US Hawks Hang Gliding Association

This topic invites model hang gliders to be discussed. Present plans, videos, aerodynamic discussions, etc. Model hang gliders have been being built for many centuries.

Start:
The Sonic Hang Glider: Project #75

This project seems to neglect dive recovery reflex or other means.
============================================================

Cool producer, Harry!

Hang glider models in scale 1:5 and 1:3 since 1980.
As an flying team we know the different types since the beginning.
All hang glider scale models are hand made accordingly by full scale drawings or brochures.

The models are all custom made, not standard serial production.

Consider a tradition of building a model of the wing that one may build or buy for full-size hang gliding.

==========================
RC choices: pure gliding vs motorized vs motorized-to-gliding :: practice weight-shift control.

Over the Christmas holiday, I finally finished the second version of my hang glider model building software.

I had written the first version for Blender years ago (you may have seen some of those images here on the forum). The new version is written in Javascript to make it more available. The graphics are much more primitive than Blenders, but it's certainly sufficient for building basic models of common gliders. It can also export a JSON representation of your glider to possibly import into other programs like Blender.

It's not ready for wide distribution yet, but if anyone would like to try it out, please let me know...

I've uploaded a working copy of the glider designer here:

    http://ushawks.org/designer

Here's a screen shot:


It's currently all line drawing (no filled surfaces), and there's no hidden line removal attempted, so some lines will draw over others when they shouldn't. But it's good enough to play with a new design or to make a 3D model of existing gliders (helpful for diagrams and our training manual). The models can then be rendered in a 3D tool like Blender for finished artwork. There are red tabs (not shown) for 3D rotation, and clicking and dragging anywhere in the window will also rotate the view.

These are the basic parameters for the model:


You lay out the glider starting with the front and rear locations for the keel (it's best to only use the "Y" dimension because I haven't tested alternatives). You also specify a point (x,y,z) for the wing tip. That lets you set the nose angle along with dihedral or anhedral. The control bar is specified with a height and a width. All units are envisioned to be feet, but meters would also be fine (just change the zoom to fit the window). The control bar height is the vertical distance from the keel (when horizontal), and the width is the width of the base tube. The "% Keel" specifies where the apex attaches to the keel (a percentage of the keel's length starting from the nose). The "Rake" parameter is how far (horizontal distance) the base tube is from the apex. The cross bar (X Bar) is specified in percentage of the leading edge and percentage of the keel (both from the nose). The "z from Keel" parameter is the "float" height above the keel. The kingpost is similarly located at a percentage of the keel from the nose, and the height and "Rake" (horizontal offset of tip) can also be specified. The trailing edge of the sail is again specified as a percentage of the keel (again from the nose) and so is the rear wire attachment point. Right now there are only 2 battens (root and tip), and they're drawn with bezier curves in the Battens panel (selector at top). The segments setting determines how many samples (lines) are taken from those two bezier curves to make the wing. The "Refresh" button (at the bottom) will copy all of the current settings to a JSON format in a text box. You can then copy that text and save it for future rendering (see example below).

There's also a batten design panel where you can define a bezier curve for the root chord section and the tip chord section. The root is on the top, and the tip is on the bottom (although you can drag them around to confuse yourself if you want).


The rotation of the battens is relative to the keel, and they will be scaled so that the root batten fits between the front keel point and the trailing edge point (percentage of the keel). The tip batten will be scaled by the same amount, so they will remain relative to each other as drawn.

Here's the JSON specification for this particular model:

{
 'Version': 1.0,
 'KeelFrontPoint': [0,5,0],
 'KeelRearPoint': [0,-5,0],
 'WingTipPoint': [15,-3,0.2],
 'TrailingEdgePercentKeel': 80,
 'RearWirePercentKeel': 82,
 'ControlBar': {
   'Height': 3.5,
   'Width': 3.2,
   'ApexPercentKeel': 40,
   'RakeDistance': 1.0,
   },
 'CrossBar': {
   'PercentKeel': 33,
   'PercentLeadingEdge': 60,
   'HeightAboveKeel': 0.1,
   },
 'KingPost': {
   'PercentKeel': 40,
   'Height': 3,
   'RakeDistance': 0.1,
   },
 'Battens': {
   'NumPerSide': 2,
   'BattenList': [
     {
     'PercentLeadingEdge': 0,
     'Segments': 20,
     'LeadingEdge': [50.0,165.0],
     'LeadingEdgeControl': [125.0,20.0],
     'TrailingEdgeControl': [275.0,150.0],
     'TrailingEdge': [685.0,160.0]
     },
     {
     'PercentLeadingEdge': 100,
     'Segments': 20,
     'LeadingEdge': [225.0,305.0],
     'LeadingEdgeControl': [240.0,295.0],
     'TrailingEdgeControl': [260.0,300.0],
     'TrailingEdge': [350.0,315.0]
     }
   ]
 }
}

That shows exactly how the glider will be built, and that's a specification that can be easily shared. If you build any gliders, please post the JSON for it so other people can use it as well. Here's the whole application (in frame design mode):


There are a lot of improvements that I can make in the future, but this is a good start. Have fun, and please post the JSON for any models you make.

I finally got around to making a pilot ... or at least a reasonable approximation:

Back in the day I was selling plans for a scale model of my Skysail for $2. Some who bought the model plans were also thinking of building the full size version. However, I don't know of anybody who did.

I exhibited one of those original model plans at the 40th anniversary gathering at the HG history marker site in Newport Beach in May, 2011. Ken DeRussey also has a set in his HG museum.

Frank

Frank Colver wrote: ... plans for a scale model of my Skysail ...

If you can find some rough measurements, I'll be happy to build a model.

Here's a model of my Wills Wing Falcon 3 195 with some guesses.

I looked through the manual and couldn't find:

1. Distance from nose to crossbar junction along leading edge
2. Length of kingpost
3. Angle of kingpost to the keel
4. Angle of control bar to the keel

Here's the data that you paste into the program:

{
  'Version': 1.0,
  'KeelFrontPoint': [0,5,0],
  'KeelRearPoint': [0,-6.1,0],
  'WingTipPoint': [16.3,-4.9,0],
  'TrailingEdgePercentKeel': 80,
  'RearWirePercentKeel': 83.8,
  'ControlBar': {
    'Height': 4.9,
    'Width': 4.65,
    'ApexPercentKeel': 49.5,
    'RakeDistance': 1.0,
    },
  'CrossBar': {
    'PercentKeel': 33,
    'PercentLeadingEdge': 55,
    'HeightAboveKeel': 0.1,
    },
  'KingPost': {
    'PercentKeel': 50,
    'Height': 4,
    'RakeDistance': 0.1,
    },
  'Battens': {
    'NumPerSide': 2,
    'BattenList': [
      {
        'PercentLeadingEdge': 0,
        'Segments': 20,
        'LeadingEdge': [50.0,165.0],
        'LeadingEdgeControl': [125.0,20.0],
        'TrailingEdgeControl': [275.0,150.0],
        'TrailingEdge': [685.0,160.0]
      },
      {
        'PercentLeadingEdge': 100,
        'Segments': 20,
        'LeadingEdge': [225.0,305.0],
        'LeadingEdgeControl': [240.0,295.0],
        'TrailingEdgeControl': [260.0,300.0],
        'TrailingEdge': [350.0,315.0]
      }
    ]
  }
}

Here's an animation of the model it produces:

Here's a model of the Colver SkySail that I made from Frank's' drawings:




Here's how it matches up with Frank's original drawings:



Finally, here are the settings I used to build that glider. I used "pixels" as the unit of measurement, so the scale should be about 1.0 to view it. Frank's drawings show the scale as 1 foot per 0.2 inches, but I haven't factored that into the model yet.

{
  'Version': 1.0,
  'KeelFrontPoint': [0,135,0],
  'KeelRearPoint': [0,0,0],
  'WingTipPoint': [353,-22,18],
  'TrailingEdgePercentKeel': 90,
  'RearWirePercentKeel': 100,
  'ControlBar': {
    'Height': 86,
    'Width': 63,
    'ApexPercentKeel': 60,
    'RakeDistance': 15,
    },
  'CrossBar': {
    'PercentKeel': 66,
    'PercentLeadingEdge': 60,
    'HeightAboveKeel': 10,
    },
  'KingPost': {
    'PercentKeel': 80,
    'Height': 65,
    'RakeDistance': 0.1,
    },
  'Battens': {
    'NumPerSide': 2,
    'BattenList': [
      {
        'PercentLeadingEdge': 0,
        'Segments': 20,
        'LeadingEdge': [100,200],
        'LeadingEdgeControl': [188,75.166],
        'TrailingEdgeControl': [415,174.166],
        'TrailingEdge': [649,200]
      },
      {
        'PercentLeadingEdge': 100,
        'Segments': 20,
        'LeadingEdge': [100,400],
        'LeadingEdgeControl': [158,429.166],
        'TrailingEdgeControl': [226,413.166],
        'TrailingEdge': [416,400]
      }
    ]
  }
}

If you want to view a 3D rotatable model, copy that specification and paste it into the "Glider Data (JSON format)" text box at http://ushawks.org/designer.Then click the "Update Model from Text" button to copy it into the model. You'll also have to switch into "Batten" mode and switch back out to cause the batten calculation to be made. When I originally wrote the program, I figured you'd only be changing the battens from "Batten" mode, so that's when it does the batten curve calculations. I need to have it make those calculations when the model is updated via text as well.

Bob Kuczewski wrote:You'll also have to switch into "Batten" mode and switch back out to cause the batten calculation to be made. When I originally wrote the program, I figured you'd only be changing the battens from "Batten" mode, so that's when it does the batten curve calculations. I need to have it make those calculations when the model is updated via text as well.

I just updated the code to do that automatically, so when you paste in a model and click "Update Model From Text", it should update the model and the battens.

I also updated the Colver SkySail model to be in feet rather than pixels so it will come up scaled nicely. Here's the updated model:

{
  'Version': 1.0,
  'KeelFrontPoint': [0,6.1,0],
  'KeelRearPoint': [0,0,0],
  'WingTipPoint': [15.9,-1.0,0.8],
  'TrailingEdgePercentKeel': 90,
  'RearWirePercentKeel': 100,
  'ControlBar': {
    'Height': 3.87,
    'Width': 2.83,
    'ApexPercentKeel': 60,
    'RakeDistance': 0.67,
    },
  'CrossBar': {
    'PercentKeel': 66,
    'PercentLeadingEdge': 60,
    'HeightAboveKeel': 0.45,
    },
  'KingPost': {
    'PercentKeel': 80,
    'Height': 2.92,
    'RakeDistance': 0.01,
    },
  'Battens': {
    'NumPerSide': 2,
    'BattenList': [
      {
        'PercentLeadingEdge': 0,
        'Segments': 20,
        'LeadingEdge': [100,200],
        'LeadingEdgeControl': [188,75.166],
        'TrailingEdgeControl': [415,174.166],
        'TrailingEdge': [649,200]
      },
      {
        'PercentLeadingEdge': 100,
        'Segments': 20,
        'LeadingEdge': [100,400],
        'LeadingEdgeControl': [158,429.166],
        'TrailingEdgeControl': [226,413.166],
        'TrailingEdge': [416,400]
      }
    ]
  }
}

That's based on the assumption that Frank's original paper was 5 squares/inch. Frank's drawing states: "SCALE: 1 FOOT = 0.2 INCH". If you have a physical copy of the page, then it's easy to verify the size of the squares, but an image could have been any original size (and drawing paper can come with different squares per inch). But in order for the scaling to make sense, the height of his glider (from base tube to undersurface) shows about 4 squares which would be 4 feet. So his paper was likely 5 squares per inch which is a common format. So measuring tip to tip, that gave me a scale factor of 32 feet per 712 pixels. That's what I used to make the model in this post.

It's also interesting to note that I designed this tool to make it easy to design models from scratch. That's why so many items are percentages rather than actual distances. That makes it easier to fiddle with in an abstract way. But my experience with building the Falcon and Frank's SkySail (actual gliders) showed me the down side of that approach. When building a model from actual measurements, it's easier to work in real distances (feet) rather than percentages. So I may consider some adaptation for that case.