Flying Flea

25 July 2016

Yesterday, a big predatory bird swooped and attacked the Flying Flea repeatedly. At one time, it caught the wing, but theFlea managed to escape. Its talons went through the starboard wing of the Flea. Everyone was excited at the spectacle.

 I think some tape would be fine to patch it up.

 

 

 

 

4 July 2016

I flew the Flying Flea yesterday after I hotglued the undercarriage wire into the slot, clipped the propeller short and made a battery connector to receive the spade type of batteries. At the field I discovered that the elevator is inverted when I used my WLToy's V911 transmitter, so the Flying Flea was flown in 2 channel mode: Throttle and Rudder. It flew easily to a great height using minimal throttle. On descent with throttle off, I can see it bobbing its nose slightly, indicating a slight stall condition on the front wing. There was a marked veering to the right, perhaps that got to do with the fact that the motor was spinning a smaller propeller?

The second flight ended prematurely when it dove to the ground and broke off the motor and the fin and rudder tore clear off the tail. It is fascinating that it requires such small amount of thrust and it appears that the front wings have high lift capability.

When I get my FS-i6 transmitter, I shall try again. All it now needs is some dabs of hotglue to re-secure the motor and tail.

4 January 2016

26 December 2015

Yes it flew!

It was over-powered. I think less than 1/3 throttle is required for level flight. I heard the propeller coming on and off intermittently at the lowest throttle I can maintain in order to reduce altitude.

Rudder control was ok, but the torque effect was significant. Even though I had set up exponential on the elevator, it was touchy to use elevator. To climb add throttle, to descent reduce throttle.  It was too windy to do meaningful flight adjustments.

25 December 2015

With the WLTOYS 8mm geared motor, a small cell but no undercarriage, the model weighed 35gm.
The undercarriage I had on the old Flea, constructed of CF rod running in aluminum tubing and commercial 1.5" wheels, weighs more than 7gm. That meant a 20% increase in weight if I use that.

The 7mm geared motor is much smaller and lighter than WLTOYS 8mm (which is a 8.5mm?). That meant a 10% reduction in weight if I use that. Also, the propeller is much smaller, and I will have less problem with trimming the torque.

The model is designed for 6-7mm geared motors, the thrust of the 7mm geared motor ought to be enough. However, I had the mount slotted to accept the 8mm and the 7mm mounting involves a square section rod so I hot glued the 8mm to the model. Looking for a reason for this decision, apart from the hassle of making a mount for the 7mm, I think 8mm is beneficial to the CG's position and I can keep the 7mm for something else.

The main wing was warped, so I raised a lower corner, mist and weighted the balsa wing. After 3 tries, the wing was corrected. I then used the remaining kit supplied 1mm CF rod and cut two lengths to the width of the main wing's ply pieces. They were inserted through the vertical ply wing mounts and superglued in placed. I think this is better and lighter than using the 4 missing kit supplied screws.

As for the undercarriage, I will leave them off for now, i.e., not going to make them, until the model had its flights. I superglued (Daiso's) small plastic tube to the hubs of the broken main gears of V977; they shall be the light weight wheels to replace the missing kit-supplied wheels.

Transmitter set up, model set up, cells charged, need to try in a big field on a calm day. I had to review how to bind and set up the Flysky transmitter.

21 December 2015

http://www.banggood.com/TY-Model-NO_7-292mm-Wingspan-Wood-Park-Flyer-RC-Airplane-KIT-p-1004972.html
T'is the season to be building, fly la la la lah, la fly la lah
From Banggood, a small laser cut balsa kit that looks a lot like the Flying Flea. The kit was selling for about $23.
Scary, the way it came in.






The balsa sheets were bagged and inside the folded corrugated board. (single edge razor blades not included)









Missing parts? Don't see the pair of wheels.










It looked ok.











The 3 sheets of balsa were complete, so was the ply piece. There're no wheels, and I don't think the wires/rods were all there.

Whatever it is, I started building.







I started by colouring the balsa sheet.


Time: 0:00







I drew the lines and coloured in with a red colour permanent marker.

Time: 0:11







Ah! The kit is missing 4 magnets! The missing magnets must have been 3mm diameter and perhaps only 1 or 2mm thick.
I have some 4mm magnets but they are too big for the formers. The awl cannot be used to enlarge the hole.

I would have saved some time if I had the 3mm magnets.



I enlarged the hole in the ply former with a round file and glued my 4mm x 3mm magnet into it.
Of course the former broke!
Then I assembled the 2 fuselage sides against the tray and glue them together. It is better to do it this way otherwise the ends might be mis-matched.

Time: 0:36



After the sides were glued to the frame, I found I needed to colour more balsa.

I painted the dashboard black and the top decking red.






They were glued to the fuselage and attention is now turned to the motor mount.

The motor mount had too small a slot to fit my WLTOYS 8mm motor.

Time: 1:43





The motor mount was marked out to receive the WLTOYS geared motor.










The motor mount was slotted ('dremel' type rotary tool rocks!) and drilled to accept the pins of the WLTOYS geared motor.

Time: 2:09







The motor mount was glued to the fuselage and I have to pre-colour yet another piece, this time the bottom of the nose before gluing the nose doublers and this nose bottom piece.








I assembled the pivot mount and glued it to the top of the fuselage. I pre-coloured the tail skid and glued it to the tail.


Time: 2:21






Work halted for the day, time for dinner.










I wrapped the pieces around a glass bottle, sprayed their top surfaces with water, some wood were used to spread the load of two rubber bands to prevent crushing of the moist surfaces, and left to dry in the morning sun.

Time: 2:30





When the pieces were dried, they were flattened out because I used a small glass bottle.

Drying and flattening time not included in the time taken to build this model.







The wings were assembled and glued 'in the air'.
Daiso's $2 5gm superglue rocks! And of course the precision of the parts were vital!

Time: 3:30







I marked out where the sewn hinges were to be. But I made the mistake by sewing from the other un-marked side. The kit is missing the plastic hinges in Banggood's photos.


Time: 3:45

It seems a dreaded task to mark out the locations, but actually it was very fast to do. 1 cm spacing.









The sewn hinge looked neat enough. it would have been faster, easier and neater if I had started the threading right on those pencil marks!


Time: 4:00

This is where the stale superglue can be used.












The hatch had two washers and these were attracted to the magnets in the fuselage.



Time: 4:10

The orange tube is the Daiso's superglue! Good product! Because it sticks well and it is cheap!











Another mistake I committed was to open up the rear slot for the rudder pushrod because I later decided that the front slot was more appropriate.



Time: 4:30












The kit is missing thin wires for the pushrods! The kit had enough carbon rod and heat shrink sleeve but no wire to make the connection. I had to use my own wires: a short length of thin music wire for the rudder horn connection and some lengths of thin flexible wire for the other 3 connections.

This took a long time to figure out and do.


Time: 5:00









Then I discovered that the motor wire was too short to connect to the board.

Flipping the WLTOYS receiver board didn't help, so the connections were fitted again and the receiver board was then hot glued in place. The sleeves were shrunk with a heated piece of paper clip.

Thang will give me an extension. He said he will give me a spare 7mm too.

Superglue was used throughout. I bought a fresh 5gm tube of cyanoacrylate from Daiso for $2.00. Good product! My stale balsa superglue (bought from model shop) didn't stick too well.












I still have to:

1. Wire up and hot glue the motor
2. Fixing the landing gear
3. Fixing the top wing. The kit is missing the 4 screws in Banggood's photo.

List of missing items in the kit:

1. 2 wheels
2. plastic hinges for the rudder
3. 4 screws
4. thin wire for the pushrods.

A Vacuum Former that also function as a stackable storage box

20 July 2016

Recently, the desire to make clear or opaque plastic canopies, wheels cowlings and nose pieces has gotten stronger. I considered folding techniques, stretch-moulding and after thinking somemore, believes a proper vacuum former will give a better result and it doesn't necessary have to be complicated or wasteful while becoming an eyesore wasting space when it is not in use.

I would prefer to have paid a small price and get a well made one then to make one myself, but since it is not available, here's an idea for a minimal fuss vacuum former.

The main components are 1) a vacuum chamber, 2) a frame for the thermoplastic and 3) the heat source. If I should make one I know I won't use it much, because that is me, however the most part of the fun is had if I should. I just have to bear in mind not to clutter and keep things simple, cheap and easy, and I think that a converted plastic storage box is a good basis to make a vacuum former.

The vacuum chamber:

1. Check that the 20+ years old house vacuum cleaner still works, measure the vacuum tube's diameter. If mine doesn't, I am not going to buy a replacement and this project will be abandoned.
2. Buy a small plastic box to be the vacuum chamber. I am only interested in very small mouldings but I want the box to be sufficiently big to store  the paraphenalia. Maybe I can find something at Daiso. At $2 each, it won't break the bank. Choose one that has sufficient depth to accept the vacuum tube and sealing rings.
3. Cut a hole in the side of the box. It should just be big enough to slip in the vacuum tube.
4. Cut two rings that will fit over the vacuum hose out of some compressible epp-type foam of perhaps 10mm thick. Maybe I can find discarded blocks or sheets of foam rubber.
5. Hot-glue one ring to the inside of the box, over the hole made out to receive the vacuum tube. Tape the other ring to the vacuum tube. Tape around the vacuum tube if necessary to have a snug fit against the inner ring. Check and adjust so that the vacuum hose fits snugly to the vacuum box. The first ring will form the seal, the second ring that is taped around the vacuum tube positions the tube against the box and forms the secondary seal.
6. Drill small holes at the bottom of the chosen plastic box. If the object to form is very small, excess holes can be taped over. Place the box upside down, i.e., with the lid downwards and  the perforated bottom topmost, stick the vacuum tube in and we have a vacuum chamber. Once vacuum forming is done, remove the vacuum tube and the vacuum chamber reverts to a storage box when it is placed upright. The box can then be used to store small pieces of thermoplastic sheets and mouldings, the second ring, plywood frames and male plugs, tape, hair-dryer, heat gun, stapler, staples etc, anything that doesn't fall out of the holes in the side and bottom of the box. We have a vacuum chamber that reverts back to  a storage box and since it is one of those plastic box, it is stackable too, saving footprint.

The frame:

There are a few options for the frame. I prefer option 3:

1. An identical plastic box, either the bottom box or the box's lid with a centre cutout, but then I have to figure out how to secure the thermoform plastic (tape perhaps?) and it will cost another $2.00?
2. A simple rectangular timber frame made of wooden strips that fits loosely around the bottom of the plastic box. Use angle brackets but ensure sufficient width of timber is still exposed as this will be used to staple the thermoform plastic sheets. But the cost will go up because I have to buy everything including the brackets, nails etc.
3. A piece of ply wood with a center cutout. It is not necessary for the cutout to follow planform of mould, afterall, it is plunge moulding. After use, this plywood piece can be stored in the vacuum chamber box for future use. The thermoplastic only has to be as large as the frame, and since the frame can be made small, I won't need large pieces of thermoplastic. Yes, I have to buy or find a piece of plywood, but I think it will cost less. A slight tweak to this option is to use ice cream sticks stapled to form the frame.

Heat source and use:

Thermoform plastic sheet will be stapled to the frame and heated on the underside by a hair-dryer. Then the heated framed plastic will be brought down over the vacuum chamber and the hair-dryer moved to the topside. If my hair-dryer can't heat the thermoform plastic sufficiently, then the options would be to 1) use the gas hob, 2) buy a heat gun, 3) use the toaster oven. I don't suppose hot water will work well. Regardless of choice of heat source, bear in mind it depends on the size and type of thermoplastic I will be experimenting on.

Small craft foam wheels and other bits

15 July 2016

Too much work for little bitty wheels! Anything round and has a hub would qualify as wheels.

Getting the CG also means getting the wing and tail incidences and the thrust line correct. They all work together and influences one another. If I view the tail as a stabilising force, the tail should be flat to the airflow, little or no lift, and the wing propped up slightly at a positive angle for stable flight. But the more it is propped up, the more 'one-speed' the plane becomes. So I could also set both the wing and tail at zero degrees to each other and use a bit of 'up' elevator for stability. From the slowflyer example, the wing has lots of positive angle to the wing, relative to the tail. I had to 'down-trim' the elevator for that CG position I want and the slowflyer flies with the fuselage nose pointing down. In other words, I have simply built in more downthrust. I have elevator control in a RC model, I could simply have set both wing and tail at zero angle to each other and used a bit of 'up' trim, and let the plane figure out itself what angle to hang the wing, probably 0.5-1.5 degrees, very minute anyway. By doing so, we can then approximate the CG placement base solely on the distance and relative size of the tail. The real CG depends on how the plane glides during test flying.

Thrustline angles are dependant on where the thrustline is relative to the centre of gravity and centre of drag of the plane, how long the propeller is, relative to them, diameter, pitch, area, shape and drag of propeller, and its spinning speed, and the flying speed of the aircraft. The wing is the biggest drag element. Just put in say 3 degrees downthurst and 1.5degrees right for general 'sport' flying unless doing aerobatic types (eh, maybe 1.5degrees eachway?). Whatever it was set to, the real thrustline depends on how the plane goes under full power during test flying. Adjust the thrust line to get similar power on and power off glide behaviour, quite impossible, but close to is enough. After CG is determined, more down thrust if plane zooms up under power and more side thrust to counter the torque. Thrustline acts on 1) CG, 2) centre of drag, and 3) the vector force which could cause a turning moment on the previous two points (vertically and horizontally), and 4) the convolution of propeller and spinning speed and is relative to the range of flying speed. Nothing much to work out, just try it in the air and observe the plane's behaviour.

For directional and roll stability, it doesn't really matter because we can manage even if it is without aileron and has only rudder control, but make sure the CG is centre to the fuselage and the fuselage is straight to start with. To improve rudder controlled planes, some form of dihedral: 1) the wing's dihedral or sweepback or 2) lateral area or CG above the CG/centre of drag, is required. Otherwise, the model will skid but doesn't translate to bank and the turning radius is large, therefore the turning response is slow. Another basic thing to watch out for is that rudder controlled planes has another quirk, if the fin and rudder is high up relative to the CG/drag centre (and nothing below it), it rolls the plane instead of yawing the plane. The rudder deflection will act like aileron and this rolling motion is not in the turn direction desired. Just lower the fin and rudder, probably don't need so much area anyway. This is borne out of my experience with my BD 5 where I got the area too big and they are way above the fuselage. So, it got to look right as well.

9 June 2016

The craft foam rimmed spoked wheels are installed on the big slow flyer and has lasted one flying session (a few landings). They seems to be working fine so the construction method used is ok.

Those wheels are for show but they are not attractive because their rims do not spin centrally. Obviously the acrylic setting tool is not good enough for such diameters.

Maybe a wheel jig like this will ensure that the cones and hub comes together correctly. I could also it to glue the silicon tubing tyres.

For the tyres, I don't think I have to slit the silicon tubing. The craft foam tyres have been strong enough, maybe gluing the tubing directly to the rim of the cones will work too. After the cones and hubs are glued, I use the rim of the cones to mark its circumference on a piece of paper. The girth (length) of the tyre will be longer, work out the maths by adding in one diameter in the formula and cut a tubing to length. Join the ends of the tubing to form a ring by heatshrink sleeves and hotglue, make sure it is not buckled. Then spray black.

Re-insert the cones and hub spoke wheels to the wheel jig's axle wire, supported by the lower ring. Slip the tubing ring to the spoke wheel, set the tube-ring to height, put some talcum filler, and then some superglue with the tip of a fine point tool.



The silicon tubing has arrived but before I try it out, one of the previous cones made snap at the joint line, so I will rejoin it and clean up the cones before making the wheel jig.

Ha! Here's an idea for the wheel jig: 1) Mark the exact centre of two identical plastic screw caps, 2) drill the centres with drill bit the diameter of the hub tubing and we are done.

If plastic cones were pre-drilled, the two plastic cones are sandwiched between the two screw caps with the hub tubing running through the caps (hot glue, so it can be removed later). Spots of superglue will secure the cones at the rim. Glue the silicon tubing around the rim. Remove hotglue from one end of the cap and remove the cap. Dab some superglue to secure the exposed hub to one side of the wheel. Then remove the other cap by breaking the hotglue bond and apply some superglue to the otherside of the wheel's hub. Trim off and we have one spoked wheel.

If the plastic cones were not pre-drilled for the hub, then the method is: sandwich the two plastic cones as before, eyeball that the rim and the screw caps are concentric, spot glue the rim with superglue, drill the cones through the screw caps. Remove screw caps and glue the plastic hub tubing, trim. Glue the silicon tubing around the rim, eyeballing that it remains central. Obviously, with this procedure, you have to hold the caps against the cones gently and consistently.

27 May 2016

The template for the spoked wheels was printed out and squares of pvc plastic taped to the paper template in turn. A pin mark was made to a plastic squares before taping to the template. This marks the centre of the cone where all spoke lines radiate from. I scribed the 17 spoke lines on the plastic using the same plastic headed dressmaker pin against a straight edge, starting from the centre pin mark radiating outwards, sighting the printed spoke lines of the underlying paper template.
When the 4 plastic squares were all scribed, I ran my Pilot G-2 0.7mm black pen over the scribed lines. The plastic squares were left to dry when I went for lunch, then a soft tissue was held against each square and the ink buffed away in circular motion. This left the scribed lines with the black ink.

The four squares of plastic have their holes pierced through and enlarged with a rotary diamond bit (the long tapered one that looks like a spike with glittery surface). Pushing the tool while rotating with the fingers had the holes enlarged and the edges protruded which ought to be useful when gluing the bushing tubing for strength. The rotary mandrel (the one that has an enlarged head holding a bolt) was assembled this way: the plastic washer was fed to the bolt, then two square plastic pieces, the paper template (this time, the paper template has the diameter of the wheel marked on the 17 spoke lines), the final two square plastic pieces and finally the mandrel. The assembly was gripped in a dremel rotary tool, held secured against a table surface with the left hand, and spun at the lowest setting. The small dots marked on the paper template can be seen through the square pieces and a cutter was introduced with the sharp edge pointing outwards. The first two pieces were cut accurately but after that, I could cut the remaining two pieces. A lot of plastic bits and dust were flung around and it was messy and it hurts when a scrap stung my forearm. I decided I don't want to cut the remaining two pieces with the cutter. I saw that the last two pieces on the mandrel were bigger than the first two pieces and were not round. I used a 60 grid sandpaper block and sanded them down. It took about the time to write this paragraph.

The segments of the discs were cut out and clear transparent tape was applied to the unscribed surfaces on one edge of the discs. The tape should be slightly longer, so that the corners of the tape will go past the spoked discs. The other edge of the discs were brought together with the scribed lines inside of the formed cones going past the first edge that had the tape on the other surface. The edge was released until it matches the first and the tape brought in contact, thus holding the coiled discs in a cone position. A little superglue was applied with a small scrap of the segment and the cone placed on a flat surface for curing. The exposed tape applied in the previous step held the cones in position. The cones were weighted down to ensure that the cones do not cure twisted. It only took a few minutes for this operation.


These were the four cones after peeling away the tape pieces. The butt joints held the cones but they were slightly raised off a flat surface at the rim. It could have been better if I had weeighted the cones better, on the joint lines.

Another problem was the blooming of the superglue on the scribed inner side and some seepage on the smooth outersides. A cotton bud and some alcohol will clean up the cones.

Now I have to put them aside until I receive the 3mm x 2mm silicone tubing from Banggood to do the tyres. They are neater than the previous pair.

 

Here are ideas on the use of and the working of the 3mm outer diameter x 2mm inner diameter silicone tubing:

• Short rings may be used to reinforce the cones and provide greater bearing surface for the 2mm outer diameter plastic tubing.
• Short tubes may be used for the same purpose as the short rings, but instead of two rings per wheel, only one tube is required.
• I have yet to figure out a way to make a tool that can slice the tubing perpendicularly with a razor blade.
• Make a tool to slit the tubing. This comprises a 1mm packing piece to which is glued a #11 blade, the tip of the blade is exposed about 1.5mm from the packing piece. The packing piece with blade is then glued to a flexible strip of pvc. The tool is used by rolling the pvc strip around the tubing and inserting a second 1mm high packing piece, pinching with the left hand. One end of the tubing is tied stationary, the other end is fed through the tool and pinched with the right hand and held at a slight tension. The tool is moved forwards the stationary end and the tubing is slit.

 

25 May 2016 (attempts at a pair of 1.5 inches spoked wheels)

There is no spoke in this spoked wheel construction. Instead, it relies on drawn lines on clear plastic cones to simulate the spoked wheels. Why not? It is only a hobby.

These 1.5-inches wheels were intended for a Roger Sommer's Monoplane, approximately 20" wing span, 3-channel, powered and controlled by WLToys F929 F939 components.

I used a dremel mandrel piece to screw on a craft foam and spun it in my electric drill to cut the disc. Previous experiments was gluing a bushing hub before spinning and cutting; this experiment with a mandrel piece, the one with a wood screw, that is used for felt disc (for polishing) works.
In this case, it was for the crankcase of the 7 cylinder rotary engine, but I put it here because it was a more convenient tooling method.

I drew a circle of my wheel diameter and divided it to 16 segments.

 I traced the image on PVC cover sheet by scribing it with the back of a cutter then going over the scribed line with a black pen. The pen markings would be wiped off if there is no scribed lines. I cut with scissors, poked the centre hole, cut a single slit and coil the disc into a cone. The first half is glued to a bushing perpendicularly and then the other half introduced with the spoke lines staggered.

The result however showed a flaw in the concept. At one side of the spoked rim, the spoked lines are staggered but at the other side, they matches. The pattern I used did not consider the overlap.

I didn't use a compass so I used my protractor to draw a circle and marked out the segments. The overlapping portion is 8 degrees and the other 16 segments are at 22 degrees.

I used a compass cutter to cut the plastic disc and it was inaccurate; I have had better accuracy using a pair of scissors. Anyway, the disc was taped temporary to the template and sribed/penned. Then a single slit was cut and the disc overlapped by the two spoke lines of the 8 degrees segment. The centre holes were widened with a round file and the bushing added to one cone and then closed off with the other cone to form a spoked rim.

I have ordered 3mm (OD) x 2mm (ID) silicone tubing from Banggood after I misplaced my previous purchase but I wanted to try completing the wheels with 2mm craft foam as tyres.

I cut strips of the 2mm craft foam, and glued them around the spoked rims. I started at one end of the strip by applying a bit of superglue. The spoked rim was brought in contact to this spot and held until there was sufficient grip. More CA was applied to the spoked rim in small sections and the rim rolled onto the foam strip. When the foam strip made a round, the excess was cut off with a NT cutter and the free end glued in placed, butting against the fixed end.

 

 

I observed:

1. The pair of wheels are inaccurate and not the same sized because the compass cutter had to pass more than once around the plastic sheet.
2. The lines were not radiating from centre and the hole is probably not centred.
3. There is a twist to the cone because of the overlap.

To improve accuracy/neatness, I think:

1. Use better template. Scribe and draw spoke lines as before, against the better template.
2. Either hold 4 pieces of plastic together in a dremel mandrel (original use of mandrel is for cutting disc) or a bolt/washers/nut assembly, or scissors cut individually.
3. Cut the 8 degrees segment away, glue half of a short strip to the middle of one side of the disc (i.e., midway between hole and rim), bring the other side and butt each other. Glue the short strip to secure. Glue will flow by capillary action to the butt joint.
4. Try using the silicone tubing as tyres when it arrives.

So I get a friend to draw this template. The concentric circles are for visual guidance of round discs.

And although primarily for guiding the cylinders in rotary engines, he drew circles with 7 and 9 radiating lines. Who knows, maybe one day I need to simulate spoked wheels that are too small or thin to do them in two half cones and I have to use flat discs instead.

15 September 2015

The monowheel prepared the night before was sanded with the bushing in a portable electric screwdriver/drill. The edge was smoothened but it is still eccentric.

The bushing was inserted to the setting block and cut.





The remaining bushing was inserted to a square piece of craft foam.

The foam was held against the side of the tool so that the bushing may be perpendicular.





The bushing was held in the electric screwdriver.

The centre pivot point of the compass cutter was held in the bushing and the cutting blade brought to the spinning foam gradually.

The wheel was cut off with a NT cutter rotating against the bushing along the side of the tool.

The tools and equipment to make this small pair of wheels are: electric screwdriver, setting tool, NT cutter, compass cutter, and sandpaper block to sand the bushing. The materials are: Superglue and Kicker, 2mm craft foam, plastic tubing.



The pair of wheels is 15mm in diameter, the monowheel is approximately 8mm diameter.
The scale is accurate to 0.1gm, and the three wheels registered 0.0gm!

The craft foam and plastic tubing are easy to cut. They are also very light.The procedure of gluing the bushing before cutting the disc results in a better wheel with less eccentricity.
It is suitable for making small wheels because eccentricity becomes more obvious with small wheels.

14 September 2015

Tape the craft foam to the cutting mat.
Set the compass cutter.
Rotate the cutting mat, not the compass cutter.
Trim if edges were still joined.
A later experiement showed that having the tape at the bottom of the craft foam will help ensure a thorough cut.

Cut the short lengths of bushings from pastic tubing. They are easy to snap clean.


Find a hole which is big enough for the plastic tubing to slide through the arcylic setting up tool.

The acrylic setting up tool was made by Tan Kee Chiao at his office/workshop with his milling machine and pillar drill. It is a series of perpendicular holes on a half inch (or was it 10mm?)  thick of acrylic.




Use the pointed end of a bamboo skewer to position the wheel disc on the drilling tool. The hole will stop the bamboo. Then the disc is held against the setting up tool and the plastic tubing is inserted from the other side of the too.







Rotating the plastic tubing while applying force has the bamboo skewer replaced by the plastic tubing.

The plastic tubing is retracted but ensure that the craft foam disc is still in position.





The short bushing is inserted, then some UHU Por glue is applied to the exposed bushing.










A wire which fits the plastic tubing is inserted from the rear, through the bushing and left aside until the glue is set.

The other wheel is similarly prepared.







Short lengths of plastic tubing were glued and crimped to retain the wheel.

The wheels are not true. The bushing and the centres do not coincide and I suspect that the disc may not be at perpendicular to the bushing. They are however, good for current purpose.




Craft foam doesn't sand well and I did not have to use a drill and a sanding block.
If I shall require a better but equally small or smaller wheel, I would have both sides of the craft foam taped before cutting oversized with circle cutter.
Then slide the axle wire through the plastic tubing, feed it through the setting up acrylic tool, wire through the taped foam disc and then the plastic tubing, super-glue one side, remove tool, wire and snap the pastic tubing to 1" length.
Spin the wheel by chucking the tubing to a drill, sand true, snap the plastic tubing again and the remaining length of plastic tubing ought to be more than enough for a few more wheels. When the stub gets too short, then they can be used as spacers.