.

In the meantime, can you provide more information? Did you start with
a simple example and work towards more complex? Did a simple example
work?

Is there anything unusual about your model or the geometry?

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.
Message has been deleted

Nicolas Gonthier

unread,
Jun 5, 2015, 5:10:50 AM6/5/15
to ope...@googlegroups.com
I find that the problem is about the fuselage, and the fact that I have rounded rectangle section. 
Without this section I don't have any problem with VSPAero.

I join the XML file.

Thanks

Nicolas
FuselageTest.vsp3

Rob McDonald

unread,
Jun 5, 2015, 10:07:54 AM6/5/15
to ope...@googlegroups.com
On Fri, Jun 5, 2015 at 2:10 AM, Nicolas Gonthier <nica...@gmail.com> wrote:
> I find that the problem is about the fuselage, and the fact that I have
> rounded rectangle section.
> Without this section I don't have any problem with VSPAero.

I understand how rounded rectangles could cause a problem -- I'll talk
with Dave and see if we can come up with a fix.

In the meantime, just switch those sections to ellipses. That will
actually do a better job of representing what you want to VSPAERO.
Since it is a thin surface code, the geometry comes through as smashed
down sections. In that case, the ellipse is the exact same thing.

Rob

Nicolas Gonthier

unread,
Jun 5, 2015, 10:23:12 AM6/5/15
to ope...@googlegroups.com
Yes, you are right. I will do use VSPAero with ellipse instead of  rounded rectangle.

B-ROB

unread,
Jun 10, 2015, 10:07:09 PM6/10/15
to ope...@googlegroups.com
Yes I noticed the exact same problem.  I took the "transport.vsp" file from the VSP hangar and tried to run...same issue.  Won't work.

Rob McDonald

unread,
Jun 10, 2015, 10:27:15 PM6/10/15
to ope...@googlegroups.com
Brent,

Were you trying to model the Transport with all the components -- when
I tried it in VSPAERO, I believe I used Sets to not pass the nacelles
and pylons to VSPAERO. You might try that to see if it helps you make
some progress.

Rob

Rob McDonald

unread,
Jun 10, 2015, 10:28:03 PM6/10/15
to ope...@googlegroups.com
Most importantly, the nacelles and the puck component held inside them
-- that component really doesn't make sense for VSPAERO. They will
cause problems when converted to a DegenGeom.

Rob

Brent Robbins

unread,
Jun 10, 2015, 11:09:15 PM6/10/15
to ope...@googlegroups.com
Rob,

I "cut" everything out except "Fuselage", "HTail", "VTail", "GearPod" & "Wing".  vspaero.exe stopped working.  You got it to work?
 

From: Rob McDonald <rob.a.m...@gmail.com>
To: ope...@googlegroups.com
Sent: Wednesday, June 10, 2015 7:28 PM
Subject: Re: [OpenVSP] VSPAero QNAN Problem
>> email to openvsp+unsub...@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+unsub...@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages