Multi-Row Radial Engine Questions

[Jane-DoH]

I remember reading about how in the 1930's they had trouble with radials that had more than two rows due to airflow cooling issues. I remember reading that engines that were developed in the 1940's such as the R-4360 got around this by offsetting the cylinders, forming a semi-helical arrangement.

My questions are

Didn't the twin-row radials have a cylinder offset as well?
Why didn't they think of that earlier -- the idea isn't exactly a leap of imagination as it would simply be an extension of the twin-row radials?

[Avionker]

I certainly don't claim any kind of expertise in this field but the first thing that springs to my mind is airflow limitation through the cowls.

If you have multiple rows of cylinders the cooling air which each row receives is going to be hotter than the preceding row, due to the air having already passed around a row, or more of cylinders. Also if the rows of cylinders are off set then the airflow is going to be slowed more than if it is passing over only one row.

The amount of airflow will be dependant on the diameter of the cowl opening which will of course be limited by drag considerations.

As I say these are just my gut reactions and may well be way off the mark. I'm sure many people, more knowledgeable than I, will be along presently to give detailed responses.

[411A]

The engine cowls on large American-made 4-engine piston transports were of the NACA pressure design, thereby increasing engine cooling airflow....greatly.
Cylinder row offset was very necessary with the R4360 engine, as if it had not been done, the last cylinder row would get very hot indeed.
Even with cylinder row offset (with attached baffling), the last cylinder row was much hotter than the others, at least on the Boeing Stratocruiser...which I operated for a short while.

NB.
Very nice airplane to fly...the engines were very smooth, almost turbine-like.
The B377 was also a pilots dream...the Flight Engineer did all the hard work.:D

In addition, the R4360 design was driven to a large extent by the specific requirements for the B-36 very long range bomber, as the engines on that type were buried in the wing.

[Avionker]

I'll get back to my avionics then.....:O

[SNS3Guppy]

Twin-row radials were offset. Same for the 4 row 4360.

Cowling on most piston engines is crucial for ensuring proper cooling. A few radial installations leave the engines completely exposed, and this is usually done only on single-row installations. A few utilize a pressure ring, a few simple pistons just use eyebrows or baffles, but most installations use pressure cowls with baffling to control airflow for maximum cooling. In most cases, cowl flaps are also used to vary the airflow and increase cooling when needed.

* Didn't the twin-row radials have a cylinder offset as well?

Yes.

* Why didn't they think of that earlier -- the idea isn't exactly a leap of imagination as it would simply be an extension of the twin-row radials?

They did think of that earlier. Offset cylinder rows are a basic design feature of radial engines.

The 4360 used a gradual offset to the right, or clockwise, as seen from the front, with airflow spiraling clockwise aft. An intricate baffling system and induction system was used to provide what limited airflow there was, to reach the rear cyclinders. The whole package was a royal pain in the butt when it came to changing a cylinder, which is a very common act on old, large radial engines.

[barit1]

Not exactly on topic...

The B-17 with its Wright Cyclones (single row, 9 cylinders) running 1200 hp. had cowl flaps, as did most large radials.

But the prewar Boeing 307, basically the same design but with earlier Cyclones rated 900 hp, had a simple NACA cowl without flaps.

[SNS3Guppy]

Ah, the relentless march of technology.

Rigging cowl flaps on large radials is somewhat of a pain in the keyster.

[Old Ag]

Careful cylinder offset was also necessary to combat vibration inertial forces.Limiting vibration in large double row radials requires masterful engineering with master rod offset, crankshaft design, vibration dampers, gear reduction and propeller selection etc. An ideal offset for cooling meant a slight compromise on the offset for vibration. Vibration in engines like the R2800 really drove much of the deign.

Old Ag

[Jane-DoH]

411A

The engine cowls on large American-made 4-engine piston transports were of the NACA pressure design, thereby increasing engine cooling airflow....greatly.

What's the difference between the NACA Cowl and NACA Pressure cowl? When was the latter first developed?


SNS3Guppy

most installations use pressure cowls with baffling to control airflow for maximum cooling. In most cases, cowl flaps are also used to vary the airflow and increase cooling when needed.

Understood. What was the first time the technology to produce a radial installation with pressure-cowls -w- baffling, and cowl flaps came into being? What was the first aircraft to have such a set-up?

The 4360 used a gradual offset to the right, or clockwise, as seen from the front, with airflow spiraling clockwise aft. An intricate baffling system and induction system was used to provide what limited airflow there was, to reach the rear cyclinders.

What made the baffling and induction system so intricate? Did the technology exist before this to produce a three to four row radial?


Old Ag

Careful cylinder offset was also necessary to combat vibration inertial forces.Limiting vibration in large double row radials requires masterful engineering with master rod offset, crankshaft design, vibration dampers, gear reduction and propeller selection etc. An ideal offset for cooling meant a slight compromise on the offset for vibration. Vibration in engines like the R2800 really drove much of the deign.

When did the technical capability first come along to be able to balance all of these variables to allow a three/four-row radial to exist?

[SNS3Guppy]

What was the first time the technology to produce a radial installation with pressure-cowls -w- baffling, and cowl flaps came into being? What was the first aircraft to have such a set-up?

I have no idea.

What made the baffling and induction system so intricate? Did the technology exist before this to produce a three to four row radial?

Technology was a progressive thing. There's really not a lot of technological increase in the engines introduced during the war years, other than they were a steady development to bigger and better using the same general technology. The biggest advent in power production wasn't the 4360 having more cylinders. The most technologically advanced radial was the R3350, with it's three power recovery turbines.

The 4360, in my opinion, is one of the best sounding of any radial ever built. Listening to the C97 idle was like listening to a symphony.

Baffling is intricate in any installation, in that minor changes in the baffles can result in hot spots, insufficient cooling and can shorten engine life or even lead to cylinder cracking. Baffling on any piston engine, not just radial engines, is crucial, particular to the individual installation. That is to say, an engine in one installation isn't necessarily baffled like the same engine in a different aircraft. In twin radial aircraft, the baffle arrangement isn't that complex, but it was especially critical in the 4360, for cooling to the aft cylinders.

[Old Ag]

The breakthroughs for the large double row radial came around 1939 for P&W and 1941 for Wright. It was all slide rule and trial and error. For the R2800, P&W broke many engines on the test stands trying to solve vibration with damper designs, rod design changes, forging techniques, bearing changes, machining changes and material changes. By the time the produced a production engine, vibration was within spec and the production changes improved longevity. The R2800's reputation for reliability speaks for itself.

In terms of vibration, the development of the R4360 was less troublesome (4 rows of 7 cylinders) than the R2800 (2 rows of nine). However, the R4360 did break crankshafts during testing due to resonance issues within the crankshaft. P&W solved the problem by strategically adding and removing metal from various locations along the crankshaft's length.

Both P&W and Wright had issues with development of their respective 18 cylinder two row radials due to vibration. But both were able to solve the vibration issues to produce two engines (P&W R2800 and Wright R3350) that would eventually demonstrate the full potential of the radial engine.

-Old Ag

[Big Pistons Forever]

The other factor in the development of the large piston engine was the advent of high octane aviation gasoline. General Jimmy Doolittle's greatest contribution to aviation was not all his incredible feats of flying it was his role in pushing the development of 100 Octane Avgas in the 1930's.

Post war availability of 115/145 Octane Avgas enabled another increase in horsepower to give enough available power to allow the development of the final (and finest :) ) generation of intercontinental piston airliners.

[1-26E]

B-29 Superfortress - Chrysler and Dodge contributions (http://www.allpar.com/history/military/b-29.html)

Here is a history of the development of the B29 and the concurrent development and production of the R-3350. Makes interesting reading. One wonders at the pace that the aircraft and engines were developed and produced.

[Jane-DoH]

SNS3Guppy

Baffling is intricate in any installation, in that minor changes in the baffles can result in hot spots, insufficient cooling and can shorten engine life or even lead to cylinder cracking.

At the penalty of sounding mildly retarded, I'm wondering if anybody ever thought of blow-in doors similar to the kind used on gas turbines? It wouldn't be used for thrust; it would be used to allow cooling air in but the arrangement would be similar overall as I understand it.

Baffling on any piston engine, not just radial engines, is crucial, particular to the individual installation. That is to say, an engine in one installation isn't necessarily baffled like the same engine in a different aircraft. In twin radial aircraft, the baffle arrangement isn't that complex, but it was especially critical in the 4360, for cooling to the aft cylinders.

So the baffles had to be arranged differently in each aircraft it was fitted to? How much harder was the baffling arrangement in a four-row radial versus a three-row radial? And how much harder was the baffling in a three-row radial versus a two-row radial?

[barit1]

I'm wondering if anybody ever thought of blow-in doors similar to the kind used on gas turbines?

Fair enough question; but blow-in doors in a jet inlet only work at low IAS where the static pressure inside is less than on the outside of the inlet. The compressor (or fan) draws a partial vacuum that overcomes spring force of the doors. As IAS builds, the static pressure inside rises to equal and then exceed static pressure outside, and the doors are forced closed.

A radial engine cowl never sees this pressure differential. (Well - The only possible situation I can think of would be if there were a cooling fan in the system to create a partial vacuum...)

[Jane-DoH]

barit1

Very good point :ooh:

[osmosis]

Suggested reading:

Piston Engines (http://www.enginehistory.org/engines.htm)

[Jane-DoH]

osmosis

Thank you

[barit1]

The 4360, in my opinion, is one of the best sounding of any radial ever built. Listening to the C97 idle was like listening to a symphony.

Pratt & Whitney R-4360-20 first start on Vimeo

[Volume]

Careful cylinder offset was also necessary to combat vibration inertial forces.Limiting vibration in large double row radials requires masterful engineering with master rod offset, crankshaft design, vibration dampers, gear reduction and propeller selection etcSo what is the exact angular difference between the two parts of the crankshaft ? I would have always assumed 180° (as it is easiest forged that way in a simple die), but with the offset of half a cylinder pitch, this would of course mean that always two cylinders fire simultaniously, which increases torque pulses hence excites torsional oscilations. If the cylinders rows are aligned, there is always only one zylinder firing at a time.
So is the crankshaft offset different from 180° or the cylinder offset different from half a cylinder pitch ?

[barit1]

I believe the 4360 crank was in four pieces - one for each cylinder row. This would be necessary because the engine is assembled in a stack: the crankcase (with its 5 bearings) is in five segments, front to rear, with the crankshaft spline joints at the bearing journals #2, #3, and #4.

Thus the crankshaft throws could be manufactured in any required angular relationship.

[bvcu]

did a bit of work on the centaurus a few years ago in the sea fury. the arrangement of baffles and the finning were a work of art. always amazed me that the americans didnt adopt the sleeve valve engine .

[Graybeard]

from Wiki:

The Knight Engine was an internal combustion engine, designed by American (http://en.wikipedia.org/wiki/United_States) Charles Yale Knight (http://en.wikipedia.org/w/index.php?title=Charles_Yale_Knight&action=edit&redlink=1) (1868-1940), that used sleeve valves (http://en.wikipedia.org/wiki/Sleeve_valve) instead of the more common poppet valve (http://en.wikipedia.org/wiki/Poppet_valve) construction.

Born in Indiana in 1868, Knight was originally a printer and newspaper publisher, publishing a midwest farm journal called Dairy produce. To cover dairy activities during 1901-02, he bought an early Knox (http://en.wikipedia.org/wiki/Knox_Automobile) automobile, a three-wheeler with an air-cooled, single cylinder engine whose noisy valves annoyed him. He believed that he could design a better engine and proceeded to do so. Knight was familiar with the slide valves used on early Otto engines, having repaired the similar valve mechanism in his father's sawmill. The slide valve had however been replaced in gasoline engines by the poppet valve, whose characteristics were better suited to four-stroke engines.
At first Knight tried making the entire engine cylinder reciprocate to open and close the exhaust and inlet ports. Though he patented this arrangement, he soon abandoned it in favour of a double sliding sleeve principle. Backed by Chicago entrepreneur L.B. Kilbourne, an experimental engine was built in Oak Park, Illinois (http://en.wikipedia.org/wiki/Oak_Park,_Illinois) in 1903. Research and development continued until 1905, when a prototype passed stringent tests in Elyria, Ohio (http://en.wikipedia.org/wiki/Elyria,_Ohio). Having developed a practicable engine (at a cost of around $150,000) Knight and Kilbourne showed a complete "Silent Knight" touring car at the 1906 Chicago Auto Show. Fitted with a 4-cylinder, 40 hp (30 kW) engine, the car was priced at $3,500.
------
The last sleeve valve engine produced in the US was in 1932.

Yes, the Sea Fury sounds wonderful - like nothing else.

GB

[Jane-DoH]

bvcu

did a bit of work on the centaurus a few years ago in the sea fury. the arrangement of baffles and the finning were a work of art. always amazed me that the americans didnt adopt the sleeve valve engine .

Didn't sleeve-valves require a lot of lubrication?

[Jane-DoH]

barit1

I believe the 4360 crank was in four pieces - one for each cylinder row. This would be necessary because the engine is assembled in a stack: the crankcase (with its 5 bearings) is in five segments, front to rear, with the crankshaft spline joints at the bearing journals #2, #3, and #4.

Thus the crankshaft throws could be manufactured in any required angular relationship.

How many other multi-row (twin, three, four, etc) radials were assembled as a stack?

[esa-aardvark]

Hello, you will find a lot about P&W engines and others at...

enginehistory.org

[CV880]

I worked on two large twin row radials in the workshops early in my career; the 14 cylinder sleeve valved Bristol Hercules and 18 cylinder P & W R2800.
Both had 3 piece crankshafts with 180 degree offset big end journals. The only major differences were, the Hercules incorporated the big end journal into the centre crankshaft section whereas the R2800 had the big end bearing journals as part of the front and rear crankshaft sections. The method of bolting the sections together, the vibration damping and main bearing arrangement was different as well.

The R4360 had, as far as I can see, a solid crankshaft with split master rod big end bearings similar to a typical car engine. There are several pictures of a cutaway R4360 on the AEHS website where the split master rod big end bolts are visible.

[mike-wsm]

http://mjr.org.uk/pprune-047.jpg http://mjr.org.uk/pprune-048.jpg

http://mjr.org.uk/pprune-050.jpg http://mjr.org.uk/pprune-049.jpg

Bristol Centaurus


Bristol Industrial Museum used to have a collection of Bristol engines, including a glorious exploded Centaurus.

BIM closed in 2006 and is due to open under a new name soon, not sure if the engine collection is still there or whether at Kemble or Patchway.

Well worth a visit if you can track it down.