Although I had seen several
Ross Yoke type engines running they had all been enclosed designs and gave
no idea of the nearly straight line motions that the yoke gives to the
piston rods. This engine was built to show the ingenious mechanism
in operation.
Unfortunately at some
stage of its 'design' (involving a 10x scale model using thumb tacks, wood
and cardboard) the yoke got turned over and and the engine finished up
rather longer than necessary. In the true Ross yoke design the crankshaft
is on the cylinder side of the yoke suspension link and the yoke has its
flat face away from the cylinders.
A mild disappointment is that the mechanism still cannot be seen working. It is a fairly high compression alpha type engine and, in spite of the heavy flywheel, it insists on running at too high a speed. (and don't the twelve second-hand ball races make a splendid din --- !)
The one really successful feature of the engine is the piston sealing arrangement.
Several people who had used Rulon or similar plastic lip seals found that although their engines initially ran very successfully, they lost all compression after cooling down and refused to restart. The reason for this is that the plastic's high coefficient of expansion causes the lip to leave the cylinder wall when it cools and this combines with its cold stiffening to prevent it re-sealing, although for some reason, possibly an extremely thin lip, this has never troubled me with my Whippet engine.
Whigmaleerie's pistons are unconventional
in that they do not have skirts but are merely steel discs carrying Rulon
seals. The top, 'hot', piston also carries the hot cap and this needs
to be kept accurately true within the bore, so it is fixed to its rod which
is guided by external bushes, with connecting rods coupling it to the yoke.
However, the bottom,'cold', piston rod is connected directly to the yoke
and the piston has to rock for the fraction of a degree necessary to cope
with its non linearity.
The seals are made from 0.015"
Rulon sheet, using a method evolved by Rob McConagy for the seals used
in his Stirling aeroplane engines.
Rings, punched from the plastic,
are clamped onto a male former which is then inserted into a die block.
The whole assembly is heated to the softening temperature of Rulon (which
conveniently coincides with the melting point of pcb solder) and squeezed
together in a vice. The result is a near perfect cup washer seal
and by correctly co-ordinating the male taper on the former with the female
in the die, a sealing lip tapering down to no more than 0.002" can be easily
obtained.
This gave an excellent seal
which combined with the very low friction of the rest of the engine to
give a really lively performance, but now, probably due to the rocking
movement of the cold 'piston' breaking the seal when cold, I too had an
engine that wouldn't start from cold. In fact the only way to regain
enough compression to restart it was to fill the jacket with hot water!
The obvious solution to this
was to use some kind of spring to persuade the cold lip into contact with
the cylinder wall, and the best idea seemed to be an internal O-ring.
However, since commercially available rings were all far too stiff,
these had to be made from 'silicone bathtub caulk' which cures into an
extremely soft, yet strong, rubber and, once the correct dimensions were
determined, this proved to be a simple job.
A length of steel tubing with a reasonably
smooth bore is spun in the lathe while a quantity of silicone is squirted
into it.
Initially a finger is used to ensure that
the inside is coated evenly and then this is followed by a length of rod,
held in the tool post and set axially true to give the correct wall thickness
to the silicone sleeve. The lathe is left running for a few minutes
until the caulk begins to cure and the opportunity taken to clean every
trace of the muck from the lathe etc. If left, it will rust any iron or
steel parts as it cures.
When fully cured, after about 24 hours,
the silicone sleeve can be rolled inside out and out of the steel tube.
It can then be stretched lightly on to an aluminium rod and square sectioned
rings parted off with a craft knife blade, best fitted in a holder clamped
in the tool post to permit accurate spacing. After some experimenting
it was found that rings approx 0.1" square by 1.05" free diameter gave
the best compromise between cold sealing and additional friction when hot:
in fact it is impossible to measure any difference in performance with
or without the rings.
Although the engine won the 1997 Model Engineer Hot Air Engine Competition it is neither as powerful or as efficient as the much simpler Whippet.
One unexpected feature of this yoke drive was the phasing between the two pistons. I had blandly assumed that they would be approx' 90 deg's apart and it wasn't until recently that I actually measured the angles and found them to be as follows:-
