Thursday, 23 July 2020

The Cambox

On the Velocette engine, the camshaft is driven by a bevel gear and lubrication of the bearings, cam and tappets is by oil that enters the cam chamber from the bevel drive housing by being pressure fed through a slot in the camshaft that aligns once per revolution with a slot in the phosphor-bronze bush that supports the shaft.

The arrangement here is entirely different; the camshaft is chain driven and there is no pressurisation inside the timing case to force oil into the cambox. An external oil supply must therefore be provided to lubricate the contact surfaces.

The design of the Velocette engine bevel drive provides plenty of opportunity to fit a good size bearing at the driven end of the camshaft. The loads are wholly radial and a large deep groove ball bearing was therefore fitted into a new bearing housing. A new shaft was made to support a new K17/2 cam, shown below.

 And the camshaft was completed by making a Vernier adjuster so that the valve timing could be easily set.

Finally, the cambox was drilled and tapped to accept a new oil feed that would be pressure fed directly from the pump. The picture below shows the new oil feed and also a restrictor that was plumbed into the system to control the amount of oil reaching the cambox. 

Oil that “escapes” from the cam chamber into the surrounding gutters is collected, shown below, and drained back into the engine, which is standard Velocette practice. Also, as per Velocette practice, oil that reaches the gutter on the inlet side is fed into the inlet valve guide. The 1/8" OD copper pipe for this oil feed can just be seen on the right of the picture.

Unfortunately, the cambox lubrication proved problematical during the shakedown of the bike and more will be said of this later.

Saturday, 18 July 2020

Machining the Timing Cases

The first machining operation is to face the inner surface of the inner timing case, ie the mating surface to the crankcase. This provides a datum for subsequent boring of the holes for the securing screws and machining of the outer face and internal surfaces.
Internally, the surface on the tension side of the chain must be machined to accept a flat strip of low friction material to avoid the chain hitting the aluminium surface and the slot for supporting the chain tension blade support (approximately midway between the chain centres) must be added.

The picture below shows the inner timing case on the milling machine.

After machining the internal surfaces of the inner timing case and the inner and outer timing case screw holes and threads, the timing case can be assembled on the engine. 

The original material that was riveted to the tension side of the timing case appears to be some kind of hard fibre. I have rebuilt a number of original 350 and 500 engines which have been fitted with exactly the same material and I find it amazing that this fibrous material has survived, in good condition, for 90+ years. Unfortunately, I don’t know what this material is or even if it is obtainable today. I therefore chose to use a modern low-friction and tough nylon-based material called Ertalon 4.6  which has desirable properties for this application. However I could only buy this in round bars and it was therefore necessary to make a ~1” wide flat strip from a round bar.

This was then riveted to the timing case. At the same time, the chain tensioner support and retaining spring were fitted.

Note that a small hole has been bored into the flat surface on the outer timing case, shown below, to locate the oil pump drive. This has been made by drilling from the timing-side crankcase and through a dummy camshaft drive spindle to ensure concentricity with the oil pump drive spindle.

Finally, the housing for the Velocette oil pump is machined and 2 aluminium covers for the magneto and camshaft sprockets are made. 

Friday, 10 July 2020

Making the Timing Case Castings: The Patterns – Part 2

Having the ability to make multiple copies of the original timing case castings in resin provides a straightforward means of making patterns for new castings that embody the features of the original engine. Nevertheless, there are differences that must be incorporated into the new patterns, in particular, the centre-to-centre distance of the overhead cam chain drive is about one inch longer and the mating surface to the crankcase is a quite different shape.

The process adopted here for making the patterns is:

     1)    By making a number of resin copies, these can be cut up and “stuck” to other pieces to develop the overall shape and dimensions. JB Weld is used for attaching pieces of resin together.

      2)    The chain drive is left is place with its dummy camshaft spindle to ensure that the inner timing case is in the correct place. The outer timing case is made to fit the inner when the latter has been completed.

      3)    The mating timing case-to-crankcase surface is made by merging the thin slice of aluminium casting taken from the original SB6 engine crankcase to the resin copy of the inner timing case.

      4)    A strip of metal is used along the long vertical side of the resin pattern to ensure it stays straight and does not warp.

      5)    Small pieces of wood are used to incorporate, for example, bosses for screws.  

      6)    P38 body filler is used when the main body of the pattern is complete to blend in curved surfaces and to fill pinholes and sections of the resin that did not cast so well.
      7)    When the resin + aluminium + JB Weld + wood + body filler is complete, the mating faces (ie, those to be machined) of both patterns is coated in epoxy adhesive and laid on a large piece of thick gasket paper. When the epoxy has set, the gasket paper is cut away with a scalpel. This adds an additional  ~1mm of thickness to the face and is repeated a few times to build up the height of the surface to ensure that there is sufficient material on the casting for machining.

The series of pictures below shows the patterns in stages of completion

And finally, the patterns are taken to the foundry and cast in LM25 aluminium. The castings are then heat treated by the foundry to the TF condition (maximum heat treatment) and this effectively doubles their strength.

On the topic of casting, it is worth a few words about shrinkage. If you google "aluminium sand casting shrinkage" then you will get the response "aluminium shrinks by over 6% during solidification". I do not know where this information comes from or how it is defined. Even if a value of 6% is taken as the volumetric shrinkage, this would correspond to a linear shrinkage of over 1.8%, which is too high.

I have seen practical linear shrinkage values quoted at around 5/32" per foot which equates to 1.3%; this agrees with my experience. Comparing the dimension of the long outer timing case between casting and pattern, shown below, gave measurements of 433.5mm and 439mm, which corresponds to a linear shrinkage of 1.3%. The smaller dimension across the timing case gave a shrinkage of 1.4%.

Monday, 6 July 2020

Making the Timing Case Castings: The Patterns – Part 1

The camshaft and magneto are both chain-driven and the entire drive system is enclosed in L-shaped inner and outer aluminium chaincases. The inner chaincase is bolted to the timing-side crankcase and provides part of the structure to support the Weller chain tensioner and the outer is screwed to the inner to retain the oil and to provide the surface onto which the external oil pump is affixed. These castings are not particularly complicated; nevertheless there are many small bosses for screws and there needs to be sufficient material in certain places to allow for machining.

So, how to make the patterns? I digress briefly ….. Back in 1975 I lived in a flat in Muswell Hill, London, while I was studying and, as luck would have it, I met a guy that owned an antique shop. I hasten to point out that the antique shop was not in Muswell Hill, for reasons that will become apparent, but rather a less salubrious part of North London. Chatting one day, he showed me a bronze statue of Venus de Milo that he had for sale. This famous ancient Greek statue must have been reproduced thousands of times over the centuries, nevertheless a quality, signed bronze will make a good price.

For some time, I had been playing around making resin castings of various objects – not easy working on the floor of a small London flat, and was having reasonable success. I borrowed the original signed bronze from my buddy and made a 2-part split mould using RTV silicone. The 2 parts of the mould were held together by a multitude of screws that were cast into the silicone and, although it was quite tedious assembling and dissembling the mould for casting resin copies, it worked and worked well.

I then perfected the art of making fake bronze statues. Bronze powder filler was added to the resin and the mould was then part filled and the resin “swilled around” to give the surface a good thick coating. When that had set, more resin was added, this time with a lead powder filler to give the “bronze” added weight and the base of the bronze was then filled with a final dose of bronze powder resin. The final cosmetic touch, after gently eliminating and then polishing any imperfections at the mating line, was to artificially age the bronze with a home-brewed chemical concoction and stick a piece of green felt on the base. 

The result is pretty convincing and quite a few ended up on the shelf of his antique shop! The picture below is the last one that I have.

So, how does that relate to making aluminium castings for a motorcycle? At the time of starting the AJcette project I was restoring the AJS K7 and the various bits and pieces of the K7 engine were scattered around the workshop. What better way to make patterns for the AJcette timing cases than to copy the K7 castings in resin and “cut and shut” them to fit the new engine?

The first step in making the mould is to make a plywood box that surrounds the original part and to insert screws a couple of turns into the wood.

The original part, with various openings blocked off with cardboard and tape, is then suspended with stainless steel locking wire approximately ½” above the base.

RTV silicone is then poured into the mould to the top of the casing.

Finally, 3 pieces of round aluminium bar are lightly glued to the casing that will form sprues for feeding resin to the cavity and vent holes for letting air out. It is also necessary to coat the already-cast silicone on the bottom part of the mould with a release agent (I use Vaseline) because the new liquid silicone will otherwise adhere to the surface and make it impossible to separate the 2 halves of the mould.

After the second pouring of silicone, the upper and lower parts of the silicone mould can be separated by unscrewing the multitude of screws (which have threads cast into the silicone), disassembling the box and removing the original casting. 

The box and 2-part silicone mould can now be reassembled and used for casting resin copies of the original. Obviously this process must be repeated for both the inner and outer timing case components.