Some questions about VSPAERO

[Gordon Yuhim Lai]

Hi all,

Recently I have been exploring the VSPAERO and I have a number of questions that I would like to be clarified.

1. In the result viewer, VSPAERO shows the contour of Delta-Cp, is there a way to view the actual Cp distribution or relate the Delta-Cp to actual Cp from the analysis results?

2. At the config page, after I have pressed "create new setup", can I still alter the command input on the right side before launching the solver?

3. In the stability analysis, is there any way to assign the amount of alpha/ beta changed and the deflection angle of control surfaces. Also, is the analysis (not only the stability option but VSPAERO as a whole) capable of modeling changes brought by deploying simple flaps, slats, spoilers or other control surfaces?

4. Are the angular momentum brought by the rotors (actuator disks) taken into account? Will it affect the results by adding yawing/rolling moment?

5. Is there any way to obtain the total Cm of x, y, z axis, and Cp, AC positions of the whole aircraft model? And where could I find the stability derivatives?

Any help would be appreciated. Many thanks.

Gordon

[Rob McDonald]

On Tuesday, February 7, 2017 at 9:33:24 PM UTC-8, Gordon Yuhim Lai wrote:

Recently I have been exploring the VSPAERO and I have a number of questions that I would like to be clarified.

Welcome to OpenVSP and VSPAERO!

 

1. In the result viewer, VSPAERO shows the contour of Delta-Cp, is there a way to view the actual Cp distribution or relate the Delta-Cp to actual Cp from the analysis results?

In thin-surface (Vortex lattice method) mode, VSPAERO can only calculate a Delta-Cp.  The VLM calculates the Cp change from top to bottom surface directly.

In thick-surface (Panel method) mode, VSPAERO actually does calculate Cp -- even though the viewer says it is showing you Delta-Cp, it is really showing you Cp.

 

2. At the config page, after I have pressed "create new setup", can I still alter the command input on the right side before launching the solver?

Yes, you can use the window on the right as a text editor to change any settings.  'Create new setup' is really there to jump-start the process, it causes VSPAERO to write out a template file with default settings.  Make sure to 'Save Setup' after you change anything in that text window.

Soon, the way OpenVSP and VSPAERO are integrated will change -- we are updating the GUI to have control of _all_ VSPAERO inputs.  This will eliminate the need for the text window displaying the setup file.  It will also eliminate the need for the 'Create new setup' step.  Instead, everything will be controlled from the GUI and the *.setup file will be written by VSPAERO before each run.

 

3. In the stability analysis, is there any way to assign the amount of alpha/ beta changed and the deflection angle of control surfaces. Also, is the analysis (not only the stability option but VSPAERO as a whole) capable of modeling changes brought by deploying simple flaps, slats, spoilers or other control surfaces?

I believe changing the delta for control derivatives would require modifying the code.  It is not a tremendously difficult change, but you would need to be set up to compile VSPAERO yourself.

You will need to gain experience with the applicability of VSPAERO (VLM or Panel) to different situations.

For example, it should do a good job at modeling elevators, ailerons, rudder for normal (relatively small) deflections.  However, I wouldn't trust it at all for modeling spoilers -- as their effect is dominated by separated flow.

 

4. Are the angular momentum brought by the rotors (actuator disks) taken into account? Will it affect the results by adding yawing/rolling moment?

The actuator disk model does include swirl in the slipstream.  It will cause flow asymmetry and moments.

 

5. Is there any way to obtain the total Cm of x, y, z axis, and Cp, AC positions of the whole aircraft model? And where could I find the stability derivatives?

The moment coefficients output are about the input Cg positions -- so if you set the Cg to (0,0,0), then you'll get the moments about the origin.

If you run with -stab, it will perturb the flow properties and calculate stability derivatives.

The current version won't find the AC position for you, but if you simply run two alphas and record the CM and CL values, you should be able to find the AC very easily.

The next update to the OpenVSP/VSPAERO integration (referenced above) will make a bunch of this sort of thing much easier.  A lot of these steps that are manual now will become automated.

Rob

[Gordon Yuhim Lai]

Thanks for the clear explanation above, Rob.

I got some more questions (or problems encountered) regarding to VSPAERO

1. .In the control group section of the command window on the left, what do those numbers mean?

2. I found the Panel Method crashing frequently whenever I put the vertical stabilizer and horizontal stabilizer together, or whenever I put the fuselage in the solver. However, the panel method works completely fine when there are only horizontal surfaces (wing and horizontal stabilizer). It also works when the vertical stabilizer alone is put into the solver, though the results as seen in the viewer contains missing mesh (triangular holes). What could be the possible reasons for these kind of phenomena?

Many thanks.

Gordon

[Rob McDonald]

What version of OpenVSP are you running?

I'm not sure what 'control group section of the command window' you're
talking about.

Does your OS issue any sort of error report when you see the crash in
VSPAERO? Does it just crash the solver, or does it bring down the
OpenVSP GUI as well?

In general, if the GUI goes down too, it is a bug in OpenVSP --
probably something randomly triggered, having nothing to do at all
with your geometry. If it is a bug in VSPAERO, it should be 100%
reproducible with the same geometry -- perhaps you can post the
offending file so we can try to see if there is anything that jumps
out about it.

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.

[Gordon Yuhim Lai]

What version of OpenVSP are you running?

    I'm running on 3.10.0

I'm not sure what 'control group section of the command window' you're
talking about.

  They are the numbers in the highlighted part (i.e. 1., 1., 10.).

 

Does your OS issue any sort of error report when you see the crash in
VSPAERO?  

  

  Yes, the OS reported VSPAERO is not responding and has to be shut down whenever vertical stabilizer and horizontal stabilizer are present. However, it is still possible to return to the config interface to kill solver afterwards (The OS don't actually shut down OpenVSP). For the fuselage, the solver just freezes as seen in the solver console (as seen in the picture below).

 

Does it just crash the solver, or does it bring down the
OpenVSP GUI as well?

 

  As I mentioned above, most of the time it just crashes the solver. Yet sometimes it brings down OpenVSP too.

 

The picture below shows what happens when putting the vertical stablizer alone in the solver.

The .vsp file is attached for your reference.

Again, many thanks.

Gordon

[Rob McDonald]

Gordon,

It looks like you made the same mistake -- twice...  Easy mistake to make -- also easy to fix.

When you create a wing, you really define the right (positive Y) wing.  Since most models will have a symmetrical wing copy, we turn on XZ symmetry for wings by default.

For a vertical tail, you rotate about the x-axis by 90 degrees.  Since the vertical is still on the center line (y=0), you end up with two vertical tails perfectly superimposed on top of one another.

This really confuses the geometry generation before VSPAERO -- causing the mixed color triangles in the GUI and also the missing gaps you see later in the VSPAERO viewer.

Other geometry types don't have symmetry enabled by default.  However, at some point, you enabled symmetry for the fuselage.  Since the fuselage is sitting on the center line, you ended up with two fuselages perfectly superimposed again -- causing the same problems.

So, turn off the symmetry on the vertical tail and the fuselage and you can get this solution...

You will certainly want more spanwise resolution, but with some experimentation, you should be able to get good results.

Although symmetry threw off your vertical tail and fuselage, it would be good to use it for the actuator disks and engine pods.  You used blank components and attachment to make those assemblies (good), so when you use symmetry, you need to choose an ancestor axis to be symmetrical about...  There was a presentation at the 2016 OpenVSP Workshop about symmetry and attachment.  That is probably the best documentation available on that.

You threw me off of the 'control group' discussion because the setup configuration window is on the right -- the current GUI doesn't do anything for control surfaces.

The version of VSPAERO you are using can model control surfaces when running in the thin-surface VLM mode.  Control surfaces can be grouped to model coordinated surfaces, mixing, etc.

Soon, the OpenVSP GUI will be modified to allow definition of control surface groups without editing the control file manually.

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+unsubscribe@googlegroups.com.

[Gordon Yuhim Lai]

Rob,

Thanks for pointing out the problems - they were really hard for me to spot them.

So how may I config the control surface groups in the setup configuration window on the right?

Thanks.

Gordon

[Rob McDonald]

The best documentation for the control surface stuff for VSPAERO is
probably the videos Nick Brake did at the last OpenVSP workshop. I
think he was in the middle of the second day.

https://nari.arc.nasa.gov/vsp

The control groups are defined by

GroupName
Comma separated list of surfaces in group
Comma separated list of rates per surface
Group deflection

Example:

ControlGroup_1
SS_CONT_0, SS_CONT_0
1., 1.
10.

The default control surface subsurface is SS_CONT_0. It appears twice
because the wing has two symmetrical copies - left and right.

The rates are 1.,1. -- Deflect the surfaces equally, each in their
positive direction. If you look at the GUI, you will see the hinge
line has a circle arrow indicating the direction of positive rotation.
The symmetry reverses direction (right hand rule, align your thumb
from root to tip), so this will provide 'aileron' action. If you want
these surfaces to behave like plain flaps, 1.,-1. will do that.

Each surface is deflected by the product of the group deflection and
its rate. So, if a rate is 0.5, and the group deflection is 10, that
surface will deflect 5.0. Rates can be greater than one.

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.

[Gordon Yuhim Lai]

Rob,

Thanks for the clear explanation. I think that is all I want to know for now. Really appreciate your help.:)

Gordon

[Gordon Yuhim Lai]

Rob,

I just found another problem. When I put the revised model in VSPAERO , the Cdi obtained are negative for both panel method and VLM. The results for Cdi are also negative for the original model. What could be possibly wrong?

Many thanks.

Gordon