Having completed the “top-end” of the engine the
“bottom-end” was next on the list for attention.
Although I have had all the aluminium components vapour
blasted and their appearance suggests they are nice and clean and ready for
assembly the reality is that the cleaning process often adds to the decades of accumulated crap deposited in threads and oilways. I therefore clean all threads
by running a bottoming tap down them and oilways by brushing in cellulose
thinners and blasting with an airline.
The oil pump, which has been stripped, cleaned and checked
for any damage is popped in first. Grove Classics sell new
ones, if needed, and I’ve used these on both the AJcette and AJS V-Twin
projects. As I’ve mentioned in the past, I have 2 ovens for heat treatment and
the larger, lower temperature oven that I use for tempering is extremely useful
for heating aluminium castings for removing and dropping in bearings etc. This
also applies to the oil pump which is an interference fit.
If it is not held together, the oil pump quickly dismantles
itself! The shim, K-110, that goes on the back of the pump holds the gears for
the scavenge part of the pump in place and if it is removed or the pump is
positioned drive-side up the gears simply fall out.
The way that I insert the pump is as follows:
Two 3/16” steel rods with a 3/16” BSW thread on one end are screwed
into 2 opposite holes in the crankcase that support the oil pump.
The crankcase is then heated in the oven to 180 - 200 0C.
The oil pump, with 2 of its countersunk screws inserted in the holes that will
NOT be used with the 2x 3/16” rods, is then held face down (to prevent it
falling apart) and offered up (the right way round!) and engaged with the 2x
rods protruding from the facing-downwards heated crankcase half. With a bit of
jiggling, the oil pump can be inserted fully into the crankcase and the
crankcase can then be turned over and put on the bench and the threads of the 2
screws engaged,
the rods removed
(I use a tailstock chuck – because it’s convenient. The
alternative would be to thread both ends of the rods and use a couple of
locknuts) and the other 2 screws inserted. Needless to say, a good pair of
welding gloves is needed to avoid a visit to A&E.
I have fitted a new timing-side bearing.
These (and the drive-side roller bearing) are a metric size
– ID = 22mm, OD = 50mm and t = 17mm. These bearings were also fitted to vintage
Nortons (I know because I’ve replaced them in the past) and probably other
bikes and they are available from Vintage Bearings or Veteran Triumph. As can
be seen, they are also now available with an oil seal which is very convenient
for the Velo engine which requires pressurization of the bevel drive chamber.
However, this does not mean that the K-150 shim can be omitted because this would affect the positioning of the bevel gear.
I have found both the original and new bearing to be tight
fits on the crankshaft and to avoid applying a lot of force on assembling the
crankshaft into the crankcase it was much easier to fit the bearing onto the
crankshaft so that the force is applied directly onto the bearing inner using a
short length of aluminium tube and then drop the crankshaft + bearing into the
heated crankcase.
The next step was to set up the crankshaft end-float and the
bevel drive. There are a number of possible ways in which the crankshaft
end-float can be determined; my preference is by direct measurement of the
distance between the mainshaft boss that contacts the drive-side bearing and
the contact face of the timing-side crankcase half
and the corresponding distance between the drive-side
bearing and the mating surface
on the surface plate (it’s actually a granite block). I
concluded that the end-float would be 0.041” – not too good for reliable
engagement of bevel gears.
As I think I’ve mentioned before, shims are available at
remarkably good prices on eBay UK. I paid 9 GBP for this collection (10 of
each) in various mm thicknesses and with 22mm ID and 32mm OD.
I decided that I would set the end-float to zero which can
be done with 2x 0.019” plus 1x 0.003” shims. The question now arises as to how
to distribute these as this will affect the position of the bevel gear and my starting
point was to put all the shims on the drive-side (which positions the bevel
gear at its most outward position) and check the engagement.
There is a good synopsis of setting up the crankshaft bevel
drive (and many other aspects of Mk 1 cammies) in FT 238 by Denis Drake and
downloadable from the VOCs website. More recently, Richard Boldry has also written
a comprehensive article on bevel gear meshing in the latest (at the time of
writing) copy of Fishtail (FT 507) so I will only add a couple of points that I
have found useful here.
The bevel gear needs to be pushed right up against the
abutting shoulder on the crankshaft and rather than putting the K-34 gear on
the end I simply made up a small spacer to allow the left-hand threaded K-114
nut to be tightened.
In this way, the engagement can be more easily seen when the
lower bevel gear + housing is fitted; if the K-34 gear is fitted it is more
difficult to see the engagement.
The objective is to get smooth rotation of the crankshaft by
adjusting the positions of one or other (or both) of the gears, with both
gears fully engaged and without backlash. The crankshaft bevel gear can be moved inwards (in this
particular engine build) by redistributing the 22mm ID shims that are on the
drive-side or outwards by adding shims (K-111/X available from the VOC) behind
the bevel gear. The bottom bevel can be moved vertically upwards by making
gaskets using different thickness of gasket paper acting as shims or moved
downwards by machining material from the bronze bush (an option of last resort).
The picture below is a partial screen shot from a listing
which offers gasket paper in various sizes from 0.25mm to 2mm – perfect for
making and testing gasket shims.
It turned out that with 0.041” of shims on the drive-side of
the crankshaft, no shims behind the crankshaft bevel and a 1mm (0.039”) gasket
under the bottom bevel that I got full engagement and smooth rotation. To get
to this point, a few gaskets with different thicknesses were made and tested.
It is not so easy to photograph with the whole assembly so I have removed the
spacer and LH nut to get better access.
Prior to setting up the meshing, I found that the threads on
either or both of the K-46/2 housing for the bevel gear bush or the K-52 nuts
can be poor after 90+ years of assembly and disassembly. The external thread on
the housing can be cleaned up using a 24 TPI thread chaser and holding the
other end in the lathe chuck.
The K-33 bevel gear is pressed out (it is held in by the
K-27 locating collar) prior to machining and pressed back after machining and
cleaned up to avoid any swarf finding its way inside.
On this topic, it can happen that the locating collar is so
tight on the bevel gear shaft that the bronze bush (K-46/3) is pressed out
instead. A Sykes-Pickavant type bearing separator is then needed to extract the
bevel gear from the bush.
I have been refurbishing these parts for the other engines that
I am building and this happened on one of the bevel gears.
An internal thread chaser can be used to clean up the
threads on the K-52 bevel gear gland nuts ….alternatively and if the nuts are
in poor condition, just buy new ones from GroveClassics.
With the oil pump in place, the crankshaft end-float and
bevel gear meshing sorted out the crankcases were sealed with Honda liquid
gasket (it’s expensive but I’ve been using it for years without any problems)
and bolted together. There are 3x ¼” diameter permanent studs and 4x 3/8”
studs (or bolts) that are in place when the engine is in the frame. I use all
of these to make sure that the two crankcase halves are firmly clamped together
after the sealant has been applied and for the remainder of the engine build.
I measured the piston ring gap in the bore which came out at
0.030” (!!). This is nearly 2 ½ times the recommended limit of, typically, 4 thou per inch of bore which gives around 0.0125”. However, with a nominal diameter
of ~ 80mm the slipper piston that is fitted is not original Velocette; so,
where to find new rings?
This measurement could also be interpreted as 79mm+ 0.020” or
79.62mm (within the limits of accuracy on old used parts) and which
manufacturer fitted nominally 79mm pistons? Norton of course. And the form of
the piston is very similar to, if not the same as, this Norton International piston.
Manufacturers did not change much if they didn’t need to and
the piston rings from other Norton motorcycles – Model 18, 16H etc have the
same dimensions for the ring thickness and radial thickness and this set of
rings
which I found on a well-known auction site cost around 20
GBP and fitted perfectly. The ring thicknesses are even given on the front of
the packet. These are quality rings, made by Wellworthy, and were almost
certainly manufactured in their Lymington works (South coast of UK near
Southampton) prior to the close of the factory in 1981.
I measure the ring gap at 0.013” and these were fitted
without change.
The next step was to set up the Oldham couplings for the
vertical shaft. These need some vertical clearance to avoid putting an axial
load onto the shaft and which would be transmitted to the bevel gear meshing.
But how much clearance? Phil Irving in his book Tuning for Speed simply states
“a small amount of end-clearance”, Peter Miles gives “at least 12 thou” in FT
141 and “not more than 25 thou” in an article on Building OHC Velocette Motors
and Burgess gives “not less than 12 thou” (Thanks Gerry).
Different thickness couplings are available from both VOC
Spares and Grove Classics, however, by taking a measurement of the gap (there are
a couple of washers and shims under the feeler gauge in the picture below)
I could not find a combination of the available couplings to
get close to these figures; this is a consequence of having a Norton slipper
piston fitted that requires an aluminium spacer (which can be seen in the above
picture between the crankcase and cylinder base) and has changed the cylinder
base – to – cambox dimension.
The easiest way for me to resolve this was to make a new
lower Oldham coupling
from O1 tool steel, fitted (before heat treatment because
it’s much easier to fettle) to give a clearance of 0.015”
and then hardened and tempered.
As this blog is getting quite long, I’ll leave the rest of
the engine build to the next time.
