This engine – in fact the complete bike – came to me in a
pile of bits many years ago. It came with a Mk IV (or V) bronze cylinder head,
which is not the correct head for this bike and I have already replaced it with
an iron head – see this
blog post (the bit about the head is some way down the page) and fitted new
valves/guides/springs etc.
Quite a bit of preparatory work has already been done: the
cambox has been repaired and rebuilt (see here
and here)
and fitted with one of the new scavenge oil pumps that I made; there is a new
piston and rings and a Nikasil plated bore (more later) and a new crankshaft
and shock absorber assembly.
And so, this
is the collection of parts that needs assembling.
Many years ago, I took the crankshaft to Alpha Bearings for
refurbishing and Max Nightingale offered to make a new crank rather than
rebuild the one that I had; this was the result.
A beautiful piece of engineering – in fact he made me 2; the
3rd cammy Velo that I’m working on (the “Cammy Special”) has an
identical crank.
The first task was to check the fit of the lower bevel gear
on the crankshaft,
which was perfect, before checking the crankshaft end-float
(which was also spot on – the crankshaft had been made for these specific
crankcases) before heating the timing-side crankcase to pop in the oil pump and
check the lower bevel gear meshing.
The meshing was not good! There was a substantial mismatch
of the tooth engagement. In this situation it is not possible to make any kind
of adjustment to the K-33 bevel gear or its housing or bush to get better
engagement and the only solution that I can think of is to move the K-32 bevel
gear inwards. There are only 2 ways to do this – either machine the crankshaft
(not an attractive option!) or remove material from the back face of the K-32
gear.
The bevel gears are through-hardened and really require
grinding. I do not have easily accessible grinding in my workshop (although my
Tom Senior machine does convert to a horizontal mill and it would be possible
to use it as a surface grinder – but this would result in grit going
everywhere) but I have found in the past that a solid carbide cutter will
machine hardened gears.
And so, I set up the gear carefully in the rotary indexer on
the milling machine table with the keyway aligned with the X-axis and first
removed 0.050” from the face with a carbide end-mill and then deepened the
keyway by a similar amount.
This machining operation turned out pretty well
and the engagement, although not perfect, is much improved.
I could probably have removed another 0.015” to get perfect
engagement but the cutting edge of the end-mill had given up by this time –
it’s now away for resharpening.
The next step was to check engagement of the K-72 and K-34
gears.
This was the best that I could get - not very good.
The only way to get better engagement was to shorten the
K-34 gear by making a mandrel in the lathe
to hold the gear
to be able to remove 0.090” from the length and to reform
the taper (I estimated to be 150)
before facing off the end.
Using a single 1.5mm thick disk between the K-96 oil pump
driving piece and the K-72 magneto drive gear (see here)
there is now good engagement and an acceptably small amount of end-float of the drive gear
when the cover is screwed down.
With the crankcases now sealed and the various gear meshings
sorted out it was time to fit the piston and barrel.
I reported some years ago (here) that I
had 4 sets of pistons/rings made in the US that were copies of a Mk 1 KTT
piston – in fact the actual piston that came with KTT 55.
This is stamped K27-7 which identifies it in the parts list
as an 8.5 to 1 compression ratio piston.
When these were made, I asked them to reduce the pin
diameter to 0.75” which is quite adequate for an engine of this size.
The ring pack is manufactured by Total Seal
and the rings are somewhat more complicated than the
traditional 1-piece rings of the period. In the above picture, the 2 rings on
the left are the 2nd compression ring (one fits inside the other) and the 3 on
the right are the oil control ring. The top ring is a single item.
There are very precise fitting instructions regarding ring gaps
and the positioning of the gaps.
At the same time that the pistons and rings were acquired,
the bores on the engine(s) for which these were to be used were Nikasil plated
– the Wikipedia entry gives
further details. There is quite a bit of technology involved here: the rings
are specific for a Nikasil plated bore, the black coating on the piston is for
reduced friction and the rings were made and delivered to the piston
manufacturer to enable them to complete the detailed design and subsequent
manufacture of the piston.
Is it all worth it? My experience so far with this
combination of technology is “yes”. The AJcette was built with exactly
the same bits and pieces and this proved very successful – it has covered quite
a few hundred miles and performed faultlessly as my run-around bike when I took
it to the Isle of Man for the Manx GP in 2019.
The ring gaps satisfied the makers recommendations and were
assembled and the piston fitted into the cylinder without a problem.
With the engine on its side on the bench so that the spark
plug hole is vertical and with the piston at TDC the combustion chamber volume
was measured using an oil-filled burette (and using R40 oil which the engine
will eventually use for running).
The volume came out at 61cc which gives a compression ratio
of 6.7. Why not 8.5 as stated in the Velo parts book? I have no idea. The
position of the top ring in the bore is exactly where I would want it (and
expect it) in relation to the top …however a very small change in TDC piston
position makes a significant change to compression ratio. Combustion chamber
volume? Quite possibly; this is not the original cylinder head – the head that
came with the bike was a Mk IV or V bronze head (which is incorrect for a Mk 1
KTT) and was replaced with the iron head that I’ve fitted. Are there differences
in the combustion chamber volumes of iron heads? I have limited experience but
there is certainly a difference in the volume between the KTP head and a
non-KTP head as I reported here on
the V-Twin project.
Does it matter? Not in the slightest, I don’t plan on racing
the bike in competitive events and 6.7 is quite OK.
More importantly, is there adequate dynamic clearance
between the valves and the piston around the valve overlap period? The correct
way to check this (at least, statically) is to remove the valve springs and
then measure the valve-to-piston clearance at various crank-angles around TDC
and check against the valve lift curves. I have not done this but a few carefully
chosen measurements coupled with the fact that the compression ratio is 6.7
indicate that there will not be a problem.
The next stage in the engine build was to check and set up
the clearance for the vertical shaft for the camshaft drive. In the parts book
there are 2 different sizes of Oldham couplings that are available to set up
the clearance (K-35 and K-35/2). GroveClassics went a step further and had some additional flange thickness
couplings manufactured (although some of these are out of stock at the time of
writing).
When I built the engine of KTT 305 a couple of months ago I
ended up making an Oldham coupling to get the correct clearance (recommended
0.012” to 0.025” – which, incidentally, also feels about right). This was
pretty straightforward to make and so I decided to do the same thing here.
So how much clearance did I have on KTT 55? Very little (none!)
as it turned out - I couldn't even squeeze in the smaller thickness flange couplings and if I had made Oldham couplings to satisfy the target clearance the flanges would have been
dangerously thin. And so, the first step was to remove material (0.020”) from
both ends of the vertical shaft. It is hard steel but machines easily with a
new sharp carbide tip tool in the lathe.
That’s the easy bit. The coupling will not seat properly on
the shaft unless the slot is chamfered and this is a bit more involved. This
could probably be done with a Dremel or, very carefully, with a flapper disk
but I chose to make the chamfer using the milling machine and the setup looked
like this.
The rotary indexer is mounted on the angle plate and the
slot positioned under the cutter as shown.
The picture below shows the “new” end with the added chamfer
(on the right) compared to an original on the left.
I now had clearance and could make an Oldham coupling to
fit.
As before, this was made from O1 tool steel in the rotary
indexer on the milling machine
followed by cleanup and heat treatment
to give a clearance of 0.015” as best as I can determine and
which seats correctly on the shaft.
Incidentally, I have found that a reasonably reliable way to
measure the clearance is to use a set of small drills as pin gauges. I do not
possess a set of pin gauges (nice to have but a substantial outlay for
something that would probably only get used a few times in my lifetime).
The drills are in 0.1mm (0.004”) increments
and by determining which fit within the slot at both
extremes of the shaft position
I can make an estimate of the clearance. This is somewhat
crude and some judgement needs to be made about the “level of fit” but I think
I can probably get an overall accuracy level of around +/- 0.002”. A feeler
gauge between the end faces is a double check.
The vertical shaft cover was sealed (hopefully…) using a
PTFE seal and a nitrile O-ring.
Does this work? I’ll let you know….this setup was engineered
by a buddy of mine that had a couple of cammy Velos and who also worked at
Ricardo many years ago. He swears it works …if it doesn’t then I can at least
go back to the old asbestos (substitute) string method without having to strip
the whole engine.
The valve timing was set up and checked, as described previously for KTT 305, which gave the following results:
IVO 44 0BTDC
IVC 74 0ABDC
EVO 64 0BBDC
EVC 60 0ATDC
The more reliable of these measurements are those of EVO and
IVC because these are made without loading from the other cam during the valve
overlap period. If they are compared with the text book data in the table given
in a previous blog post here
then they are very close to those of the K-17/4 cam used in early KTTs
At this stage, the engine was installed in the frame.
The last detail to set up on the engine was the magneto.
This is a square ML with slack-wire advance (my preference) and has been
completely rebuilt mechanically and electrically.
I had to make new 3/8” studs – BSW one end and BSCY on the
other end, together with the cable holder that screws into the support arm.
Rather than simply bolting on the magneto, the rear timing
case and then putting on the sprockets and chain it is worth spending a bit of
time on preparation. First, do the timing case inner and outer mate properly?
I have some on other engines that don’t mate so well and it
is much easier to sort this out before assembly rather than discovering a poor
fit after the sprockets and chain have been assembled.
Secondly. chain tension is critical and it is much easier to
check and correct this before final assembly
and, as with all chains, the tension needs to be checked in
a number of different positions of the shafts to make sure there are no tight
spots.
I have also drilled and tapped the engine sprocket with 2x ¼” BSCY
holes 1 3/16” apart so that the sprocket can be removed easily with a simple
puller. With the engine sprocket removed I can remove the chain at the magneto
end and I have a puller that will get behind the sprocket. These are the 2
pullers that I use.
(this particular engine has an endless chain – most chains
now in use have a spring link which makes it easier to get a puller behind the
sprocket).
With the preparatory work done the inner timing case,
sprockets and chain were fitted and the engine timed at 380 BTDC
fully advanced using a sliver of cigarette paper between the points.
The fully-retarded timing was checked (the retarded timing
is not so important …. more for curiosity – most magnetos give around 400
of spark retard) and this came out at 30 ATDC.
Although there are a few details remaining that’s about it
for KTT 55s engine build.
