Flow Field Query

[Dennis Haimerl]

VSP ver 3.36.0 is being ran on a Windows machine. I am trying to obtain the flow velocity vector behind a wing. Once I understand how to obtain a velocity vector one chord length behind the wing T.E. on the chord line and wing centerline, I should be able to expand and find additional values that I need behind a flapped wing just forward of the h-tail L.E. I understand how the coordinate matrix works, but have not been able to determine how to apply the values to vspAero. The only information I have on this is shown below.

1) https://www.youtube.com/watch?v=H48Fv1rIIOs&t=1s

2020 VSP WS: VSPAERO GUI & Tutorial – Advanced by Justin Gravett

At the 32 min. mark the video shows a slide for Flow Field Query and states:

“Add the following to the *.vspaero input file and run from the command line:”

NumberofSurveyPoints=n

1 x1 y1 z1

2 x2 y2 z2 //for however many points needed

******

n xn yn zn

2) https://groups.google.com/g/openvsp/c/8TF3DCy7BWA. by Will Walton (Thanks Brandon)

At the bottom of the vspaero file, include:

NumberofSurveyPoints=n

1 x1 y1 z1

2 x2 y2 z2 //for however many points needed

******

n xn yn zn

and it will output a file with the .svy extension, that includes x,y,z and the corresponding u,v,w velocity components.

I have viewed the “Modifying and Executing a Design Variable File” video regarding running vsp.exe from the command line with a *.des file, but I don’t understand how to run the vspaero.exe with the survey points as stated in the 1st example.

I don’t at all understand what is meant in the 2nd example by including the survey data “at the bottom of the vspAero file”.

I have studied the included swept wing TR1208.vspscript file to see if I could apply the data there, but I’m not sure of the correct syntax to use. I don’t have experience with C++, though I have experience with Turbo Pascal and Visual Basic syntax so I was able to follow much of what the file was doing.

I’ve attached a couple of images of where I am at with the model. I am modeling for aero results, not presentation. I have provisions for Frise ailerons that are not quite complete at this time. The CL and CM data results using VLM are even closer to Datcom and limited flight test data than the TN-808 comparison I shared earlier. I will have a few questions regarding my use of the Blank module for translating and rotating the geometry from the FS & WL reference to a wing AC reference. The CL vs AoA is working, but I’m having issues with the wing incidence angle and CM values. I need to work out the downwash values first, so more about this later. 

[Brandon Litherland]

Dennis,

Sorry that process wasn't more explicit.  To accomplish the point flow survey in VSPAERO, you'll be running from the command line.  You'll see the command to enter into your terminal/CMD at the very top of the VSPAERO printout in the GUI window.  Something like: "C:\Path_to_VSP\vspaero.exe" -omp 4 "C:\Path_to_model\modelname_DegenGeom".

To add the survey points to your input file (*.vspaero), you'll add the lines described in your post above to the end of the text file and save it.  Then you'll be able to run VSPAERO from the command line and it will catch that input and write the *.svy file.  I'm attaching example VSPAERO and SVY for a default wing analysis.  Let us know if this doesn't clear things up.

[Dennis Haimerl]

After spending some time attempting to interpret the proper method and syntax, I finally figure it out to be “vspaero.exe -interrogate Flow_DegenGeom” at the command line. I keep adding “Flow_DegenGeom.vspaero” for the file name which lead me down a rabbit hole. Now that I am getting the survey data, the next step is to see how the results correlate with NACA Report No. 648.

Is see what appears to be a single point wake angle vector value in the printout, but I’m not sure what this is referring to. Do you know of a method to find the center of the wing wake withot having to plot a large number of flow survey points?

This will give me a lot of data to look at. It would be nice if the GUI had a place to add flow survey data points similar to the Cp slicer. If I wanted to learn C++ or Python I could probably write something, but I'm not fully retired yet and it is getting hard for me to learn new languages. I suspect that additions to the GUI would be a ways down on the wish list as it seems the effort is more toward adding additional solvers, which is commendable.

When I get back to the full airplane model I may have questions about how to find meshing issues when the solver won’t complete a run. I think I may have an issue with the fuselage/v-tail intersection. I need to investigate this further when I have more time to be sure.

If you think anyone would be interested I could try to post some of my downwash comparison results. Thanks very much for the help.  

[Brandon Litherland]

The -interrogate option should output the SVY file based on an existing solution (if you have one).  Running as I describe above will run a new case and output the results.  I'm still a bit unclear on exactly what you are trying to obtain from VSPAERO.  From your first post in this thread, it sounds like you would like to have the flow UVW components one chord length aft of the trailing edge at the center of the wing (Y = 0), aligned with the chord.  You also want to have UVW data some distance from the horz. tail leading edge which is simply more X Y Z coordinates in the list of survey points. For now, you need to manually add these values and run from the command line.

If you are interested in the flow field at various slices, you can add slices in the VSPAERO GUI and it will write out the quad DAT file containing the velocity field for all points in the quad tree. This was discussed in a prior post of yours here https://groups.google.com/g/openvsp/c/FEnQiVtBvfc/m/1IjanMgOEQAJ.  For what you are describing here, I still think survey points are your best bet.

[Dennis Haimerl]

In the past I have wasted a significant amount of time learning softeware that in the end did not correlate well to wind tunnel or flight test data. My first goal with VSPAero was to verify that the lift curves and moments developed from the deflection of highly translated Fowler flaps was comparable to experimental and flight data as well as DATCOM calculation methods. So far the correlations have been acceptable.

My current goal is to determine the downwash angle near the horizontal tail and the location of the wing wake centerline at high angles of attack with the flaps fully extended. I’m investigating this for two reasons. First, I want to have some degree of confidence that the horizontal tail has adequate margin from dangerous tail stalling. Second, I would like an estimate as to the wing wake center to avoid placing any appendages in the wake when operating at high AoA’s. I have experienced some experimental airplanes that exhibited tail shake from the wing wake prior to entering a stall.

It appears that the flow survey will only output data based on the highest AoA if a multiple angle solution is held in the DegenGeom files. This has not been an issue for what I am doing. Now that I have generated some flow survey data I have a good idea of the vertical location limitations for the horizontal tail. I can now calculate CL Max. using DATCOM for the maximum up and down elevator deflections, then apply the range of downwash angles with elevator deflections using VSPAero. When completing a comparison to Hoerner data using this method I found very good correlations.

I do not have wake data yet. I do not see an "Export to *.csv" button in the VSPAero GUI as I understood from a previous thread. I only found an "Export to *.csv" button in the Comp Geom and Degen Geom GUI that seem to contain only model geometry, not VSPAero results data. I assume that the wake centerline for a give U, V location is determined by the W location where the flow angle is largest in relation to the relative wind (free stream). This could be calculated from a large amount of flow data, but it would be time consuming. Since the wing wake can be depicted in the Viewer, I assume that the U,V,W location data is held in the adb file. Is there a way to access that information?

The next step is going to see where the problem is in my complete aircraft model that keeps it from solving. It was working in VLM but not Panel Method prior to adding the vertical tail and changing the coordinate system using Blank. There may be a messing issue with the Vertical tail. I am also look forward to solving for static stability in the future. I has been a little hard to learn some aspects of the user interface, but I am impressed with the results I am getting.  

[cean wang]

Nice to see the quad tree.

---

发件人: ope...@googlegroups.com <ope...@googlegroups.com> 代表 Brandon Litherland <litherlan...@gmail.com>
发送时间: 2024年3月14日 5:06
收件人: OpenVSP <ope...@googlegroups.com>
主题: [OpenVSP] Re: Flow Field Query

[cean wang]

I tried to modify the vsploads file to output the UVW values in a new Wake file. But this Plain Flap model generated an error. 

terminate called after throwing an instance of 'std::bad_array_new_length'
  what():  std::bad_array_new_length

You can see in the Wake file, Case = 3 has very big numbers.

---

发件人: ope...@googlegroups.com <ope...@googlegroups.com> 代表 Dennis Haimerl <lightde...@gmail.com>
发送时间: 2024年3月14日 20:20

收件人: OpenVSP <ope...@googlegroups.com>
主题: [OpenVSP] Re: Flow Field Query

[Dennis Haimerl]

Cean,

I assume you are showing a code written in C that you are using to access one of the *.adb files. After 30 years starting with HPLG, I was hoping that I would not have to learn my 7th programming language (Python looks more familiar to me). I’ve only written code occasionally when I cannot find affordable software to complete the task at hand. I also assume that the slice.jpg file is showing the wake, being those vectors that are clustered together. I now have flow field  vectors at the horz. tail L.E. for various flap deflections for the partial span flaps being used. I was hoping to obtain the vertical location of the wing wake with flaps fully deflected with respect to the horz. tail that is just short of 2 chord lengths behind the wing T.E.