STL File Import

[hausi...@gmail.com]

Hi Rob

I am new to OpenVSP and tried to import a stl-file from SolidWorks in order to analyse a wing with VSPAERO. I can see its geometry in OpenVSP but I couldn't find a way to analyse it with VSPAERO. Is there a procedure to acomplish a CFD Analysis starting with an imported stl-file?

Many thanks for your help

Hans

[Rob McDonald]

No, a STL import is just a shell of the geometry, it doesn't have the
underlying form or smooth surfaces. A few of OpenVSP's analyses will
work with it (wetted area, wave drag, etc) but you can't feed CFDMesh
or the vortex lattice mode of VSPAERO directly.

Usually the best approach is to use the STL as a guide to develop a
native OpenVSP model that matches your geometry. If you want to get
as close as possible, you can use FitModel to assist the process.

Hope this helps,

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.

[Jayrajsinh Solanki]

I believe if we can import geometry with meshing in STL file format ,so we will be able to do analysis....

[Brandon Litherland]

You can bring an STL (preferably in ASCII format) into VSP using Import.  You can also run VSPAERO using Panel mode if you like.  However, be very, very careful about trusting any results.  The STL from CAD will not typically have the same type of grid resolution that would be appropriate for an aero analysis.  For example, the grouping at the leading or trailing edge could be too rough while other sections where the aero gradients are low are too fine.  Furthermore, unless the model is already simplified and defeatured sufficiently for an aero run, you'll have a lot of extra bits that will mess up your results.  See an example below of a simple unit sphere in VSPAERO.  The pressures were calculated but the solver did not compute drag because there were no wakes attached because no sufficiently sharp TE was detected.  In another example, we successfully test an imported STL of a NACA 2412 cambered, finite wing using Panel mode.  This worked just fine.  For this case the difference between the VLM and Panel modes was about 4% in lift but again, this is an arbitrary case.

As Rob states, if at all possible it is best to recreate your geometry or some representation of your geometry within VSP.  Hope this helps!

[Jayrajsinh Solanki]

Dear  Brandon Litherland  ,

  Thank you very much for your guidance , however i am going to import somplified nacelle geometry (created with cst method) in vspaero with meshing.. i am asking for further guidance that which things should i keep in mind before imporng geometry with mesh as well as while running .

Thank you  

[Brandon Litherland]

That still all depends on what you want to do with it.  If you intend to use VSPAERO Panel mode on the part, then you'll want to have cell clustering appropriate for aero (dense at the nose or leading edge/trailing edge).  If you are using it for placement or meshing, then it can be more rough.

[Jayrajsinh Solanki]

Dear  Brandon Litherland ,

  Thank you  very much , however i am stuck as  i was able to import  the nacelle geometry , but to create the fan effect i am using the actuator disk, the problem with that is every time i run it..vsp stop functioning.. can you give me more guidance about how to create fan effect ? with above mention  method.. 

 Thank you very much in advance for responding  

[Brandon Litherland]

It sounds like you are trying to put a fan inside a nacelle (ducted fan or turbine...?).  Can you send a snapshot of your model or just attach the model?

A ducted fan analysis will only run in Panel mode and even then you'll have to treat the results with a healthy amount of skepticism.  Also note that actuator disks are intended for Propellers but can sometimes be used for fans.  The assumptions of the Actuator Disk method have to be satisfied for it to work.  For example, if you were to try and treat the exit of a rocket nozzle as an actuator disk with 1 RPM and nearly pure thrust, it will run but the answer is meaningless.

[Brandon Litherland]

You can run a propeller analysis using the -rotor command from the VSPAERO command line input.  However, this spins the entire model about the axis of rotation.  I think there might be some nomenclature miscommunications going on.  

Rather than asking how to model something, can you tell us what it is that you are trying to accomplish?  Are you trying to model the installed thrust of a ducted fan?  Are you just trying to put a propulsor into an existing aircraft model?  In many cases, you can get what you need done in a different way.

On Monday, August 31, 2020 at 2:35:21 PM UTC-4 jayrajs...@gmail.com wrote:

So basically what i have done that, above shown fig is in stl with surface mesh and;  i am trying  to put fan inside of it and creating a mesh in openvsp...but as soon as i run vsp..it stops functioning.. 

please suggest.. 
Thanks in advance.

On Monday, 31 August 2020 at 23:13:54 UTC+5:30 Jayrajsinh Solanki wrote:

this is the snap shot of model and i am trying to put fan  inside of it trying to make analysis.

[Brandon Litherland]

I see the issue, I think.  This STL is a surface mesh and not a watertight solid.  VSPAERO isn't going to run this.

You'll most likely want to remake the nacelle using either General XSec or Edit Curve type.  From there, you can put an actuator disk or propeller in the model and run Panel mode in VSPAERO.  See below.

On Wednesday, September 2, 2020 at 9:01:19 AM UTC-4 jayrajs...@gmail.com wrote:

Dear  Brandon Litherland , 

  Thank you very much for responding.

 The thing i am trying to achieve is to fnd the distribution of cp on nacelle lips; with considering the what will be the effect of ducted fan on cp  on lips. Today , however i tried with propellor but i was not able to do it.

just finding a way to solve this issue.

for your reference i m attaching simple geo along with it.  And the geometry is very rough. 

hoping a furthur guidance.

[JPS]

Dear  Brandon Litherland , 

  Thank you very much for responding.

 As you mentioned that by recreating the geometry by General XSec or Edit Curve type; does that mean by running the vsp script ? if not, then could you please explain ?

 another question; Is there any way that we  could create a watertight solid other than vsp ; which will be useful in VSP?

Thank you in advance.

Have a nice day.

--

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/cf49e42f-9884-4576-95a3-104209c34311n%40googlegroups.com.

[Rob McDonald]

That is expected.

CompGeom works on individually water-tight components.  A disk has no volume and is not a water-tight component.

That said, the panel method does not need the panels of the actuator disk to insert an actuator disk.  It should still work.  The disk comes through as a center point, a normal vector, and a diameter.

That said again -- do not expect the actuator disk + duct model to actually represent what you want it to.  This will be wrong, it will not match your physical system.  How are you going to do you verification that the results are useful?

Rob

On Fri, Sep 4, 2020 at 8:27 AM JPS <jayrajs...@gmail.com> wrote:

Dear  Brandon Litherland ,

 Moreover, i am facing an issue with actuator disk..

As you above mentioned that VSP require water tight solid, i was able to create the water tight solid..i imported the geometry and tried to add actuator disk;  but now the problem is with the actuator disk..as soon as i  run comp geo vsp was able to calculate the my nacelle surface  geometry, but was not able to calculate the actuator disk surface..

And I am facing the same issue when i create the geometry in the VSP. i created the duct and i tried to add disk.. i  am using 3.21.2

Hoping a guidance from you..and thank you very much for responding..

have a good weekend !!

To view this discussion on the web visit https://groups.google.com/d/msgid/openvsp/CAEfzvNDfaMM4yJ9AqX8_Xs7K_%2B%2BJ_f76yqQe0mxhLO%3DqzRfTXA%40mail.gmail.com.

[Jayrajsinh Solanki]

Dear ROB,

 Thank you very much for responding. i am going to compare the results with the result obtained with the high fidelity methods.

 Moreover, what will be solution of above problem ? according to you..which will match my phsical system.. with using the panel method.

Please guide..hoping a response..

[Rob McDonald]

Using rotating blades should be able to match the physics much better -- but will be several orders more expensive.

Potential flow methods work via superposition -- a bunch of fundamental solutions are summed together.  We usually think of these as sources, sinks, vortices, and doublets.  However, a uniform freestream is also a solution we use.  In potential methods, we usually treat the vortices as having unknown strength.  That is what the matrix solution is all about.  However, the freestream component has strength known ahead of time (velocity and direction are known).  So, it can go on the RHS of the matrix solution.

The actuator disk solution in VSPAERO uses an analytical solution for an actuator disk in isolation.  Imagine an actuator disk of known strength in a known freestream flow.  Conway worked out the 3D velocity increment everywhere in space caused by that disk.  VSPAERO takes that disk influence and adds it to the RHS of the flow field along with the freestream.

Consequently, the disk can influence the rest of the flowfield -- it changed the freestream.  However, the disk's flow can not be altered.

So, if you have a wing-mounted engine in cruise, where the slipstream of the prop is basically unaffected by the airframe, this approach probably works quite well.

However, if you have a situation where the airframe would substantially deflect the slipstream, then I would not expect it to do as well.  It might work out -- so validation tests may prove promising.  But I certainly wouldn't use it without convincing myself that it was producing useful results.

Brandon has cautioned you about this in several past emails -- he wasn't as blunt about it as I have been.

Rob

To view this discussion on the web visit https://groups.google.com/d/msgid/openvsp/c603c244-befb-4602-b47f-f4e8cd9c6793n%40googlegroups.com.

[JPS]

Dear Rob & Brandon

  Thank you very much for your responses.they were very helpful.

 

Regards

To view this discussion on the web visit https://groups.google.com/d/msgid/openvsp/CAEppYpHEBu02K96A6EML7Gho1GmzZFLpVChYSvvJYqovx-ndzA%40mail.gmail.com.

[muku]

Hi!

I am a new user of of OpenVSP.

I want to import a stl file with OpenVSP and perform supersonic calculation.

In the case of stl, I couldn't use the From model of Sref,  so I decided to use manual.

but, After importing stl data,  I don't know the unit. Please tell me.

and,I can't calculate more than M1.1 with stl data, please tell me how.

Muku

2020年9月10日木曜日 5:26:30 UTC+9 JPS:

[Rob McDonald]

The dimensions are whatever the model was made in.  STL files have no way of conveying dimensions.  I have no way of knowing how your model was built.

Only thin-surface mode in VSPAERO will work in supersonic flow.  So, you can't use a STL file read in supersonically.

You could use the wave drag calculation, but that is it.

I recommend re-building a VSP native version of your model using the STL as a guide.

Rob

To view this discussion on the web visit https://groups.google.com/d/msgid/openvsp/67e41242-2a8a-4ee6-9f51-ab94557020b9n%40googlegroups.com.

[muku]

Dear Rob McDonald

Thank you for your reply. It was very helpful.

I have two new questions.

 Regarding rebuilding the VSP native version of the model, is it possible to rebuild the VSP file by "save as" the stl file?

 Is it possible to change the air density with VSPAERO? If you can, please tell me the setting location.

Muku

2020年11月20日金曜日 1:50:32 UTC+9 Rob McDonald:

[Rob McDonald]

No, Save-As will keep the mesh in the *.vsp3 file, but it will stay a mesh.  By a Native model, I mean a parametric model built up from OpenVSP components.

You can change the air density on the Advanced tab under the field 'rho'.  Many VSPAERO analyses ignore it, so you may not need to change it.

Rob

To view this discussion on the web visit https://groups.google.com/d/msgid/openvsp/4bd6ff53-fea4-4005-9278-740c60cfadfbn%40googlegroups.com.

[muku]

Dear Rob

Thank you for answering. It was helpful .

I have new question about modeling.

Is there a way to change the curvature when using a pod or fuselage?

Muku

2020年11月27日金曜日 4:04:10 UTC+9 Rob McDonald:

[Brandon Litherland]

A Pod is a specific shape that is only defined by the length and fineness ratio.  A Fuselage or Stack can be skinned or lofted into almost any shape.  There is a series of tutorials on the Ground School here 
https://vspu.larc.nasa.gov/training-content/chapter-1-vspfundamentals/skinning-blending-and-section-modification/skinning-introduction/.

I highly recommend watching the Ground School tutorial on Basic Modeling if you haven't gone through the process before.

[muku miyagi]

Dear Brandon Litherland

Thank you for answering.

I was able to make a torso!

Next, I want to make wings.

When using fit model, is it possible to match the shape of the airfoil of the point cloud data and the airfoil of the model even if they are different?

By the way, the airfoil is not existing in openvsp.

Muku

2020年12月24日(木) 20:16 Brandon Litherland <litherlan...@gmail.com>:

You received this message because you are subscribed to a topic in the Google Groups "OpenVSP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openvsp/K73DA2UlGUU/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openvsp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openvsp/ed8c4c07-317e-4d56-a9ec-d331315f4281n%40googlegroups.com.

[Rob McDonald]

Using fit model to match an airfoil may be possible -- but it is certainly not recommended.

If you insist, you should match the rest of the shape of the wing without worrying about the airfoil.  First match the position, size, and planform.  Then the twist and dihedral.  You should be able to do a very good job of matching the entire leading edge line and trailing edge line of the wing -- before you even think about airfoils.

If you really want to match the airfoil shape using fitmodel (you don't), then change the airfoil types to CST and use those parameters to match the upper and lower surfaces to match.

If you know the airfoils are not in OpenVSP's built-in types -- does that mean you know what the airfoils are?  If so, you will be much happier using the file airfoil type.  Get a file either from the UIUC Airfoil Database, or make your own file (if they are points you designed yourself).

If you don't know what the airfoil is -- but the mesh you are working from is a structured CFD mesh, you should be able to extract some of the curves, reformat them into an airfoil basis and use those files for the points.

What are you really trying to accomplish?  If you already have a CAD model, why are you trying to get it into OpenVSP?  That may still be a good idea, but there may be a better way to accomplish your overall goal.  What is that goal?  What are you really starting from?

Rob

To view this discussion on the web visit https://groups.google.com/d/msgid/openvsp/CAENN_OBphVYev6ktpMgkWWvG78NE3636w5reQV61pSE6NKtKzA%40mail.gmail.com.

[Rob McDonald]

As Brandon mentioned -- a Pod is a very simple component with only two parameters.

In many ways, the Pod was created as a first component for debugging OpenVSP.  It was very simple to develop and we use it to test things all the time.  In practical use, there isn't much reason a user would ever use the Pod.  It is too limited for most purposes.

The Fuselage and Stack components are much more flexible and allow you to adjust curvature a number of ways (I'm not sure I understand your question).  In general, I recommend you use the Stack component instead of the Fuselage component.

Rob

To view this discussion on the web visit https://groups.google.com/d/msgid/openvsp/d8322067-bd96-4506-8a9a-be18c3170c54n%40googlegroups.com.

[muku miyagi]

Dear Rob Mcdonald

I have stl data, but I heard that it is not suitable for calculating with vspaero.

Therefore, I would like to convert stl data to cad data corresponding to vspaero.

After creating a model corresponding to vspaero, I would like to calculate the drag coefficient and lift coefficient.

Ideally, I would like to change the shape of the wing of the created CAD data and calculate.

Muku

2020年12月25日(金) 9:25 Rob McDonald <rob.a.m...@gmail.com>:

You received this message because you are subscribed to a topic in the Google Groups "OpenVSP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openvsp/K73DA2UlGUU/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openvsp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openvsp/CAEppYpFvEN9%2BjaVd8b1RX1v9HBWOZyJaGXkCPdkwhK2cZT%2BZzw%40mail.gmail.com.

[Brandon Litherland]

Hi Muku,

If that's all you need then getting the model close to the STL shape will be fine for VSPAERO.  Be aware that VSPAERO is an inviscid, vortex lattice or panel solver that has it's own set of assumptions.  Thickness doesn't really come into play all that much.  You probably don't even need to match the airfoil all that well.  Twist and chord will get you most of the way there and then the camber will do the rest.  Don't start by trying to make the most accurate model possible.  Start by making the basic components about right and then run your analyses.  Check that the model is stable and that the results are about right.  If they are significantly wrong, locate the error in the model and then try again.  Keep doing this until your model is sufficiently detailed and your results are accurate or trusted.  

Check the OpenVSP Ground School as mentioned previously or see the videos from the 2020 OpenVSP Workshop.  There are a LOT of great demos there.

Remember, "All models are wrong... but some models are useful." - Box  

and " Since all models are wrong the scientist must be alert to what is importantly wrong. It is inappropriate to be concerned about mice when there are tigers abroad. " - also Box. 

[muku]

Dear Brandon Litherland .

Thank you for your detailed advice.

I will make it happen.

I want to change the topic and ask a new question.

I can't find the literature, so I'd like to ask a question here.

Does the panel method work even when the air density in the upper sky is low?
This is because when approximating from the Navier-Stokes equations, I thought that the approximation method would change if the air density changed.

muku

2020年12月29日火曜日 1:36:39 UTC+9 litherlan...@gmail.com:

[muku]

Dear Brandon Litherland and Rob McDonald .

I have an additional questioon.

I tried using two versions of openvsp(ver 3.16.2 and ver 3.21.1)

Then ,there was a change in the CL and CD even with the same model.

Should I look at the new version(ver 3.21.1) in this case?

muku

2020年12月30日水曜日 22:41:35 UTC+9 muku:

[Brandon Litherland]

My first question is to ask how big of a change did you see?  If we are talking about a few pounds of lift, then that can come from any number of little tweaks to the code or setup.

if there is a significant change, then we need to ensure that the run was set up in exactly the same way between the versions.  This includes the preconditioner, wake nodes, wake iterations, etc.

[muku]

Dear Brandon Litherland.

Sorry for the late contact.

Regarding the matter that the calculation result differs depending on the version, it has been solved!

The cause was that the model created for each version was different.

Thank you for your polite reply.

muku

2021年1月8日金曜日 21:20:15 UTC+9 litherlan...@gmail.com: