[David Paule]

This one got delayed due to the lack of photos. Please see the bottom for those.


Craig wrote -

This week we planned to test the oil pump (off the engine) by running it with a power drill. We mounted the pump housing to a board and ran a plastic tube from the pump outlet to a needle valve with a pressure gauge between the pump and needle valve. The plan was to run pressure/volume curves for several different viscosity oils, including the original castor oil.

Unfortunately, we didn’t get to do any testing. We assembled the pump with a new return spring (old one missing) and tried to run the pump with the drill. We found that the cam lobe that runs the pump was hitting the inside of the oil pump housing. This presented a mystery. These were all parts that had run together before. Was the shaft bent? We couldn’t tell without removing the gear from the shaft so as to examine it. How would the shaft get bent with both ends supported in bushings? It seemed unlikely that a blocked oil outlet could have done it. There was no sign of something going through the gear teeth. Also, the shaft didn’t really look bent. The mystery would only be solved by removing the gear from the shaft, and the shaft from the housing.

The gear was pinned and keyed to the shaft, and the pin came out easily with a pin punch. The gear should have then slid off the shaft, leaving the key in the shaft. I tried roll head bars. No luck. I didn’t want to put too much force on the bars for fear of breaking the housing. I thought about heating the gear, and tested it for hardness with a file. Hard. Chances are it is a simple high carbon steel and heating it enough to do any good would anneal it. I then supported the housing on a thick aluminum plate with a hole in it for the shaft to project into and proceeded to drive the shaft through the gear with a hammer, using an aluminum drift to protect the shaft. It was slow going, but eventually the shaft dropped through the hole.

The results: The shaft was straight. The bushing had been doing the supporting of the gear as I pounded on the shaft so I expected it to have some damage. It had been loose so I planned on replacing it anyway. What I didn’t expect to find was to find a keyway neatly cut through it by the key that was still in the shaft. Oops!

Mystery solved. The bushing outside diameter had a steep taper and had a larger diameter shoulder on the large end. Apparently the bushing had popped loose in operation. This gave it big clearance as the bushing walked out of the taper bore until it was stopped by the gear. With the new large clearance the shaft was free to wobble. The cam lobe hit the housing at the same place every revolution because our new cam spring pushed it there.
Before we go any further, we are going to need a new cam bushing. I recommend that we go modern and give it a uniform outside diameter.

Dave asked why not use a gear-puller.

Craig answered -

I have a couple pullers but the housing casting wrapped around the gear so closely that I couldn't get two legs at 180 degrees. Roll-head bars are miniature pry bars (like crow-bars). Once again, I couldn't get two at 180 degrees, and I put enough force on one as I thought prudent without a hint of motion.

I haven't a clue as to why they tapered the bushing. It seems like a very difficult way to get an interference fit. Also, if the bushing pops loose, as it did, you have the bushing shoulder applying axial load to the back of the gear, and the cam lobe applying that same load to the other end of the bushing and wearing it and the case. Odd indeed.

Roger added -
My vote is for photos. I always prefer having pictures. Frankly, I was having trouble visualizing the problem with the oil pump drive shaft and bushing - which I now realize was simply because the problem is so bizarre. Hats off to Craig for figuring it out. Craig brought the disassembled parts over to my house yesterday on a motorcycle ride and the problem immediately became obvious. But it is so unusual that I've still want to do some measuring, think on it, and look under the microscope in order to prove to myself that the taper that Craig found was deliberately but there by Anzani. If so, it is a remarkable engineering error because it creates an undesireable lateral force out of the predominately radial force on the bushing....the result being that a portion of the reaction to the force on the gear teeth ends up loosening the bushing!!

I've seen a steel shaft manufacture a remarkable imitation of a deliberate taper into soft materials while it is rotating. We've probably all seen that. All it takes is for one part of the rotating shaft to be under more radial load that the other. About 40 years ago some manufacturers tried rotating their overhead cams directly in the aluminum head castings. The result was to wear a perfect taper. But I don't think that is the case for the Anzani - though I still want to look again. My initial thought is the same as Craig's...the taper appears to be in the wrong direction to be caused by wear.

When I mentioned that, Craig pointed out that we have already seen that the wrist pins were deliberately machined as tapered parts. So there is a chance that tapered parts were some sort of Anzani design philosophy. If so, we can expect to find it elsewhere in the engine.

I don't see the advantage. A bushing with parallel OD and ID is simple to make and constrain. It can move soak up a lot of end-wise or lateral movement without affecting the radial clearance that it is tasked with controlling. That is not so with a taper. Tapers are difficult to make, incredibly hard to locate against a shoulder, and once it looses preload it no longer functions as a bushing.

Dave added -

One thing that's not obvious is that the oil pump shaft is supported at three points. If you look at the Oil Pump Housing Front photo, you'll see the shaft sticking out. That end fits into a bushing on the rear crankcase half. Then the gear goes on, and the bushing fits in the hole. Finally, the end we can't see rides in the tachometer drive cover. All three bushings need to be in good alignment.

Incidentally, it's not clear yet whether the oil pump shaft cam is part of the shaft. The journals on either side of it are not part of the shaft....