Filling in the Holes

There are a number of different ways in which the now-superfluous holes in the timing-side crankcase, primarily the 2 large holes previously used for the camshaft bushes, could be filled. As this part of the structure needs to be strong, a method that retains structural rigidity whilst being machinable yet not too invasive on the crankcase itself is required.

Epoxy fillers, such as Devcon and JB Weld, are not really sufficiently strong for this job, although JB Weld is used later in the project for filling in the threaded tappet block holes.

Welding, on the other hand, is certainly strong enough but the amount of heat that would be needed to fill these relative large holes and the danger of distortion and warpage to the overall structure that could result makes this a potentially hazardous option.

In the past, I have had good success using Lumiweld for filling holes in gearboxes with ruined threads and crankcase damage on other engines and therefore opted to use it here.

The “weld” part of “lumiweld” is misleading; it is not welding, it is a brazing process, Lumiweld melts at 390^oC whereas aluminium, typically, melts at more than 250 ^oC higher temperature. The strength of lumiweld is of the same order as aluminium. For the record, literature values for the UTS (Ultimate Tensile Strength) of lumiweld and 6061T6 (a medium to high strength aluminium alloy) are 60 kpsi and 45 kpsi respectively.

The process of applying lumiweld is, however, quite different from traditional welding, brazing or soldering. To get lumiweld to adhere to the aluminium surface, the surface must be scraped vigorously with a stainless steel rod during heating; this is of paramount importance. After cleaning the surface of grease and roughing with coarse emery cloth, the overall process that I have evolved for filling holes is as follows:

     1)    Block one end of the hole to be filled with a piece of stainless steel plate using a g-clamp or some kind of improvised jig. This is important to stop the molten lumiweld from simply running out of the bottom of the hole and it will not adhere to stainless steel allowing the plate to be removed later.

      2)    The structure can either be pre-heated in an oven or use oxy-acetylene to warm it up initially. As the structure comes up to a high temperature concentrate the heat in the area that is to be filled.

      3)    Melt a small piece of the lumiweld into the hole in the flame and then stir it around the hole vigorously with a stainless steel rod, keeping the heat on the structure. Keep the stainless steel rod out of the flame as it will simply melt it!  Eventually, the surface is seen to  “wet” and the lumiweld starts to adhere to the surface around the bottom of the hole.

      4)    Add more lumiweld into the existing puddle and continue stirring, working up the hole in the stirring/surface wetting process.

      5)    Eventually, when the top of the hole is reached, add some lumiweld around the outside of the hole (this ensures that the edge of the hole is properly covered) and build up an excess (ie a little “mound”) immediately above the hole (or rather, where the hole was). On cooling and solidifying this mound will contract and this ensures that there is enough material to machine back to a flush surface.

The picture below shows the side of the crankcase after hole-filling that was blocked off with a piece of stainless steel plate.

And the next picture shows the inside of the crankcase from where the hole was filled.

And finally, the crankcase is machined to clean-up the excess lumiweld and to form a recess, centred on where the drive for the overhead cam (which consists of a gear driven by the crankshaft and 2 sprockets for the overhead cam and magneto drives respectively – identical to the K7), into which an aluminium housing will be inserted to support the OHC drive itself. 

Here, the smaller holes in the crankcase will be retained but the larger holes around the outside (those with threads) that previously supported the tappets and a breather will later be filled with JB Weld.

The AJcette Project ….how it started ….part 2

Over the years, I have acquired many Mk1 OHC Velocette bikes, all as “projects”, and many other assorted bits and pieces. And so, rather than attempt to make the upper-part of the engine, ie the cylinder/head/cambox an easier solution would be to use Velocette parts, if at all possible, and graft these onto AJS crankcases.

However, AJS crankcases of the period are either very basic - the SV engine uses bushes rather than bearings for the crankshaft and are easily obtainable, or quite sophisticated - the OHC engines have strong crankcases and support a 3-bearing crankshaft, but impossible to obtain secondhand.

Nevertheless, AJS engines evolved fast during their days at Woverhampton and by the time the marque was sold to the Collier Brothers in 1931, the 350 Big Port (which, by this time, had been officially designated this moniker by AJS) had been developed into the SB6 which had a 3-bearing crankshaft not dissimilar to the K7 of 1928. By a stroke of luck, a good friend of mine (JT) just happened to have a pair of SB6 crankcases in his shed and so the first mock-up of an AJS crankcase/Velocette barrel combination was tried.

It transpires that if a pair of SB6 crankcases are spaced apart by about 0.4” then the spacing of the cylinder base retaining studs across the engine matches that of a Velocette Mk1 OHC cylinder barrel. And if the cylinder base flange holes of the barrel are elongated by ~ 0.020” in the longitudinal direction then the barrel will slide on to the retaining studs easily.

The first step, therefore, was to make a suitable spacer to move the crankcases apart properly (rather than a bunch of washers that had been used for a ”what-if”). Luckily, these crankcases have a locating flange when they were originally machined. The advantage of this from a production perspective is that if the crankcase mating surface, the flange and the main bearing machining is carried out in one operation but independently on the drive-side and timing-side crankcase halves, then one crankcase half can be matched to any other crankcase half and the main bearings will be in alignment. It is therefore possible to machine a spacer that is also flanged, with male and female sections on appropriate sides, and that will then maintain the concentricity of the drive-side and timing-side main bearings with the spacer installed.

The plan, therefore, was to use a large piece of 7075T6 aluminium as a spacer by machining the final thickness and the male/female parts of the locating flange in the lathe and to then send the assembly to a local company to water-jet cut the outside profile of the crankcase.

A sizeable piece of aluminium plate was therefore mounted on the faceplate of the lathe (for reference, the faceplate is 15” diameter) and machined.

And after water jet cutting to the outside profile of the crankcases….

The spacer and one of the crankcase halves was then mounted in the milling machine and the holes for the crankcase studs were machined and the excess material in the cylinder bore was removed.

At the same time, a slitting saw was used to remove a thin “slice” from the timing case surface that would later be used for part of the casting pattern for a new inner timing case.

Note the number of machined “holes” in the timing-side crankcase. There are holes for inlet and exhaust camshaft bushes and the cam followers/tappets in the OHV engine. None of these are needed in the OHC engine and need to be filled before work can be started on making a new OHC drive.  

The AJcette Project ….how it started

There are plenty of things in a workshop that can be dangerous. Large pieces of metal spinning round in a lathe chuck, milling cutters, disk cutter and other sharp tools; there are poisonous liquids – I use cellulose thinners for parts washing, 2K paint, methylene chloride for removing old tank sealer, all of which are very nasty chemicals and, of course, there are hot things from welding or heat treatment that will burn your fingers nicely! Accidents do happen, but if care is taken you can get to the end of each day without ending up in A & E.

There is something else that is not usually thought of as being dangerous but, nevertheless, can easily take a year out of your life. And that is having a bit of spare time to sit down and let the mind wander. Whilst I was restoring the K7, in a moment of idleness I mused “I wonder if I could convert my side-valve 350 AJS to overhead cam” ….and that was how the AJcette project started ....and another year went by.

Sometime previously, I had bought a 350 side-valve AJS on ebay. It was in Scotland and, as I live on the South Coast of England, was a long drive to go and pick it up. When I got there, it turned out to be in very original condition, non-running, no registration documents and, for some reason, somebody had painted it in cheap and very bright red paint. They hadn’t made a very good job of the brush painting, the mudguards and tank had been painted “in situ” without even cleaning the dirt off!

 

Luckily, the artist that had painted the bike had used some cheap household paint that simply wiped off with a rag soaked in cellulose thinners to reveal what was almost certainly the original enamel beneath.

One valve spring had broken and, as these are conical valve springs and I didn’t have anything that even closely resembled one of these in my workshop, I TIG welded the 2 halves of the spring together and, amazingly, it worked fine for a few hundred miles until I eventually sold the engine. I would not do that on an OHV engine! I found that the bike was good for about 46 mph flat out with me lying flat on the tank and clocked by the speedo of a following Velo; even that speed is probably optimistic.

As far as I can ascertain, all the AJS 350s of this period (1920s) had the same frame/forks/hubs/gearboxes etc (although ratios are different between SV and OHV/OHC bikes). This would therefore make an excellent starting point for a K7-lookalike …..I just needed an overhead cam engine and close ratio gears.

This story is about how the bike pictured above turned into this: