Anzani Engine Project

[David Paule]

On a whim, we pulled the oil pump out of the box recently and took a look at it. I had previously disassembled it, so we weren’t able to examine it all together at this stage. We were curious, since the Anzani uses a pulsator tube instead of an oil pressure gauge, and therefore probably has low oil pressure compared to today’s engines, if we could get an idea of the volume of oil that is being circulated. I had thought that the Anzani’s oil flow was characterized by low pressure and high flow rate. Apparently it’s both low pressure and low flow rate. I think it was Craig who pointed out that the flow rate had to be low, since the oil system is a dead-loss system, to allow sufficient range for the aircraft. Roger said that the flow rate on modern engines is hundreds of times larger than this.

Here's a discussion about this.

It's worth bearing in mind that we didn't have the oil pump piston return spring out of the box of parts to assess.

Craig wrote -

Oil pump thoughts -
I think I have figured out how the oil pump works, and it isn’t pretty. The pump piston is driven in its cylinder by the pump cam. The cylinder has two ports: the output port that has a ball check valve, and the inlet port. The inlet port has no check valve and is always open to the oil filled compartment in the oil pump housing. As the piston reciprocates in the cylinder, the oil is forced in and out the inlet port. Because the inlet port is on the small side (about .16 inch or 4 mm dia.), the oil in the cylinder experiences pressure oscillations. The pump depends on the positive pressure pulses being high enough to unseat the ball check valve and squirt a little oil past it. The designer probably gave it a cam that rapidly drives the piston up and slowly withdraws it so as to make it more efficient.

The advantage of this design is that it only uses one check valve. The disadvantage is that its output is highly dependent on the viscosity of the oil. A low viscosity oil will just squirt out the inlet hole without generating sufficient pressure to unseat the check-valve. With the total loss oil system you don’t have to worry about the oil getting hot except on a hot day. Then the lower oil viscosity will result in lower flow just when you wanted more.
The real disaster, however, comes in the modern day when people like us decide to use modern oil. Modern synthetic oil has much better lubricity and the lower viscosity buys you…oops. The lower viscosity buys you no oil flow and a seized engine.

In order to use modern oil in the Anzani engine the oil pump has to be modified with a second check valve. I would hesitate to even use a 90 weight synthetic or modern castor oil because of the uncertainty in the viscosity of the castor oil used in 1918. Besides, we don’t know what gave these engines their poor reputations. The good news is I think we can easily modify the pump to take a second check valve.

Dave responded -

I'm not sure this is the way it actually operates.

I think that the intake port gets sealed off by the pump piston as it passes past the holes. So from that point on, there's an oil charge that's going to have to get past the outlet check valve because there's no other place for it to go. If this is correct, and I'm not looking at the hardware at the moment, just a sketch of it, it's a positive displacement two-cycle pump. The inlet "valve" is probably akin to a two-stroke motor valve, although I confess to being even more ignorant about those than I am about this. I think I've seen that sort of thing on model airplane engines when I was a kid.

Here's how it works, step by step. On the opening stroke, as the piston retracts, first it leaves a vacuum in the chamber when the check valve closes and seals. Then it slides past the intake orifice and that vacuum, plus the ready supply of oil with some head (a foot or two), fills the chamber. Then on the compression stroke, the piston pushes excess oil out the intake holes until the holes are sealed off. From that point to the end of the compression stroke, all the oil in the chamber is being pushed out the pump.

This depends on the stroke being reasonably large, but except for the loss of oil past the piston (which doesn't have sealing rings, so there might be some loss there) it supplies the same amount of oil regardless of viscosity. Perhaps more when it's thin than when it's thick, if the thick viscosity has difficulty filling the chamber. Thick oil would stress the pump more, obviously.

Proper setting up of the pump would have the piston opening no farther than the extent of the holes. Ideally, it should compress as close to the check valve or end of the chamber as possible.

Craig replied –

With regard to the oil pump, that is how I first thought it operated. However, I didn't like the idea that the piston, after giving its shot of oil, had to retract without the port being initially open. This would require the oil in the chamber to flash to vapor. It would also require the piston return spring to be strong enough to do it. Otherwise the piston/cam follower would just hang in the up position. The answer to this question is easily found by screwing the pump fully into its chamber and looking to see whether the piston travels beyond the port.

As to the crankcase compression idea, two-strokes have a valve at the inlet to the crankcase. It is either reed valves or a disc valve, or at least a piston port. Maybe not a check valve but at least something to allow the gas to be compressed. The oil pump has none (I think).


Roger added -

Well, Craig, Dave......those are interesting thoughts on the oil pump. I can see that moving oil via a pressure pulse instead of positive mechanical pressure keeps the pressure down and also avoids any potential dangers that are always possible with positive pressure pumps....not that I can recall ever seeing engine damage done by hydraulic "overpressure" from the oil pump, but in the early days they might have worried about it happening.

Maybe that is why so many of the oil systems we see on early engines were fairly low pressure. I've wondered about that. Surely they knew how to make the pressure higher if they had wanted to.

Here's an interesting aside: Yanmar is a large manufacturer of high quality diesel engines, and they still believe in low pressure for their diesel motors. Their engines have a direct drive positive pressure oil pump with no low pressure shutoff and a typical idle operating pressure of less than 5 psi. However, there are two different high pressure shut-offs that kick in about 35 to 40 psi.

The two stroke engine description should include one other pressure cycle. On a two stroke engine the volume of the crankcase is used to store a charge pulled in as the piston goes up and the pressure in the crankcase goes down.

After being pulled into the crankcase, the charge is compressed by the descending piston and also swept by the crankshaft counterweights - both of which move the charge in the direction of the transfer ports that feed it onto the top of the now-rising piston.

Just for grins, I think we ought the lift and duration of the oil pump cam. That would shed some light on which philosophy they used for the oil pump - was it pressure pulse or positive pressure.

Later, David wrote -

Got to thinking that if the piston is 1/4 inch diameter, and there's 20 psi of oil in the oil pump housing there (much more head than is likely, plus sea level atmosphere on it, and a vacuum on the other end of the piston) there's only about one pound of unbalanced force trying to prevent the piston from opening up. That's not a huge spring. And it would be a static preload on the piston (which may be why there was wear on the cam and lifer, in spite of being immersed in oil). That is, at the extreme open position the preload would be about a pound, and the cam pushes against the spring from there.

Then Craig -

Dave, You're right. I had also come to the conclusion that it wouldn't take much to vaporize the oil. That would be a unique way to operate a pump and now I am really curious.

[David Paule]

Yesterday was one of those gusty bright Fall days. The leaves were falling, probably why it’s called “Fall,” and because of the conditions and leaves all over the place, we kept it simple.

Here’s what we learned about the gear train. These gears are at the very back of the engine, behind the cam and the tappets. There is no propeller reduction drive for this engine, and the maximum speed is 1,300 rpm.

1. The crankshaft gear has 20 teeth and is held on with a nut that has left-hand threads.

2. It drives the oil pump gear. This 40 tooth gear directly drives the oil pump at half engine speed. For future reference, the oil pump plunger runs at a maximum of 10.8333 pulses per second at maximum engine speed. You can see the oil pump gear in the left-hand picture below.

3. The oil pump gear, in turn, drives a small 15 tooth idler gear. This idler gear, the only one in the engine, runs at 1.333 times engine speed.

4. It drives the cam gear. The cam gear has 40 teeth, and rotates at half the engine speed just like the oil pump gear. The cam is attached directly to this on its forward face. The cam gear itself is located coaxially with the crankshaft, forward of it, and rides on a bushing on the crankcase itself.

Those gears are all inside the crankcase and are lubricated. The crankshaft sticks out of the back of the crankcase, and the accessory drive housing fits up to that. The accessory housing holds the oil pump, which you can see in the right picture, and supports the single magneto on the shelf

This whole housing assembly is mounted aft of the engine mounting points. Since the engine usually was mounted to a formed steel forward bulkhead, as was the usual practice in WW I, that means that the magneto and oil pump are probably in the cockpit.

Above the shelf, to the right of the oil pump in that right-hand picture, you can see the hole where the crankshaft fits through.

5. The final gear on the crankshaft is attached to this crankshaft stub. It’s a 21 tooth gear. I couldn’t find the nut that holds it on, and this worries me a little.

6. And it drives the 14 tooth magneto drive gear. The magneto, then, turns at 1.5 times crankshaft speed. In the right-hand picture, the magneto drive gear is placed down and to the right of the hole, in that open space there.

Roger pointed out that there’s a gear to snag a shoelace on each foot.

These gears, being external to the crankcase, will need to be lubricated before each flight. On some of the early aircraft, you can see an access panel on the right side, low and forward. It’s probably to let the mechanic do this. Also, since there’s no way to retain the lubricant, that whole area will probably get messy. It would make sense to provide some sort of light cover assembly that can be removed for cleaning, if that wouldn’t adversely affect the magneto cooling.

I wonder what was used to lubricate those gears?

We determined that the magneto fires twice each revolutions, for three sparks each time the crankshaft revolves. As this is a four-stroke engine, with each of the six cylinders wanting and needing to fire once every two crankshaft revolutions, it works out perfectly.
.

[DutchNostalgicWings]

Once you got the engine ready and fired up I would be gratefull to hear from you.
Overhere at Hoogeveen airport in the Netherlands, a friend (Douwe v/d Werf) is building a Pietenpol Air Camper and he has an Anzani engine which he hopes to use for this project. The engine he owns has been overhauled but still needs to be installed in the airframe and tested so we would love to hear your experiences working with this engine.

The Pietenpol probably needs another year before its first flight.

If other members have user reports for this engine, they are welcome to share them as well.

Cheers,

André Jans

The engine, Douwe and the plane.

[David Paule]

That's a very pretty engine!

Does it have the carburetor? How about intake valves controlled by the cam?

The Pietenpol will look really good with it. Is your friend's Pietenpol set up for the lighter weight of the Anzani? Originally, the Pietenpol was designed for the heavier Ford Model A, I believe.

I'll be in touch as we get closer, and of course our progress will be reported right here.

Incidentally, our Anzani is a later model, slightly, with the domed forward crankcase. I'll post a picture later when we get to it. We have a similar exhaust collector, but it doesn't have the flanges at the bottom. We have the prop hub and a prop waiting for it.

[DutchNostalgicWings]

David,

this is the only shot I have so far as the engine is still at his home and not on the airport. I will ask him to bring it so I can shoot close up pictures of what he has right now. We already have our doubts if the engine will be suitable for the Pietenpol as mentione dit only produces between 50 and 60 HP power which might be on the lower end of what's needed.

[David Paule]

We'd be interested in a picture of the back of the engine, from the right rear; another of the top of the engine; and another of the bottom of the engine. Perhaps one of those might show the top of a cylinder and the valves.

Then I can comment upon the differences between yours and my slightly later one.

Thanks!

David

[David Paule]

Today's work explored the lubrication system some more.

The oil pump is located at the back of the engine, actually behind the mounting bulkhead. For most installations, that will put it in the front of the cockpit. There's a large inlet and a small pressure outlet. The oil seems to go externally to the engine to a fitting at the top back of the engine, and another at the bottom front.

In the first picture, the inlet fitting is shown loosely installed at the bottom of the oil cavity. It's got a white tag hanging from it. The hand is holding the tachometer drive cover next to the tach drive shaft. The tachometer drive is 1/2 engine speed, in the reverse direction (counter-clockwise here).

The second picture is a poor one of the inside of the oil pump. Unfortunately it doesn't include the piston. It does show the brass pump body, the tapered oil pressure outlet line with the inside showing, and the check valve spring, retainer and ball. The check valve retainer has a square hole, so that even when the ball is blocking it, some oil can escape. This probably helps the return spring push the pump piston back.

The return spring isn't shown. In fact, I'm not sure we have it. The piston weighs 18 grams, and we know the dimensions of what the spring must be, so I'll see if I can find a stock spring. If not, I'll design one.

In both cases the oil flows through the case to the crankshaft, where it goes through passageways to the ball bearings - this engine uses ball bearings - on either side of the two throws of the crankshaft, there's no bearings in the middle. There's a bearing at the very front of the case that the prop hub seems to retain, and one more in the accessory case at the back.

In the case, the front case has ample drainage and the rear case minimal drainage. The low drainage in the back is probably a way to ensure that the gears are lubricated. At the bottom of the case, there's a drain plug for the front and the back crankcase halves, and on top, a breather for each.

At the back, there's an internal copper tube that goes from the pressure oil fitting to the rear crankshaft journal at the cam. That copper tube appears to be cast in place.

Incidentally, the casting and machining on this engine is first-rate.

The drain plugs are clearly not intended for daily use. They are at suitable locations for a possible oil return system, and I believe that I have a service bulletin recommending that. If so, then we don't know what the oil return pump should be.

[DutchNostalgicWings]

Originally Posted by David Paule

We'd be interested in a picture of the back of the engine, from the right rear; another of the top of the engine; and another of the bottom of the engine. Perhaps one of those might show the top of a cylinder and the valves.

Then I can comment upon the differences between yours and my slightly later one.

Thanks!

David

Will do!

Thank you for your help by email so far.

[David Paule]

Last week we decided to run a test of the oil pump as it is right now. Since the plunger-type pump didn't have a return spring, I weighed the piston and estimated the spring rate we'd need. Then Roger and Craig measured the inside diameter, outside diameter and length that it should have, and at the moment a set of springs is on order from McMaster-Carr.

Also, we've got some castor oil and some Aeroshell 15W50. There's some Aeroshell straight-weight oil on order, too.

We hope to take at least a first pass at that this coming week.

One of the key unanswered questions that we'll have, is how much oil flow is enough?

While we were talking about the lubrication requirements I brought out the oil breathers. There are two of these, one for the forward case and one for the aft case. If you look closely at the blue engine you can see the copper ends where they go. On this engine the breathers are somewhat different. The picture shows one assembled and one opened up.

Left to right, you'll see the upper cap, then the oil screen and the oil drain, then the breather body. When it's assembled, there's a bit of play axially, so that the screen assembly can be pushed upward. I think that the internal pressure in the case does that. It's necessary for the oil vapors to escape, with the path being up past the conical bottom of the screen, through the screen, and out the top. Any oil that wants to can dribble down the narrow drain, where it'll fall on the crank journals.

When the engine's off, gravity pulls the screen assembly down (down is to the right; left is up for these in this photo) and that keeps large dust and small squirrels out. And even at that, it won't prevent squirrels from putting their acorns in the vent and clogging it.

Unlike a more conventional oil screen, this one is simply a guard to prevent debris falling into the engine. It has no filtering capability - or rather, it's got a negative filtering capacity. Remember that the oil system is a dead-loss system? If there's any debris in the oil, it'll get retained. It's the cleaner oil vapor that escapes. The crankcase has two drains, and it's important to drain those sumps often, apparently, since whatever oil is down there gets gradually dirtier.

If we can get the oil pump to lift about 12 inches then we can make a drain tank, suck from that, and have an oil return that way. We can put a screen in there somewhere, if we can achieve that.

[David Paule]

Looks like we are going to be weathered out for the second consecutive week.

And one of the guys is on vacation....

Sorry!
.