ISTR hearing that a radial engine must have an odd number of cylinders per bank. But I don’t remember why. Something about some cylinders never firing if they were an even number. I suppose you could call a two-cylinder boxer engine a radial, technically. Aside from that, are there any radial engines with an even number of cylinders per bank? Assuming not, what is the mechanism that causes (IIRC) cylinders not to fire?
Take a look at this illustrated article. It doesn’t answer the question directly but you can see how it could be just a minor quirk of geometry.
It helps on the firing sequence to have an odd number, that way the torque that it´s applied to the crankshaft is more even. Suppose a five cylinder engine (sorry for the dots, with blank spaces it doesn´t show right):
…1
2…O…5
…3…4
The fire sequence would be 1-3-5-2-4, evenly spaced around the clock, so to speak.
A 6 cylinder:
…1
…2…6
…O
…3…5
…4
That leads to 1-3-5-2-4-6; notice the "jump between 5 and 2, it skips two cylinders intead of one.
Why not making them fire in simple sucessión?, the folowing cylinder is not in the best postion to excert the correct push on the crakshaft, the geometry of the vectors isn´t good.
And… for a 4 stroke engine, the following cylinder is not in the correct stage of the combustion sequence as you can see in Shagnasty´s link.
Actually, on an even numbered engine the next cylinder [is] in the correct phase. But that still leaves the problem of bad firing rhythm, as you nicely point out. Sequential firing doesn’t have a problem of bad geometry, just really inconsistent firing rhythm. Your basic choices are for a six cylinder engine:
BSBSBSSBSBSB
or,
BBBBBBSSSSSS, with B = Bang and S = Suck for power and intake strokes.
I had always wondered about this, and just the other week got an answer. Four stroke engine only fire every other revolution.
Intake (piston goes down)
Compression (piston goes up)
Power (piston goes down)
Exhaust (piston goes up)
Lather, rinse, repeat.
Radials fire every other cylinder on each rotation. So in a 9 cylinder radial on the first revolution pistons 1, 3, 5, 7, 9 fire. On the next revolution 2, 4, 6, 8 fire. If you had an even number of cylinders (8) then 1,3,5,7, would fire on the first revolution, and on the next rotation 1, 3, 5, 7 would fire. The even numbered cylinders would just go along for the ride.
Not only would an even number of cylinders produce an uneven firing pattern, but you would have to change the way the cams work. The cam in a radial is a large ring that turns at 1/2N the crank speed, where N is the number of cylinders per row. There are multiple exhaust and intake lobes, (arrainged in two “tracks”) and in 2N crank rotations, all the cylinders will have used all the lobes one time.
There is no way to make this work for an even number of cylinders except by ommiting one jug from what would otherwise be an odd-cylindered engine…and that would result in the engine being bulkier than required, but given the visibility problems (esp. when taxiing a taildragger) associated with radials on a single-engined airplane, it wouldn’t suprise me if leaving the top jug off has been tried.
A big problem with an uneven firing order is that the pilot wants to hear a nice even firing rate, and wants to know the instant it becomes uneven. If the engine doesn’t run smooth normally, then that is a safety issue, because it makes problems less obvious.
Thank you Kevbo. You just reminded me of a display that’s down at the San Diego Aerospace Museum. It’s a radial engine model that rotates around to show how it works and then splits open while still running to show you even better.
I had forgotten about the cam design. It’s clever and elegant. I guess I know where I’m going tomorrow afternoon.