enginehistory.org

[raustin]

Hello to everyone, I'm very new to the society (joined this year) and I've always been interested in engines and more so, aircraft engines.

One reason for joining is to ask questions about engines and their different arrangements and one in particular to start with.

I've not been able to figure out why successful radials have an odd number of cylinders. This is probably a very basic question that has been answered hundreds of times in the past, but I would like to know.

Also would what ever the reason is, be the reason why 'X' configured engines and two 'Vee's' connected haven't been successful.

Best Regards
Robert Austin

[jschauer]

What I was told when I was in school was the reason for the odd numbers is for ignition timing. Take a nine cylinder engine, the timing is 1,3,5,7,9,2,4,6,8 If it had only eight cylinders the timing would be 1,3,5,7,2,4,6,8???!!!! You would then have two cylinders firing next to each other, not good! In the multi row radials it is different, sort of, i.e. an R-1830, it has 14 cylinders with two rows of seven, but you still have an odd number in each row.

I don't know about the X engines, good question. Maybe the development wasn't worth the time during the war years?

[raustin]

Thanks for that, I thought it was something to do with the firing order. Does that work the same for a two stroke too?

[jjuutinen]

How about the Jumo 222? It has 6 four cylinder banks with radial style conrods and it is a 4-stroke?

[wallan]

No, it doesn't work with a two stroke. As all the pistons act on the same crankpin, the cylinders fire one after the other. In a four stroke radial, as each piston plays follow-the-leader, for smooth power application each cylinder alternates between complimentary actions, (induction - power, and compression - exhaust, strokes) so if there are an even number of cylinders, at some point in the cycle you get two adjacent cylinders applying power, causing an imbalance. The easiest way to imagine it is to draw a circle, and mark six points around it. Number them 1 to 6, and mark against them a + for power stroke, and a - for induction stroke, alternating as you go around. If you start on power stroke at 1, when you get all the way around, you find that 6, followed by 1, are both on the same part of the 4 stroke cycle

[raustin]

I was thinking more in line of a sleeve valve two stroke, where it would fire each time a piston hit TDC (figuratively speaking). But the explanations make sense, thanks.
The Jumo 222 though would seem to be two 180 degree 'Vee's' sharing a common crank, going by the explanation in another thread! Each Vee plane lying at 90 degrees to each other and where one opposing pair of cylinders operating as a+ and a- and vice versa for the other pair.
Wasn't the RR Vulture of similar configuration and where the Germans got the idea from for the 222 in the first place. They would have had plenty of examples of the Vulture to copy.

[jjuutinen]

Robert, Jumo 222 is a radial with 6 banks of 4 cylinders each. Each bank has a camshaft (not cam plate), one master rod and 5 articulated rods and there is one crankshaft.

Vulture was a true X-engine and it and the Jumo 222 have nothing in common except being 4-stroke liquid cooled piston engines.

A good basic description is to be found in "Jane´s Fighting Aircraft of the Second WW".

[raustin]

It sounds a bit like the Wright Tornado then. I've bought the book about it and am waiting for it to arrive.

This is one of the reasons I decided to join the society though. I've a lot of misconceptions it seems.

[gryan]

As disclosed above, for radial two-stroke engine you can have either odd or even numbers of cylinders per row. Firing order is 1,2,3,4, etc. You fire each cylinder in turn as it approaches TDC.

An interesting example of a radial two-stroke engine would be Nordberg's industrial radial engine. This was mounted with its crankshaft vertical and drove a generator located under the engine. That set up was used to generate electricity for aluminium smelting.

The Nordberg certainly never flew (too big and heavy for that) but it did have some innovative features. It had an even number of cylinders. There was no master rod. All pistons were attached to a "sleeve" by slave or link rods. The sleeve was rotatably mounted on the big end journal.

To prevent the whole show tangling itself up in knots (too many degrees of freedom) Nordberg designers restrained the sleeve. They used a true motion scheme. The big end journal was free to rotate inside the sleeve. The sleeve was restrained in such a manner that it executed an orbital motion inside the crankcase. The centre of its orbital path coincided with the centreline of the crankshaft mains and the radius of path was the radius of the crankshaft throw. It was indexed on the big end journal so it did not rotate relative to any fixed point on the crankcase. That is, it would present a constant aspect to any point on the crankcase.

There were two methods employed to stop the sleeve from rotating uncontrollably. In early engines they used a gear train to control the sleeve. In later engines they used a system of linkages that tied an opposing pair of conrods together, each of which was also linked to the sleeve (looks to be a variant of a Watts Link). I've been trying to find the patent for this one but no luck so far.

I wonder if any of the Nordberg radials are still in operation?

[szielinski]

Not really a radial engine, but a barrel 2-stroke with a cam-crank, and odd number of cylinders and an even number of lobes can have alternate firing (non-sequential).
Eg 9 cylinders, 4-lobe cam-crank (2 complete sinewaves) gives a firing order of :-
1,6,2,7,3,8,4,9,5,1,6,2,7,3,8,4,9,5.

[gryan]

I found it! After expending a not inconsiderable effort searching the patent database I was lucky enough to locate the true motion mechanism developed by Nordberg for their radial engines. The patent number is US 2584098. The inventor was D I bohn. This patent discloses a brilliantly clever invention; a linkage that results in all pistons executing the same motion no matter which station they are located at.

Nordberg radials were two stroke engines with an even number of cylinders. Note that two-stroke radial engines can have any number of cylinders whereas a four-stroke radial must have an odd number of cylinders.

At this stage it would appear this mechanism requires an even number of cylinders to work as there is the requirement for two of them to be 180 degrees apart. I'll search further to see whether anyone developed variants of the mechanism to allow for odd cylinder numbers per row.

Regards

Gerald

[jreiter]

I am a little late on this subject. A radial with a cam ring has an odd number of cylinders in each row because the geometry of the engine only works out with an odd number.
I remember when I was studying for my A&P. The cam ring and firing order were hard to understand. My father and other mechanics could not make a good explanation. I used the chart for cam lobes and direction of rotation for the number of cylinders in a row. I chose a 3 cylinder engine with single cam lobe turning opposite engine rotation at 1/2 speed. I drew that out.
Assuming that the intake valve opens at TDC and the master rod is in #1 cylinder. The #1 intake valve will start to open at TDC. 240 degrees later the crankshaft has rotated to bring #3 piston to TDC. At the same time the cam ring has rotated 120 degrees in the opposite direction to bring the lobe to the position to start to open the #3 intake valve. That is why the firing order of a single row radial is odd cylinders on the first rotation and even cylinders on the second. Works the same for a 2 lobe cam ring turning 1/4 speed in the same direction as the engine. A 5 cylinder can use a 3 lobe turning opposite at 1/6 speed or 4 lobe turning 1/8 speed the same as engine rotation. On a 9 cylinder it is a 4 lobe at 1/8 or 5 lobe at 1/10 speed cam ring. There is a slight degree difference between cylinders caused by the action of the master rod and articulating rod paths. But the radial is a compromise as all engineering is.