Tuesday, 26 March 2024

Drawing formers and ribs for a model airplane with Excel

26 March 2024

How to easily draw formers and ribs for a model airplane that looks about right?

MS Excel, Insert, Illustrations, Shapes: Lines, Curve; Basic shapes, Oval.

Curve can create rib shapes, planform of fuselage.

Oval can create elliptical formers, planform of wings

Ask Gemini for more info.



Thursday, 21 March 2024

AutoGyro

21 March 2024

I watched on Youtube a video of indoor autogyro. They are hard to set right and even when they are trimmed, the performance is mediocre at best. When some of them tripped against the wall or ceiling, most does not recover. Of those that flew without accidents, their ascent is steep with the propeller thrust and their glide descent is also steep, looks like they are using relatively torquey rubber motors.  They mostly have 3 to 4 blades of narrow chord on a rotor that is balanced and spins well with just a wave of the model. In comparison, normal airplane models seem so much easier.

Here's an idea for the design of a 4 bladed autogyro rotor for rubber-powered indoor flights.
It appears that the autogyro can be trimmed by increasing or decreasing the angle of attack.
On a clock-wise spinning autogyro, when viewed from top, twisting the blades to increase or decrease will result in variation to the unequal lift, thus changing the direction of the flight like a flat wing's ailerons. Most autogyros have the rotor at an angle of attack to the flight path, some dihedral also for slight stability, largish horizontal stabiliser and fins but the overall length of the fuselage is short, thus it isn't too effective, especially for directional stability, thus requiring fine adjustment of the CG. The 'roll' though can depend on the previously mentioned blade adjustment, but if the model has horizontal stub wings, that can increase the adjustment ease and aid in roll and pitch stability.

To get a rotor that I can adjust the angle of the blades, the fixation to the hub need to be adjustable, here's how:
The hub can be a piece of firm foam, either a circular disc or a square disc. Cut slots to the disc, dissecting at 90degrees for a 4-bladed rotor or at 120 degrees for a 3-bladed rotor.
The slots will accept the blades bamboo rods and by friction alone, allow adjustability. Glue the rods to the hub only when satisfied with the dihedral and blade's angle of attack.
The top and bottom hub surfaces can be reinforced with 1/32" ply pieces or just plastic pieces. Insert a plastic tubing through the centre of the hub.

As re-written by Gemini:

Adjustable Rotor Design for Indoor Rubber-Powered Autogyro

This design proposes a simple and adjustable rotor system for a lightweight, indoor rubber-powered autogyro. The key feature is the ability to fine-tune the angle of attack for each rotor blade, allowing for precise trimming of the aircraft.

Core Components:

  • Hub: Constructed from lightweight but firm foam (disc shape - circular or square). Slots are cut into the disc at 90-degree intervals for a four-bladed rotor (or 120 degrees for three blades).
  • Blades: Bamboo rods are inserted into the slots, allowing for adjustments in blade angle. Friction holds the blades in place initially.
  • Reinforcement: Thin ply or plastic pieces are glued to the top and bottom of the hub for added strength.
  • Drive Mechanism: A plastic tube is inserted through the center of the hub to accommodate the rubber band drive mechanism (not described here).

Adjustability and Trimming:

  • The friction-based slot system allows for easy rotation of each blade, enabling adjustments to the angle of attack. This directly affects the lift generated by each blade and is crucial for trimming the autogyro for stable flight.
  • Once the desired blade angles and dihedral (optional) are achieved, a small screw or locking pin can be added at the base of each slot to secure the blades in place.

Considerations:

  • Balancing: Adding or removing material from individual blades for angle adjustments can disrupt the rotor's balance. To maintain balance, consider marking or numbering the blades for re-installation in their original positions.
  • Dihedral Angle: While 90-degree slots allow individual adjustments, achieving a true dihedral effect (where opposing blades have a slight upward angle) might be challenging. Pre-cutting the bamboo blades with a slight dihedral angle before insertion could be an option.

This design offers a simple and effective method for achieving adjustable blade pitch in a lightweight indoor autogyro. With some refinements for securing the blades and potentially achieving a dihedral angle, this concept has the potential to improve the flight stability and controllability of your rubber-powered aircraft.

5 March 2024

Designing an Autogyro Rotor

Here's an outline for the design of a simple autogyro rotor for a rubber-powered model plane.

Balancing Lift and Area:

  • The autogyro needs less wing area compared to a fixed-wing aircraft for similar lift due to its spinning rotor.
  • While the target wing area for the model plane is 24 square inches (12" span, AR 6), the rotor blades can have a combined area of around 16 square inches.

Rotor Design:

  • The rotor will have two 8" x 1" balsa wood blades, totaling 16 square inches, which is 2/3 the wing area and has a span of 167%. The rotor disc area is however, much larger.
  • A 4" "zero lift" hub will connect the blades, resulting in a total rotor span of 20 inches.

Hub Construction:

  • The 1/2" wide hub will be made of three parts: a 4" horizontal balsa beam, a flexible plastic strip 6" long, and a short 1" length of plastic tube.
  • The plastic tube will be glued perpendicularly to the horizontal beam to serve as the central pivot point.
  • Stoppers, which can be strips of 1/16" balsa, will be attached to the plastic strip to create a slight upward angle for the blades (static coning) and enhance stability.

Blade Construction:

  • Two airfoil blades will be made from 1/16" balsa wood.
  • The trailing edge of the blades will be thinned to 1/32" to 1/64" for the rear half to improve airflow.
  • The leading edge will be sanded to a slight taper for the front quarter with a slight radius underside.

Assembly:

  • Chamfer the ends of the horizontal beam and glue the plastic tube perpendicularly.
  • Glue stoppers to the flexible plastic strip at a slight angle.
  • Punch a hole in the center of the plastic strip and insert it through the bottom of the horizontal beam.
  • Apply glue and secure the stoppers against the chamfered edges, then let dry.


Building the Rest of the Autogyro:

Tail Assembly:

  1. Motor Stick: Cut a 10" x 1/4" x 1/8" balsa motorstick.
  2. Motor Bearing: From an aluminum can, cut a strip and bend one end to form a bracket, use a pin so that the propeller shaft can pass through. Bind and glue this to the motorstick as a motor bearing.
  3. Tail Hook: Bend a Z-shape from a paperclip and glue it to the motorstick.
  4. Tail Feathers:
    • Fabricate the horizontal stabilizer and vertical fin from 1/16" balsa square strip.
    • Cover one side of each with tissue paper.
    • Glue them to a 4" long, 1/16" balsa square tail boom.

Rotor Mount:

  1. Mast and Support: Cut a mast and overlapping angle support from 1/8" balsa square strip.
  2. Shaft: Straighten a paperclip and cut a 1.75" length.
  3. Attaching the Shaft: Glue 1/2" of the shaft to the top of the mast, leaving 1.25" protruding.

Final Assembly:

  1. Glue the tail boom to the motorstick's rear end.
  2. Assemble the propeller and rubber band on the motorstick.
  3. Find the balance point of the motorstick.
  4. Glue the rotor mount to the motorstick, aligning the mast top with the previously identified balance point.
  5. Insert the rotor onto the protruding wire shaft, add a glass bead, and glue another bead on top to secure it.

19 February 2024

Teetering Offset Hinge for 2-bladed Indoor Autogyro
Have the 2 bladed rotor prepared.
They are two identical airfoil blades connected to a rectangular strip of plastic. 
Mark out the centre of the rotor and draw a slanted line at 30-45degrees on the plastic strip.
Crease on the slanted line and bend up 5-10degrees on each blade, this will the 2 blades to form a cone angle, works like dihedral and decalage and each blade will have a negative incidence of 5-10 degrees.
Fold a rectangular strip of aluminum (from aluminum can) into a 'M' shape.
The spin shaft pass through the 'V' in 'M' and close the 'V' until it is back to back, glue the shaft in position.
Bend the two 'I' in the outerflaps so that it flares upwards.
Glue the outer flaps to the centre of the rotor along the slanted line.
Now we have a rotor with a perpendicular shaft that will spin freely in the autogyro model plane.

5 February 2024

This is an idea of construction for a small hub for 2 bladed rotors that may be suitable for small and very light autogyros. The rotor is teetered and hinged so that there is some form of correction for asymmetrical lift.

Material list:

  • Paper cut to hub-shape
  • Right angle balsa/foam spacer pieces 
  • small plastic/paper tube
  • thin pvc sheet that is flexible and springy for 
  • shaft either of CF or paper clip
  • bearing made from beads and or thin plastic/metal washer 
  • Disc stoppers made with paper/paper card/plastic/wire/rubber etc.

Draw a big T on the plan. Cut 2 pieces of paper to the shape of the hub. Lay 1 piece of hub-shaped paper on the plan. Cut a piece of tube to the horizontal hinge length. Insert a temporary shaft into the tube and lay it over the hub-shaped paper. Cut a piece of tube to the vertical tube length. Insert another temporary shaft into the tube and lay it over the hub-shaped paper. Cut 2 spacer pieces to shape and glue to 1)hub-shaped paper, 2)horizontal and vertical tubes. Sand spacers flush to top of the 2 tubes and glue another hub-shaped paper over it all. Remove the temporary shafts. Cut the horizontal tube and remove the centre portion which will be the pivot tube. Cut the thin pvc shape and glue the pivot tube at the offset angle in the centre of the thin pvc shape. Glue the blades on the thin pvc strip. 

On the actual shaft, glue a stopper, insert the vertical tube, bearing and stopper. Glue the blades to the thin pvc strip. thread short shaft through the horizontal tube and secure the ends from falling out.

Now, this is building up the rotor hub, if we used paper clip as the shaft, we could bend to shape and if it retains under stress, it would be simplified.

Material list:

  • small plastic/paper tube
  • thin pvc sheet that is flexible and springy for 
  • shaft from paper clip
  • bearing made from beads and or thin plastic/metal washer 
  • Disc stoppers made with paper/paper card/plastic/wire/rubber etc.
Cut the thin pvc shape and glue the pivot tube at the offset angle in the centre of the thin pvc shape. Glue the blades on the thin pvc strip. This is the rotor.
Bend the paper clip; first a 90 degrees bend, another 90 degrees bend so that it runs parallel, and another 90 degrees bend so that it is perpendicular. Maybe the bent shape will open up, maybe it will remain. Insert stopper, bearing, horizontal pivot tube, bearing and another stopper. Then the assembly is inserted to the vertical pivot of the model and add bearing and stopper. 

Another method

Material list:

  • Balsa block/laminate
  • Aluminum brackets from aluminum can
  • small plastic/paper tube
  • thin pvc sheet that is flexible and springy for 
  • shaft from paper clip
  • bearing made from beads and or thin plastic/metal washer 
  • Disc stoppers made with paper/paper card/plastic/wire/rubber etc.
Make a rectangular hub base from balsa block or laminates of thinner balsa. Insert and glue a straight length of paper clip, perpendicular to the hub. Fabricate 2 identical brackets from aluminum can or pivot tube from plastic tube with spacer, glue to ends of hub. Measure the distance between the brackets and cut the pivot tube. Cut the thin pvc shape and glue the pivot tube at the offset angle in the centre of the thin pvc shape. Glue the blades on the thin pvc strip. This is the rotor. Offer the rotor to the brackets and insert a paperclip shaft.
Insert vertical shaft to the vertical pivot of the model and add bearing and stopper.

Another method, using paper

Material list:

  • Paper
  • small plastic/paper tube
  • thin pvc sheet that is flexible and springy for 
  • shaft from bamboo/toothpick/carbon-fibre rod
  • bearing made from beads and or thin plastic/metal washer 
  • Disc stoppers made with paper/paper card/plastic/wire/rubber etc.
Cut the paper to hub template. Fold in half. Glue the shaft and pivot tube. Cut to release the central pivot tube. Cut the thin pvc shape, glue the blades. Glue the central pivot tube at the offset angle in the centre of the thin pvc shape. This is the rotor. Offer the rotor to the pivot tubes in the paper hub and insert another shaft.
Insert vertical shaft to the vertical pivot of the model and add bearing and stopper.

29 December 2023

Spinning propeller. spinning rotor
Keep the disc of propeller or rotor spin true, or just less wobbly. That means the bearing holes must have some play but not too much and be spaced a good distance apart.

Bearing structures could be made from thin aluminium sheet with the holes pierced then folded to the hub.
The hub may be 2 identical sheet balsa pieces to anchor and spaced apart the folded bearing structures.
Since the hub is sheet balsa, it will provide good gluing surfaces to the 2 blades.
Distance between bearings, the "height" or "depth" of sheet hub, may be 1/4" for 3-7" prop and 1/2" for 20" rotor. 

Propeller shaft is perpendicular to sheet hub. Rotor shaft can be angled to create a conical disc.
Propeller and rotor shafts may be flat to glue directly onto the sheet hub or they may be thin carbon fibre rods so that the angle of attack may be adjusted. If adjustable rods are used, the holding structure may be tissue/paper, pressed over the rod and glued with superglue on to the sheet hub.

To fold the bearing structure, a former can be used. A former can be 2 pieces of hardwood (ice cream stick), with one edge square and a pin perpendicular to the edge to locate the hole in the bearing structure. Then a plier to fold the aluminium over the same edge.

The blades may be like a paddle where there is no twist between root and tip. Or, the blades may be twisted to form something of a helical pitch.

There is no perceivable benefit to have tip weight to the propeller blades but it may be beneficial to have tip weight to the rotor blades to introduce some momentum stability while the rotor spins.

Rotor constructed above does not have flapping hinge.
Thin rotor blades may bend under flying load and cause something like flapping hinge but there's hardly any control of the hinge axis.
Flapping hinge is beneficial to rotor blades to avoid roll-over.
The flapping hinge may be set at an angle to introduce negative angle as the blade hinges up.
The limit of hinge may be a pair of simple thin aluminium bracket at the flapping point.

6 July 2023

Foam blades with CF rod as spars
  1. Blades from foam dinner plates. 
  2. Glue CF rod to blade.
  3. Balance.
Cardboard hub
  1. Draw and mark hub strip on cardboard. 
  2. Punch holes according to mark.
  3. Fold ends to form isoceles triangle.
  4. Make aluminium bearings from drink cans. 
  5. Cap the top of the isoceles triangle cardboard hub with the top folded aluminium bearing and add the bottom bearing to the base of the isoceles triangle.
Assembly
  1. Insert CF ends of the blades through the isoceles triangle hub.
  2. Insert Shaft through the bearings.
  3. Adjust and balance.
If the isoceles triangle is used with the base as the top, we have a flat top bearing surface which is better since the lifting/dragging force pushes against the top stop of the shaft. The sides are used as tensile members. The ends of the sides can be lapped to the counterpart side.  

3 July 2023

Paper blades
  1. Rectangle blanks, equal lengths, width at 150% of chord. 
  2. On underside, mark at 50% chord along span.
  3. Crease fold the longer 100% of chord width, the crease line will be 50% and the long edge meets the marked 50% chordline.
  4. Slit along marked chordline, glue paper tabs to secure, top of chord and bottom of chord. 
  5. Cut 1/32" balsa spars to width of 40% chord.
  6. Insert balsa spars to paper blades' pockets, ensure they are identically placed in each blade.
  7. Spot glue the balsa spars to the tip and root of the paper blades.
  8. Join the balsa spars to the hub and a rotor disc is formed. Could use the balsa spars itself to make the bottom hub.

22 March 2023

Today I am thinking about leverages for pitch roll and yaw stability.
If the rotor is flat, it can't provide any leverage.

Pitch stability
It could be an elevator at the back or a lifting plane at the front. Which is more effective? In the front lifting plane, it would be at the verge of stalling. Front fin would destroy yaw stability, so my choice would be a rear elevator.
A wing can be pitch stable itself, this involves having lifting portion and stabilising portion of the wing front and rear respectively.
A rotor with a dihedral angle will have this form of pitch stability as well.

Yaw stability
It would be a fin at the back. 

Roll stability
It could be by the center of gravity, but more likely some dihedral is required so that together with the fin, there will be less slip.
A dihedral rotor ought to have roll stability too, but I am thinking it is not constant and need supplementation.
To supplement, I think the pitch and roll stability can be enhanced with a lifting plane just in front of the rotor.
So actually, the rotor could be flat without any dihedral if the pitch and roll stability are mainly by a lifting plane. Using the spinning rotor as a additional lifting/dragging device. 

In a beautiful world, a conical rotor only need a fin (for yaw stability).

Differential thrusts changes the pitch, the yaw and affects the roll slightly.  

3 March 2023

Paper blades and ribs.

Ribs: Glue halved paper ribbons to spar.
Blade area: Cut blanks, do a fold, glue to spar and ribs.

That way the ribs can be adjusted for incidence and curled for airfoil.

1 March 2023

Conventional: Tube bearing, bent up tabs, strings and disc

As many rotor blades as needed is made identical with longish spars. 
At root end of each spar is secured an aluminum tab from aluminum can.
Have the central plastic tube surrounded by the aluminum tabs and secured the exposed aluminum tabs against the tube and open up the spars like petals around the pistil. Then glue the blades on each spar, secure with a central disc and adjust.
If the blades are already on the spars, just arrange them with a jig on the tip of the blades.

A disc of card, foam or anything like that with a central hole, is used to lock the tube's perpendicular position, the position of each blade around the tube, and the dihedral angle.

Spars can be made from balsa strips, or carbon rods.
Tabs or joiners to the tube, can be aluminum sheet as mentioned above, or cotton thread.

28 February 2023

How about a cone as the rotor?  It wouldn't spin. What makes it spin? Oblique airflow pushing the blades forward. So what is needed? Some openings in the cone for the air to rush by obliquely? Or just some negative incidence? But there appears to be zero incidence blades and why not?

Try a paper cone with 50% cutouts, the root and the tip will be solid rings. At the top is paper cross pattern to locate the central hub which is a paper tube. Acting like flat blades and adding area to rotor surfaces.

An inverted cone = some camber, more chord = more camber. Very draggy.  But autogyro is draggy and drag contributes lift?

Will location of opening affects anything? It's a 360 conical disc, why should it? Shape of openings? Is circular holes, oval holes, triangular holes, half circles one way or the other way or root or tip? 

27 February 2023

Hub for 3 or 4 bladed rotors, with a different assembly
  1. Cut out the top and bottom aluminium shape, punch hole in the centre. In case it is a 4 bladed rotor, the hub piece is a cross-shaped sheet. If it is to be a 3 bladed rotor, then there is 3 spokes.
  2. Make a work base that will support the spindle, glue the work plan over the work base.
  3. Lay the bottom hub piece through the spindle, over the plan and line it up to the work plan.
  4. Glue the arms over the hub piece (don't need chamfering, think of it as adjustability).
  5. Bend up the arms to dihedral height over the work base, use a dihedral jig. Start with one arm, then turn and do the bend on the next arm until all arms are bent to the dihedral jig.
  6. Remove the arms and bottom hub.
  7. Lay the top hub upside down through the spindle, over the plan and line it up to the work plan.
  8. Bend up each spoke of the top hub to a jig, fold down the tab and remove.
  9. Re-insert the arms and bottom hub to the spindle.
  10. Drop the top hub to the spindle and glue the tabs to the arms.
  11. Strip identical spacer strips (same height as the jig) and glue to top of the arms and the bottom of the top hub. Alternatively, a tube can be used but it shall be inserted between steps 9 and 10.
  12. If the top spokes are too weak, glue balsa strips to reinforce the top diagonal spokes.
  13. Remove assembled hub from work base.

Hub for 3 or 4 bladed rotors, with a round tube as upright
  1. Cut out the top and bottom aluminium shape, punch hole in the centre. In case it is a 4 bladed rotor, the hub piece is a cross-shaped sheet. If it is to be a 3 bladed rotor, then there is 3 spokes.
  2. Make a work base that will support the spindle, glue the work plan over the work base.
  3. Lay the bottom hub piece through the spindle, over the plan and line it up to the work plan.
  4. Glue the plastic hollow tube to the hub piece, vertically.
  5. Glue the arms over the hub piece (don't need chamfering, think of it as adjustability).
  6. Bend up the arms to dihedral height over the work base, use a dihedral jig. Start with one arm, then turn and do the bend on the next arm until all arms are bent to the dihedral jig.
  7. Chamfer a piece of balsa sheet to meet the vertical tube and the arm, strip to make the corresponding number of diagonal struts.
  8. Glue the diagonals to the work piece.
  9. Lay the top hub piece through the spindle, over the tube and bend to meet the diagonals and glue in place, 
  10. Remove assembled hub from work base.

Hub for 3 or 4 bladed rotors, built up 
  1. Cut out the top and bottom aluminium shape, punch hole in the centre. In case it is a 4 bladed rotor, the hub piece is a cross-shaped sheet. 
  2. Make a work base that will support the spindle, glue the work plan over the work base.
  3. Lay the bottom hub piece through the spindle, over the plan and line it up to the work plan.
  4. Glue uprights to the hub piece.
  5. Lay the top hub piece through the spindle and over the uprights. Bend to meet the uprights and glue in place, 
  6. Glue the arms over the hub piece (chamfered for dihedral).
  7. Bend up the arms to dihedral height over the work base, use a dihedral jig. Start with one arm, then turn and do the bend on the next arm until all arms are bent to the dihedral jig.
  8. Chamfer a piece of balsa sheet to meet the upright and the arm, strip to make the corresponding number of diagonal struts.
  9. Glue the diagonals to the work piece.
  10. Remove assembled hub from work base.

10 January 2023

hmm.... just a criss and have the blades as the cross. Reinforce with aluminium strips for the bearing points.

6 January 2023

Foam blank with 45degrees bent to provide negative incidence and dihedral
Tin can aluminium strips with double 45 degrees bend to support the blades at top and bottom.
The bottom strip may be bent at the axle hole or bent into a curve.
The axle passes through the bottom strip, foam blank, and top strip, anchored at the bottom, guided by a wire loop, through the rotor assembly, a single bead and secured at the top with a 90degrees bent or stopper.

Foam blank may be extended with paper sails supported by strips which are glued to foam blank, treat the foam blank as a spar.

6 December 2022

No suitable nozzles? How about carbon rods, paper clips and some ready at hand stuff? 
  • A short piece of carbon for the blade shaft and at both ends, tie a Z bent paper clip
  • Glue blades on the Z bent wire flat over the table
  • Bend for dihedral and twist for incidence
  • Glue the spindle rod perpendicular to the center of rotor, reinforce with thin card patch or other means.

5 December 2022

Re-using Covid ART nozzles for spinning Hub and dihedral blades. 2 blades rotor.

If the chosen carbon fibre rod fits tightly in the nozzle hole:
  • ice cream stick with hole in middle
  • thin aluminum strips to ends of ice cream stick and blades
  • Bend the aluminum strips to give equal dihedral and angle of attack 
  • Covid ART nozzle hotglued to ice cream stick, over the hole, I prefer bottom.
  • insert carbon fibre rod through the nozzle and the ice cream stick
If the rod spins freely (but not too wobbly) in the hole in the Covid ART nozzle kit, insert the rod through one nozzle, sandwich the icecream stick and then another nozzle. Once satisfied, glue the nozzles to the ice cream stick.

11 November 2022

Spinning Hub
  • A piece of round or square cardboard with a centre pivot hole. Or a pyramid formed from 4 triangles or ring of 4 triangles, so that this pivot hole is set above the cross-piece pivot hole. (can also have a square cardboard supported by 4 perimeter cardboard spacers or a central single piece spacer with a larger hole in the centre) 
  • A piece of cardboard cut to a cross shape with a centre pivot hole.
  • The 4 arms are bent to 15 degrees angle, superglue applied to lock the 15 degrees dihedral and decalage.
  • The ends of the 4 arms are cut to 15 degrees angle.
  • The pivot holes are cored to fit the shaft and reinforced with superglue, since they are made from cardboard.

Flapping Blades
  • Root of each blade (4 in total) is a small triangle with a 15 degrees cut to match cross piece.
  • Leading edge is a piece of balsa or foam, glued to the root piece.
  • Root chord and tip chord is balsa or bamboo.
  • Covered on top with 70 g/sm photocopier paper
  • Flapping hinge is a trapezoidal piece of flexible foam.
Assembly:
  1. Pin the hub in the centre pivot hole, use the same 15 degrees jig to raise and superglue the 4 arms so that the arms is raised to 15 degrees. When one arm is set, rotate the hub and do the next arm.
  2. Remove pin, place scrap pieces of flexible foam under the cross piece and re-pin. This raises the entire cross piece, letting the blade to slid under,  
  3. Glue flapping hinge to underside of each blade, slip and glue the flapping hinge under the cross piece, make sure a gap of 3mm is between the cross piece and blade root.
  4. The 15 degrees jig will support the blade while the glue set.

23 September 2021

Sure Fire Autogiro
Rake: 3/8" over 6"
Slant: nil
Dihedral (each blade): 2.3/4" over 13"
Angle of Attack: 0
Blade x2 clockwise: 11" x 2.3/8" paper
Downthrust: similar to Rake
With horizontal stabiliser

Cierva Autogiro
Rake: nil
Slant: nil
Dihedral (each blade): nil
Angle of Attack: negative 1/8" over chord
Blade x4 anti-clockwise: built up
Downthrust: nil
With roll paddle and elevator

Can't find more.

If blades can be set zero, one less to get wrong.
Rotor hub can be made from a piece of flat wood, paper clip bent to give dihedral where blades are patched.
Blades can be balsa built up with paper


22 September 2021

How about plastic tubes for the hub? If doing a very light version, bent plastic tubes from cotton bud may be enough. 

Bend 2 lengths of plastic tubes in their centre by crimping. (the dihedral effect)
Glue bent plastic tubes at right angle to each other. (4 rotor blades)

Prepare jig to hold the crossed plastic tubes evenly.
Example of a jig: one centre pin, 4 spacers to support the crossed tubes.
Pierce centre, enlarge for shaft.
Use a temporary shaft, tie the crossed tubes with thread and glue.

Prepare 4 rotor blades.
Using previous example of jig: add a triangle to one of the spacer. The triangle set the angle of attack of each blade. Can add blocks so that each blade will be located at the same place.
Glue each blade in turn.

Or, to make the angle of attack adjustable by gluing a bamboo stick to each blade and insert the bamboo into the tubes. Doesn't fit? Sand down before gluing to blade or split the tubes' ends.

A 2 blade rotor with doweled blades would be:
1. Pierce a hole in the centre of a plastic tube, enlarge hole to shaft diameter.
2. Bend up for dihedral, not in the centre of the tube, just a short distance away to avoid skewed hole.
3. Insert blade, adjust to angle. Just eyeball and spin the rotor.

Shaft shall be from a paper clip? It is easy to bend.

9 April 2018

On Sunday, I tried Version 2. Version 1 had a bamboo boom and vertical tail. Version 2 had foam fuselage, horizontal stabilizer and 2 tip fins and  affixed to the other components of Version 1.

I tried running start to spin the main rotor fast at full throttle. The initial launches rolled right and crashed within a foot away. I finally got a left arc about 2 m away after having adjusted the model to have left rudder, left weight, and left rotor. The 8.5mm motor and 56mm propeller seems incapable of overcoming the drag.

The fooling around stopped after I noticed my sewing pin is bent. Maybe next time, I can try rotors with flapping tips. From the anticlockwise prop (from front), anticlockwise rotor (from top), I can have left thrust, left weight, left rudder and left rotor. Or I am also thinking of normal elevator and rudder and if one rotor does not balance, twin rotors. I need also better pushrods then the soft wire can afford.


Version 2

These did not grip the thin soft wires I am using and the nuts get loose easily.


Version 1, as the motor stick is screwed, thrust direction is adjustable. The hotglued boom came loose easily and the bamboo size lacks torsion strength.



While making Version 1, I originally thought of making it a pusher.


5 October 2017

Previous was using plastic tubing, how about just beads?

This will limit the contact of the axle to the two points on the hub.







The ice cream stick option has the blades at 0 degrees. Pack with beveled pieces at the blade holding area, or instead of beveled pieces with balsa blocks that has angled saw cuts (similar to sketch on right), the blades will be held at an angle. Or, maybe we can leave it at 0 degrees, and use flat blades that have toe-in/toe-out near the tips by angling it upwards, which ought to propel the blades.

Or how about a spinning disc, spun by peripheral winglets? The centre area is slow anyway and can't possibly have much difference whether it is spinning slow, fast or not at all. When the disc stop spinning, it's still a larger flying surface than is possible with the normal blades' area.




Change the kite to an autogyro? I think a tractor propeller and a large tail is good.















1 February 2016

Even-Bladed Hubs

Too much theory only ends in theory and I get confused. Maybe just try the following. If I set the blades carefully, I could have some coning too.
2mm dowels from Daiso (Basswood?) for hub and blades leading edges.
2mm foam blades (from Daiso) are glued to the 2mm dowel leading edges and wrapped in "Scotch" tape.
Large heat-shrink sleeves to set the angles and dose of superglue when set is as desired.

How about a Cootie but with this blade hub?

16 September 2015

On the following chain of thought, of which all hypotheses are without any substantiation or even personal experience:
  1. Negative angle of attack is necessary to get the rotor spinning in the correct direction against the model's direction.
  2. Lift or drag of the rotor amounts to the same thing, it lifts the model so long as there is some element of vertical vector.
  3. Negative angle of the rotor blades is not too critical, even if it is at -45 degrees, there would still be 'lift' (point 2 above), except that we shan't go there because everyone will tell you it's wrong and I won't be needing that! (And also for point 12 below.)
  4. The angle the entire rotor is presented to the wind has to be positive. I think this is the most important aspect. If it were negative, the model sinks instead of rising.
  5. Combining point 3 and point 4, if there is not enough negative angle of the blades, at some point it could become a positive blade angle to the wind if the positive rotor angle is excessive. If a blade angle is positive to the wind, then it would retard the spinning rotor because it would attempt to spin in the opposite direction. The rotor would then stall, and hence a faster spinning rotor is safer than a slower one.
  6. Carrying on from point 5, this mean that the negative angle of blade must be greater than the positive angle of rotor.
  7. The torque of a free spinning rotor is unlike a motorised rotor. The torque is in the same direction as the rotor direction. Hence, if a tractor propeller has left torque, the rotor should spin clockwise when viewed from top because the blade going against the direction of the wind will create more lift, in this case, the left half of the rotor has more lift then the receding right half of the rotor.
  8. Since the rotor is functioning as a wing, dihedral would be desirable if we are using rudder to turn the model. Air flowing through the rotor will tend to cone the rotor in the correct direction, yet the faster it spins, the more centrifugal force of the blades will flatten the cone.
  9. Should the blades be like a normal flatbottomed airfoil or should it be mounted inverted instead? If it is mounted like a normal airfoil then the rotor would have more lift because of the downwash. If it is mounted inverted, we are having air directed upwards and therefore there would be less lift and more drag.
  10. Carrying on from point 9, the main objective should be to have the blades advancing with the least drag and therefore, a thin non-cambered foil is desirable. Something that has a smooth entry and exit.
  11. When the model flies faster, the rotor spins faster and the model rises, so downthrust is required to 'lift' the tail.
  12. If there is no elevator to lift the tail, the flyable CG range is very limited and the downthrust has to do the 'lifting' of the tail. There is no lift when downthrust equals negative angle of blade and hence the tractor propeller should be set at a greater negative angle.
  13. For a simple rotor, the negative angle of the blades shall be fixed. If the positive rotor angle is 8 degrees and we desire a negative effective angle of 3 degrees, this means the blades are set at 11 degrees and the effective downthrust angle is 3 degrees.
  14. When the propeller is lower than the rotor, and assuming it is not a very long way infront of the rotor, there is a upwards turning moment. Therefore the downthrust angle has to be set greater than 3 degrees. However, if we have a stabiliser set at a positive angle leewards of the propeller, the stabiliser would have some downthrust to pitch the model down when forward speed is increased.
  15. A simple autogyro with the following parameters: A coned rotor with blades set at 11 degrees negative is set 8 degrees positive (round disc don't really stall) to the datum line. Motor is set at 3 degrees negative and stabiliser at zero degrees.

Tuesday, 19 March 2024

A Farman Moustique

19 March 2024

Hmm, I didn't solder the motor up to the capacitor and didn't fly it. Now I don't even know where is this model.

The advantages of capacitor plane:
Use direct drive small prop, simplest.
Doesn't look scale size? Stick paper prop for static display, remove to fly.
It will be more durable than geared motor and larger prop, less work, weight and damage.
You like pushers? Absolutely no problem, at least much lesser than rubber powered.
You like jets? No problem, the small prop is hardly noticeable anyway.

For tractor, choose medium nose because direct drive small prop is so light, rubber powered would benefit from longer nose.
Snub nosed planes not suitable, the cowl block most of the propeller.
Use slightly longer prop, it will have more thrust and shorter power run. The long decreasing power cannot be avoided.
Hmm, shall I shop for more donor plane?

26 October 2022






Photo log

Model completed with electric model pushed into nose. Electric motor is taped between 2 wooden picks.

Close up of motor

The 3 cylinders will be glued to the motor to represent the Anzani 3 cylinder engine. The holder of the capacitor will be under the nose of the model, not a scale feature, but easy.



future flying site
6 October 2022

Model has been tissued. The silver tissue does not seem to shrink: when wet it loosened and expanded but when it became dry again, it doesn't appear to shrink smaller than the original size. I could repeatedly mist and dry to see if it improves.

Black tissue markings on the tissued model looks ok and oil-based pen was also used for thin line markings.


If I were to print on the tissue paper?
Will the tissue shrink when heat is applied and remain shrunk?

I have still to:
  • properly shrink the tissue for a neater appearance
  • mark the lines for ailerons
  • tissue on the cockpit outline
  • mount the motor
  • mount the capacitor
  • make the cabane
  • make the 2 wheels
  • make the 3 cylinders engine
  • make the tail skid 
  • make a pilot
I will try to use 3D pen to make the cabanes, wheels, engine, skid, pilot.  
Make outlines to represent them, so there will be lesser weight and drag

12 August 2022

I'll make another fuselage and pair of wings.
The wings made from sliced polystyrene is too flexible. I will use the 2mm compressed foam and I will join the 2 wings into a unit for strength.
This also requires a new fuselage because the old fuselage was for the wings to butt against the sides of the fuselage.
The new fuselage will be cut from 2 pieces of expanded polystyrene.
I had a problem cutting the planform of the expanded polystyrene, it wasn't neat. So for this time I will try some changes.
  1. Cut the rear taper from a single piece of 10mm foam. It will be a straight taper, the curve will come from joining the rear ends together.
  2. Cut the 10mm tapered foam to 2 pieces, square them up to make the left and right sides of the fuselage. Use a small piece of double sided tape to hold the 2 pieces together temporary, the taper sides are facing inwards.
  3. Thumbtack the card side templates to the flat side (outside) of the 2x10mm tapered foam. Square up the two templates.
  4. Cut the side profile and then separate the two 10mm tapered foam profiles.
  5. Using the same card side templates, cut the dihedral wing seating on the two 10mm tapered foam profiles.
  6. Join the two foam profiles to become the fuselage, join the rear ends together.
  7. As the intended motor mount is a toothpick, the nose of the fuselage is not tapered, so the curve for the nose planform is next cut. This ensures a solid foam nose.
  8. Tidy up the fuselage, put in spacers where necessary to achieve the curvature on the rear fuselage.
  9. Cover with tissue and decorate etc but leave the wing seating bare.
For the new wings.
  1. Cut a rectangular piece of 2mm foam, the centreline is where the two leading edges meet. Draw this centreline.
  2. Curl the entire piece.
  3. Cut along the centreline.
  4. Cut the wing roots using the card template of 10 degrees.
  5. Glue paper reinforcement where the rigging lines will pass through the wings.
  6. Lightly glue the wing roots together. 
  7. Cover the top only with tissue and decorate.
After the wing is glued to the fuselage, add tissue covering to camouflage the joints. Then add the pylon truss and then the rigging.

2 August 2022

Haha! Change change!
The last wire of the motor snapped off from the charging plug on the capacitor just now and I changed my mind and think it is easiest to have the motor and capacitor separate.

The motor could be heatshrunk to a wooden pick and then the pick can be inserted to the polystyrene fuselage nose. Not happy with the thrust line? Remove and pin again. The pin is to one side of the motor can, so the penetration point can be rotated around for fresh bite, it need the same entry point as the first attempt since there is nothing (like the capacitor) to restrict the anchoring points. 
Covering the motor (without the capacitor) with a fake engine would be easiest because I only have to deal with making a hole that allows the motor to pass through.
The capacitor could be held in a mini bracket/clip and hung underneath the fuselage. Lowered CG would be more stable and the capacitor on the clip could be slide fore and aft to allow fine adjustments. The wires would be hanging, but I think that's ok for a simple toy. 
Hold final adjustments? A bit of UHU Por.

Since both motor wires need soldering, I could also replace with a similar or even bigger motor for a bigger propeller and I could choose to have the propeller rotate either direction!

I think the model will be very light and need very little thrust. The current motor spins too long on the capacitor. I could use a lower pitch propeller? 

Just dreaming, no hands-on.

1 August 2022

This is for a capacitor powered model.

The outline is from a plan by Walt Mooney available from Outerzone. I print it to A3 paper and it's about 16-17" wingspan. It is definitely not the largish model that Mr. Mooney had designed.

It was quick work to cut the wings, tail, rudder and landing slabs from thin polystyrene sheets which I had sliced a long while ago. I kept the tail surfaces straight without the scalloped trailing edges because 1) the wings' were not scalloped, 2) I didn't fancy making so many scallops that have to appear uniform. It is very easy to use straight edge and NT cutter to cut them out. Make multiple slices to cut the foam without tearing.

I intend to curl the left and right wings before covering with tissue on their top surfaces. The horizontal tail will be top covered but the rudder and the landing structure will be covered both sides. 

For the fuselage, I went from building up from 1mm balsa to thicker polystyrene sheets to a chunky sheet split in 2 with internally tapered rear fuselage to now: 2 pieces of polystyrene blocks and hot wire cut like a solid balsa model. All I need is 2 templates of the side, and 2 templates of the planform (which could be half-planform). The polystyrene is so light that I think a built up model will be equivalent or heavier because of the glue. 

I don't know how I will assemble the capacitor and motor. Maybe I will glue them into a single sub-unit so they can be positioned for many models. My motor is wired up to use a pusher propeller, so I must remember to have left thrust. I would want the motor and capacitor be integral and rely adding weight to rear fuselage for balancing. The motor will protrude in a non-scale fashion. Currently I am thinking of placing the capacitor at 90 degrees to the motor, socket up, glued/secured on a thin ply or wood with a firewall for gluing to the fuselage. The business end is then covered slightly with dummy engines.