OpenVSP Analysis softwares not Responding.

[Ogweno Emmanuel]

Hello, I've been using OpenVSP for about 3 months now on my pc. The analysis software, despite a few hitches, used to work just fine. However, lately, I've been experiencing some challenges with my current designs. 

Whenever I try to run parasitic drag analysis and VSPAERO analysis the whole thing stops responding. In some instances, it took 3 hours to complete a single VSPAERO analysis, but the results I got didn't make any sense.

I've tried changing to different versions of the software but they all end up not responding. 

How can I fix this? Or is it something to do with my model design? 

Here's a model I tried to analyze.  

[Brandon Litherland]

Without more context to your issue, we can't provide much detailed guidance.  Based on your model alone, you have very minor dihedral but also have Rotate Foils selected.  This causes a gap at the root which is visible if you view from the front that causes VSP to crash or give unpredictable results.  Also, you have Steady stability calculations turned on which means at least 7 VSPAERO runs per combination of Alpha, Beta, and Mach (6 in your case).  So VSPAERO will be running a minimum of 42 different cases with 5 wake iterations each.  You see how this can get out of hand.  Furthermore, your wing is 0.52 length units in total span with an area of 0.066 square length units.  I have to assume that this is in meters and is a very small RC glider or something.  With Steady selected, your Vinf is activated and is set at 100 length units/second but your Mach is set to 0.073 (about 25 m/s).  This can also cause problems.

Fixing the dihedral gap allowed VSPAERO to run just fine.  Without knowing what you're trying to get out of it, I can't provide better guidance on your analyses.

[Ogweno Emmanuel]

I've tried to run the analysis after turning off Rotate airfoils but it still hangs. Was that the only change that you made? I further turned off steady stability calculations but it still didn't work. 

[Brandon Litherland]

What version of OpenVSP are you using?  

Did you download it or build it yourself?

Are you analyzing using VLM or Panel mode?

What is the desired outcome of the analyses?

The Rotate Foils option was the only thing that I changed before I ran it.  The alpha sweep from 0 to 10 using 6 steps worked as expected.

[Ogweno Emmanuel]

I'm using OpenVSP 3.26.0

I downloaded it from the OpenVSP website

I was analyzing using panel method. I tried VLM too

The desired outcome is basically to get the analysis results. I want to get a basic understanding of the design, like obtaining the neutral point, spanwise load distribution, sweep results, etc.

[Ogweno Emmanuel]

My main intent is to design a flying wing model, analyze it using the software, and after a satisfactory design, build the model

On Tuesday, January 18, 2022 at 12:44:12 AM UTC+3 litherlan...@gmail.com wrote:

[Brandon Litherland]

I made some changes to the panel distribution and the dihedral.  May not be what you intended but it's a start.  You'll find that in Panel mode, the wingtips have odd peaks in the solution.  The GMRES iterations are trying their hardest to force that down and in some cases this will throw things out of whack.  Try not to use wing section breaks except where your wing has a real discontinuity in the shape.  I converted the model to feet just to see if that helped the peaks.  It did somewhat but not entirely.  I'd stick with VLM for now.

Give this model a try and see if it works out for you.

[Ogweno Emmanuel]

Hello, I realized that if you change the root cap type under plan from round to flat, the model simulates just fine.

[Brandon Litherland]

I rechecked the vsp74.vsp3 file that you uploaded and the caps are both Flat type.  When were they round?

And, yes, roots should not be rounded unless there is some reason to do so such as having Y-separation in the wing sections.

[Ogweno Emmanuel]

I'm not sure when they went round but as soon as I changed that the software responded. I'm highly appreciative of the support you've given me.

[Tom]

Should the root cap be" flat" or "none"? If "flat" what is appropriate use of "none"? When a wing section pierces another component?

Thanks,

[Brandon Litherland]

In almost all cases, leaving the root cap as flat is acceptable.  This is true for wings in isolation and for wings protruding from bodies.  Choosing None can be useful if you know that the root will always be located at zero Y and you want U=0 to start at the root instead of the root camber line.  For example, if you need to sample the UW surface of a wing and you know that you will only ever use VLM VSPAERO mode, then leaving the caps off is fine.  In practice, however, we just remember where the root and tip U locations are.  There are, of course, other unique cases where you would use one or the other but the default Flat is almost always the best way to go.

- Brandon

[Rob McDonald]

Even in the special case Brandon mentions, leaving a wing's root cap to None is likely to cause problems.  A wing's root cap should almost always be left as Flat.

Every object in OpenVSP should be self-watertight.  If you violate that, unexpected things are probably going to start breaking.

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.
To view this discussion on the web visit https://groups.google.com/d/msgid/openvsp/555a20c2-97b1-4d8f-b729-4b45b82ce4bcn%40googlegroups.com.