Non linear CMy vs CLt graph

67 views
Skip to first unread message

Arshad Basha

unread,
Mar 28, 2026, 9:44:57 AM (8 days ago) Mar 28
to OpenVSP
hey! i recently started using openvsp and ran into this issue. 
i am trying to simulate an ultralight called x air falcon and the results have given non linear graphs additionally indicating the aircraft being statically neutral in pitch. 

Here i am posting some pictures of what i got. Any help would be invaluable. Thankyou for you time.

Cmy vs CL.pngDisk.pngFV.png
V-27_aoa-0.png
ISO.png
SV.png
Overview.png
TV.png

Rob McDonald

unread,
Mar 29, 2026, 2:59:59 PM (7 days ago) Mar 29
to OpenVSP
Notice the hot spot on the wing right above the fuselage.  That is likely the problem. You should either bury the root of the wing inside the fuselage, or have some gap between the wing and the fuselage.

You also likely have too much chordwise resolution on the wings.  Also, you should use less LE / TE clustering when you use a thin representation for the wings.  I.e. change the cluster values to be closer to 1.0.

Rob

Arshad Basha

unread,
Mar 29, 2026, 3:29:08 PM (7 days ago) Mar 29
to ope...@googlegroups.com
Thanks a lot for your response. I'll try to correct the model with the suggested changes. 

However, I ran an analysis without the fuselage and an xflr5 analysis before starting open vsp. Both, including rudimentary hand Calc suggested statically neutral behavior in pitch. But a pilot and a manual I found online on the aicraft said the aircraft was pretty stable. I have decided on some cfd to collect moment data as the final thing I can do on my own. Not exactly sure on that either.

Sorry this might be a little unrelated. But if you have any advice I would appreciate it.

--
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 visit https://groups.google.com/d/msgid/openvsp/42132b25-3fb1-414a-a5e3-12f29361af5cn%40googlegroups.com.

Rob McDonald

unread,
Mar 29, 2026, 3:31:33 PM (7 days ago) Mar 29
to OpenVSP
Where was the CG?

You can make pretty much anything stable or unstable by moving the CG.

In codes like this, the CG is the reference point for taking moments.  You can have it anywhere -- and shift it later.  Or you can put it where you expect it to be and the coefficients will have intuitive behavior from the start.

In most codes, the CG is at (0,0,0) by default.  Did you move it?  Is that where you want it?

Rob


Arshad Basha

unread,
Mar 30, 2026, 12:49:53 AM (7 days ago) Mar 30
to ope...@googlegroups.com
Thanks for getting back to me. The CG datum and static margin were given in the British microlight association's report. Right now, it's above the rear landing gear, which makes sense. I also made sure that the CG was set to open vsp's origin. The MOI looked ok too. 

At this point, I'm unsure how to conclude anything short of getting the aircraft and flying it. 




Thanks for your help. I'll look into what you suggested regarding the non linear data and get back to you.

Rob McDonald

unread,
Mar 30, 2026, 12:54:06 AM (7 days ago) Mar 30
to OpenVSP
Perhaps I misunderstood what you're saying, but if the CG is directly above the landing gear, then that could be your problem.

The CG limits are clearly 4.5"-10" in front of the datum -- the datum is the axle of the gear.

Moving the CG 7.25" forward (from the axle to the middle of the range) will add a tremendous amount of stability to this aircraft.

It will also keep it from falling over on the tail in a slight breeze.


Including the root of the wing in the analysis (by dropping the fuselage) will shrink the effective tail area.  I.e. destabilize the aircraft.  It won't be a huge effect, but if you're neutral now, then  you don't need much to be stable.

Rob





Arshad Basha

unread,
Mar 30, 2026, 9:28:04 AM (6 days ago) Mar 30
to ope...@googlegroups.com
you are probably right, i should have seen it says fwd. I thought the fwd of CG aft would go toward the tail. :)
As the pilots and fuel were around the datum point, I didn't suspect anything. NTM the CG kept sitting on top of the LG naturally after I loaded various components. Though that is probably due to the imprecise nature of loading in this model.
Thanks. 
Here are some iterations and fixing the CG. I also moved the model to coincide CG and origin for more accurate MOI. Not sure if thats how that works but the long.static stability graph is more stable now.


TV2.png
FV2.png
Cmy vs CL (2).png
ISO_2_shaded.png
SV2.png
Cmy vs CL (CG coincided with origin for more accurate MOI- more stable).png

Rob McDonald

unread,
Mar 30, 2026, 11:55:43 AM (6 days ago) Mar 30
to OpenVSP
The VSPAERO calculation doesn't know anything about the mass properties of the aircraft.  It doesn't know anything about inertias.

It really doesn't know anything about the CG -- what we call the CG is really just the reference point for the moments.  We could use any reference point and translate to any other, but it is convenient to use the CG.

OpenVSP has a mass properties calculation that can calculate the moments of inertia.  However, it is a bottoms up calculation, so you would need to include a lot of detail about the aircraft and its construction to get an accurate result.  Even if you did everything to calculate accurate inertias, that would mean nothing to the VSPAERO calculation.

Finally, the calculation of inertias of an aircraft does not affect its static stability.  It will affect its dynamics, but it doesn't matter for anything static.

Rob

Arshad Basha

unread,
Mar 30, 2026, 3:28:38 PM (6 days ago) Mar 30
to ope...@googlegroups.com
I see. That makes sense. Time to reiterate. 

Thanks for the clarification and your patience. It has meant a lot to me.

Arshad Basha

unread,
Mar 31, 2026, 6:13:14 AM (5 days ago) Mar 31
to ope...@googlegroups.com
New results. I removed masses and just assigned CG instead of trying to get CG through component masses. I will be trying the latter as well, but as you suggested i had to bury the wing completely into the fuselage. Keeping the wing out wasnt working either. 
Question: If i am trying to model a pylon wing, would this problem still arise?

On Tue, Mar 31, 2026 at 12:42 PM Arshad Basha <arshad...@gmail.com> wrote:
I ran a simulation with the aircraft unloaded (without mass) i got different CM/CL curve than when aircraft is loaded, this made sense as Cm/CL depends on the Xcg location which is taken pretty forward (at he LE of the wing) when running the solver without assigning mass. Even the equation of static margin is dependent on Cm/CL which is dependent on how the Xcg comes out. As static stability in pitch depends on Cm/CL slope (slope being negative) which inturn depends on Xcg, at a constant tail length, the loading will determine the static pitch stability which also determines inertia. Making a proper inertia loading necessary for proper static stability in pitch values. 

This atleast has been my thinking until now. If i am going wrong somewhere or you think i am missing something, do let me know. 

Thanks

Full.webp
CMy vs CLtot.webp

Arshad Basha

unread,
Mar 31, 2026, 6:13:23 AM (5 days ago) Mar 31
to ope...@googlegroups.com
I ran a simulation with the aircraft unloaded (without mass) i got different CM/CL curve than when aircraft is loaded, this made sense as Cm/CL depends on the Xcg location which is taken pretty forward (at he LE of the wing) when running the solver without assigning mass. Even the equation of static margin is dependent on Cm/CL which is dependent on how the Xcg comes out. As static stability in pitch depends on Cm/CL slope (slope being negative) which inturn depends on Xcg, at a constant tail length, the loading will determine the static pitch stability which also determines inertia. Making a proper inertia loading necessary for proper static stability in pitch values. 

This atleast has been my thinking until now. If i am going wrong somewhere or you think i am missing something, do let me know. 

Thanks

On Mon, Mar 30, 2026 at 9:25 PM Rob McDonald <rob.a.m...@gmail.com> wrote:

Rob McDonald

unread,
Mar 31, 2026, 11:03:12 AM (5 days ago) Mar 31
to OpenVSP
A Pylon wing would be designed to get the wing above the fuselage enough to get it away from most aerodynamic interference effects.  In that situation, the VSPAERO calculation would be fine.

Rob

Rob McDonald

unread,
Mar 31, 2026, 11:05:59 AM (5 days ago) Mar 31
to OpenVSP
You know the CG range of your aircraft -- from the page of the POH you posted.

Do not use the 'Calculate CG' button in OpenVSP.  Just use the CG range provided.  Check the most forward and most aft locations.

Also, you should look at a book on static stability such that you understand how the calculations are performed and how you can translate the reference point from one place to another.  There really is no need to run VSPAERO twice for this stuff, you can run it once and figure out how the Cm/CL curves change with CG location.

Finally, stop saying anything about inertia in this conversation, it has nothing to do with anything here.

Rob

Arshad Basha

unread,
Mar 31, 2026, 11:38:07 AM (5 days ago) Mar 31
to ope...@googlegroups.com
Thanks for your suggestions. 

Arshad Basha

unread,
Mar 31, 2026, 11:38:13 AM (5 days ago) Mar 31
to ope...@googlegroups.com
I see. Thanks for your insight. 

Reply all
Reply to author
Forward
0 new messages