Home‎ > ‎Part Index‎ > ‎

Flywheel

A website dedicated to the construction of an accurate 1/2 scale replica of a 1937 Aero Douglas Motorcycle


The flywheel is unusual in as much as it houses the entire clutch mechanism.  The friction material used is cork, and the full size component uses keystone shaped pieces which are inserted into an aluminium clutch plate.  It rotates at engine speed next to your left foot and once disengaged cuts drive from the crankshaft to the primary chain leading to the gearbox.  It is held onto the taper on the crankshaft by a captive nut and lock ring making it self-ejecting - except it self-ejects on its own, whilst out on a run.  This is a known quirk with this flywheel - it has even happened to me once! 


Inner Hub & Friction Plate

The internal components were designed around commercially available needle roller bearings.  The full size clutch has hardened grooves machined into the body of the clutchplate to receive loads of 1/4" x 1/4" rollers, there are two of these grooves and a these make up the bearings.  I did not want to replicate this due to the complexity of producing the rollers and hardened grooves, and the fact that it's a bleeder trying to slide the full size one over its spigot without the rollers coming out!  Two commercially bought bearings were used, a needle roller and needle thrust roller with hardened thrust washers.  As on the other components, the full size one was measured and a few drawings made.  The half size counterpart has slight dimensional differences due to the required space of the needle roller bearings but it is so slight you wouldn't notice even with the full size one sat next to it. 

The first component to be made was the inner hub.  This is connected to the sprocket and the friction plate, and slides on the centre spigot inside the flywheel releasing the clutch.  It must be said here that the clutch springs force the friction plate onto the inside of the flywheel towards the engine side.  The friction plate is NOT clamped as per usual but totally relies on the spring pressure for grip. 

The inner hub can be seen here in the Myford ML7.  As much machining has been done as possible in a single operation meaning that so far, it hasn't been taken out of the chuck.  This ensures that all the diameters are totally concentric - something that is very very important in all of the flywheel components.  The outside shapes are quite complex, internal and external radii, lips and undercuts all replicate the full size.  The hardest part was trying to imitate a separate pressed oil-slinger plate, which is held between the friction plate and the inner hub, in the solid. 
Apologies for the poor photographs - my camera isn't upto taking close up pictures. 

The bore was turned a specific diameter as the needle roller bearing is to be pressed in.  it presses in upto a step in the bore to accurately locate it.  The sprocket is then pressed and dowelled on the end, providing a cap and therefore totally enclosing the needle roller.   The closest sprocket i could find for scale and number of teeth turned out to be a 6mm pitch sprocket.  I purchased one from HPC gears rather than making one, this is because im going to purchase a chain from them so in theory the two should mate perfectly.  Both the end of the bore which the bearing sits in and the bore through the sprocket cap are only very slightly larger than the shaft that it runs on in an attempt to prevent dirt from getting in.  The rear of the inner hub contains a counterbore housing a needle thrust bearing and a lip to position the friction plate.  12 3/32" holes were drilled for the rivets used to secure the friction plate using a vertical dividing chuck.

The spigot that the bearing runs on is a separate component from the flywheel body and inner hub and is described further down. 



The friction plate is made from cast aluminium with the keystone shaped holes cast-in.  I decided not to follow due to the difficulty of producing the shaped cork friction pads accurately so that they were sure to stay in position - the last thing i want is a loose pad floating around in the flywheel.  A piece of aluminium plate, 3mm think, was found in the junk draw.  it has been clear anodised making it look like it could have a slightly rough/cast surface - perfect.  It was roughly hack-sawed to shape then held in the 4-jaw to have the bore machined.  This was a push fit on the step left on the rear of the inner hub.  It was then taken to the 3-jaw and its outside diameter gingerly turned.  It was left slightly over diameter to allow for a skim once riveted to the inner hub.  The holes were spaced using a rotary table on the milling machine and deburred.  The cork discs will be pressed in later.

The two photographs easily describe what i cannot.  The imitation separate oil-slinger plate can be seen as can the rivets, sproket-cap and counterbore for the needle thrust roller in the rear.  The rivets are hammered into countersinks in the friction plate and look amazingly like the full size ones.  
The needle roller bearing was then pressed in using the arbor press with some suitable protection then the sprocket, which is counterbored on the inside, was pressed on over the stepped-end with some loctite just for good measure.  All in all it turned out very well, with the outer diameter not requiring a final skim as it only ran out by half a thou - the extra few thou left on the diameter still provided enough clearance inside the flywheel body so it was left. 




Cork Inserts

The cork was given to me and was a cork floor tile slightly under 5mm thick.  A rotary punch was made using an ordinary bit of steel 1/2" diameter, drilling a hole up it 3/8" then turning a taper to a knife edge on the outside.  This was then held in the drill press on a slow speed and used against a bit of wood cut out the cork in neat little discs.  The holes in the friction plate are 8mm so there is a fair amount of squeeze involved. 

After hearing many stories from my fellow club members of them re-corking their old nortons and triumphs, and after reading about it in a vintage motorcar maintenance book i took the plunge and soaked all my cork pads in water - this also helped to clean then and remove all the loose bits after being cut.  An insertion tool
was made, this is no more than a piece of steel with a tapered hole  through it.  The inner hub and friction plate were clamped down to the drilling machine table as in the photograph, it's amazing how large a 3/8" bolt look when compared to the plate.  A depth stop was provided underneath giving a positive and equal depth to the cork inserts hopefully meaning that there should only be very little to grind off the friction face.  Two drill bits were used to provide support between the friction plate and the depth stop.  Each time the jig was lined up, a piece of cork put in 'wet' and pressed down home.  The tapered hole is smaller at the extreme end than the hole it lines up over so that the cork can be easily pressed in then expand to fit.  This took quite a long time to achieve but the results were good although due to the nature of cork some protruded further than others through the back of the plate. 


The results can be seen here, there is a fair amount of cork to be ground away to level all the pads, this will be done with a dremel in the lathe later on.  The cork needed time to completely dry out and so this part was left for over a week on top of a shelf with a cloth over it to keep out the dust, being aluminium the wet cork inserts had no corrosive effects. 



Sliding Sleve

Whilst the friction plate was drying out it was time to tackle the sliding sleeve. A tougher steel was used for this (En8) because it becomes the running surface for the needle roller bearing in the inner hub meaning that allot of time and care was needed to achieve the correct finish and size.  The sliding sleeve is a top-hat shaped sleeve and does what it's name suggests, it carries the inner hub and friction plate, which rotate
on it, and slides up and down the

inner spigot inside the flywheel body.  So  the sliding sleeve is always rotating with the flywheel and therefore engine, springs press against the outer end of the sleeve and this force is transferred through the sleeve flange, through the inner hub via the needle thrust bearings (as seen in the photograph), through the friction plate providing grip to the driving plate of the flywheel and as the sprocket is attached to the inner hub it becomes driven by the engine.  To engage the clutch, a cam operated lifter presses against the sprocket end of the inner hub forcing the hub against the springs and the friction plate from the driving plate.  Its complicated to explain but simple in operation. 

All the important diameters were rough turned first to within 15+ thou so that the outer diameter or bore will not effect the other when final turned.  It was then parted off, rotated and faced.  A radius was added so that the sleeve can slide fully up against the inside of the flywheel body.  The clutch springs press against the flange on the sleeve and pass through the flywheel body, held by a separate ring which is held onto the flywheel nut. 






Flywheel Body & Driving Plate

The flywheel body was made from a lump of steel 6" in diameter and 2.5" long, hacksawed from a billet on a rapidor manchester hacksawing machine - it took just over 5 hours to cut through it!  This was then held in a colchester lathe and the rear rough-turned to size and a pilot hole made through the centre.  The idea is that by rough-turning the steel then removing it from the chuck, it can flex and twist thereby removing any stresses within so that when replaced in the lathe and the final 10 or so thou removed, it remains the shape you want.  Any wobble on a flywheel is clearly noticeable and looks rubbish meaning it must be avoided at all costs.

The cut billet was turned in the colchester lathe.  This could've been machined in the myford but it would have taken allot longer due to the much smaller cuts taken.  The billet was held in the outside jaws of the 3-jaw with the rear turned first including the outside diameter.  The bore was machined at this stage.  It is threaded to accept a castellated ring for the self-ejecting nut but stepped to give a lip to press the insert containing the taper matching the crankshaft up to.  As the crankshaft has not been fully assembled yet, the taper in the flywheel will be pressed in later.  The stepped bore was bored to the depth of the thread (26tpi)  so could be used as a gauge when screwcutting.  A test screwcut was already made and to hand to use as a guide, when this went in it was right!  This gauge was later made into the self-eject ring, so nothing here gets wasted!  The centre spigot is 30mm diameter and has a grease groove around it as the sliding sleeve slides on this part. 

It was then turned around and held in the 4-jaw with soft packing.  It was clocked up, which took allot of time as it had to run dead true and the face turned with a recess in the centre.  The rim is just over 1/2" thick but the face is less than 1/8" thick so to prevent it vibrating, plasitcine was used to full the back in. 

All the holes were drilled on the milling machine on a vertical dividing chuck by holding the central spigot.  packing was used around the hole to be drilled in the rim to prevent it twisting whilst under pressure.  The spring holes are 9/32" to accept 1/4" stiff springs and the 4 tapped holes are 8BA to hold the brass lock ring.  The outer 6 holes are counterbored from the front and take 4BA home made screws of the same pattern as the full size ones which hold in place the driving plate on the inner side of the flywheel.  The slit in the screws were made by first fitting a 1/32" slitting saw up in the milling machine spindle and getting the height correct by bringing it up to a centre held in the dividing head, and with a powerful magnifying glass setting it so that the point is in the centre.  The slitting saw should be rotated and any 'wobble' taken into consideration.  The screws were then held in the chuk and several passes were taken.  The manufacture of the driven plate is not described here because it is a metal disc with 6 tapped holes around it.




Everything Else

The steel spring retaining ring is counterbored to accept the clutch springs.  It is approx 3/16" thick and has a notch milled into its outer circumference.  This plate holds the springs against the flanged sliding sleeve providing the required force for the friction plate.  This sleeve is prevented from rotating by a brass flanged ring with a matching 'notch' pressed into it.  The flanged ring was turned from the solid with the notch and holes produced before it was finally parted off.  The indentation was produced by rounding the end of a 3/16" bar and polishing it to remove any marks, placing the steel retaining ring into the bore and then in a vice, squeezing the rod into the notch the right amount.  I had to make two flanged rings because i squeezed too hard the first time and fractured the brass.  If i was going to make another one i would use steel as it doesn't fracture as easily when pressed and it's the correct material - but at the time i had none. 

The flywheel is held onto the taper of the crankshaft by a Nut that serves three purposes.  The main body of the nut houses a thread which matches that on the crankshaft and when tightened onto the crankshaft pulls the tapers together.  The nut is s tightened using the hexagon form which is smaller than the main body of the nut providing a flat ring.  A castellated tubular lock ring is brought down on top of this section to both lock the nut in position and to provide the self-release mechanism when needed.  The thread on the outer end of the nut is used by the brass threaded pressure ring which holds the steel spring retaining ring in position and is done up tight on the end.  Finally there is a working half scale grease nipple in the end which forces grease down a groove through the thread and into the needle roller bearings in the inner hub. 



 


Cork Grinding

It is ideal that the maximum amount of cork is in contact with the driver plate and to do this the corks must be cut/ground once they have been forced home into the friction plate.  A Dremel was used with a rotary sanding drum held in the tool post of the lathe.  If you unscrew the nose piece of the dremel you will find a thread, which can be made through a plate of aluminium or whatever you have to hand meaning that you now have a toolpost grinder which is suitable for this type of operation.  I doubt it would be suitable for very accurate work due to the lack of rigidity or decent bearings in the dremel.   The bearings were taped over to prevent muck getting in on both sides and the lathe bed was protected.  No matter what people say any grit will slowly grind your lathe bed away. 
Here you can see the rear of the cork pads being 'tidied' up a little bit.  The fronts of the pads were skimmed down in 5-10 thou steps with the drum leading on its front edge slightly to provide clearance.  This method worked very well and provided a smooth flat area on the corks with no tearing or ripping of the surface.  It's amazing what a floor tile can become! 





Components and Assembly