Thursday 16 June 2022

The AJS V-Twin Project: Postscript

The last time that I posted anything about the V-Twin was at the end of 2021 when the bike first ran on the road. This does not mean that nothing has been happening in the last few months; since that first road run I have been slowly ironing out the problems that I found. This is simply a part of the development process and I fully accept that changes and modifications are required as problems are uncovered.

Back In February, Alex Rollings from the Classic Motorcycle Channel came down and made a YouTube video of some of the bikes and we had a bit of fun running them up and down the road. Since the time that Alex visited I have been working on the new main project – the AJS 33/7 Trophy rebuild.

However, during the past few weeks progress has slowed a bit because we had a new addition to our household at the beginning of May thanks to the completely pointless and terrible war in Ukraine. Tanya and her daughter, who until early May lived with their family in Kyiv, are now living with us. They are both absolutely charming and Tanya, who is younger than both of our own daughters, is now getting accustomed to the V-Twin prior to her attempt on the vintage-class World Speed Record.

In case her husband is reading this ……I’m joking! Hopefully they will be more secure living in the leafy lanes of West Sussex rather than being in Kyiv.

If you have been reading my blog over the past year there is one feature that you may observe is missing from the bike in the above picture; the cambox scavenge pumps that were located at the top of each timing case have gone, to be replaced with simple aluminium covers. Apart from the brass plates that were attached to these covers on the original, the engine now looks more like the first incarnation of the bike.

This was not done for visual effect but rather the evolution of the lubrication system to try and resolve the problem of oil coming out of the cambox, particularly on the rear cylinder.

I have already described thelubrication system of the bike in some detail and also the problem that I had with the single cylinder AJcette cambox lubrication and I was expecting to have to have the same issue with the V-Twin – only worse because of the +/- 250 inclination of the cylinders. In spite of the fact that I had already plumbed in a bypass valve in the cambox feed line to dump excess oil into the crankcase, flow restrictors to reduce the oil flowrate into the camboxes and camshaft-driven oil pumps to scavenge the oil from the gutters in the Velocette camboxes I still had oil pouring out of the slot where the tappet exits the cambox.

The first step was to introduce an additional flow restrictor (each restrictor contains 2x 0.6mm orifices plus upstream filters) into the oil feed line to the camboxes. This is identical to those already in the feed line and now gives 4x 0.6mm orifices in series for each cambox.

By this time, I had a standard “run” for the bike of approximately 2 miles to assess the amount of oil that exited the camboxes. With the additional flow restrictor and with the dump valve fully open there was still too much oil being thrown out.

I did not want to reduce the amount of oil going into the cambox any further because invariably this will end up in damage to either the cam or the rocker skids (or both) and so it was time to try and make a quantitative assessment of how much oil was reaching the cambox by measuring the amount of oil going into the cambox. I’m sure there are inline flowmeters that could be inserted into the oil feed lines (I used to do this in my student days when I worked for Perkins Engines to perform a heat balance on truck diesel engines) but this level of instrumentation is not so easy for the home engineer ….and not cheap. The alternative is to measure the amount of oil exiting the cambox through the scavenge pumps which is at least some indication (oil entering = oil leaving through scavenge pump + leakage)

I therefore disconnected the scavenge pumps on the downstream side and rigged up 2x plastic measuring jugs, seen below:

…..and how much oil was measured? Nothing! Not a drop.

After a bit of pondering, I figured there were 3 possible reasons for this; firstly, the oil could not be sucked up the slight gradient from the collection points on the cambox to the pump inlet; secondly, the fact that the oil was being collected from 3 separate locations on the rear cylinder and at least one of these (the front gutter – exhaust side) probably didn’t have any oil would lead to the pump preferentially sucking air rather than oil; finally,  although gear pumps are self-priming, the fact that no oil was reaching the pumps would result in the pumps running dry.

To check the last of these hypotheses, small T-pieces with a detachable cap were plumbed into the inlet of both pumps:

that could be used to prime the pumps and the oil feed lines.

After priming the pumps, another test was carried out to measure the amount of oil exiting the pumps. Again, nothing!

Clearly this was not a solution and I therefore discarded the pumps:

(pity, it was a lot work making these) and inserted a couple of model engineering inline check valves into the cambox scavenge oil lines and fed these into the timing cases.

and made cover plates to replace the pumps:

The oil flow is now under gravity the whole way from the cambox to the timing case ….time for another road test.

Whilst not perfect, this seems to work fairly well and has reduced the rear cylinder cambox oil leakage to an acceptable level – essentially a fine mist that coats the rear mudguard. To continue road testing I have fabricated an L-shaped splash guard and attached some oil-absorbent material using paper clips.

While this may not be the most elegant of solutions it is a pragmatic way forward to be able to put more road-miles on the bike whilst containing the small amount of leakage oil.

During these road tests, some other minor modifications were made. There is now a new rear brake pedal

that gives more leverage and I fitted a larger guard to avoid sticking my foot into the clutch.

Both carburettor needles have been dropped one notch to weaken the otherwise rich mixture and I’m now running 35psi in the back tyre and 30 psi in the front which has improved the cornering.

It is now time to continue with running in what is essentially a brand new bike and to see what fine tuning is needed in the coming months. I'm hoping to take the bike to this years Manx GP (I go every year with something "old" and still have ferry bookings that were made in 2019 for the 2020 event and now bounced twice to 2022). Let's hope.....

No comments:

Post a Comment