3 row radials: pros cons
- Thread starter tomo pauk
- Start date
Ad: This forum contains affiliate links to products on Amazon and eBay. More information in Terms and rules
Readie
Chief Master Sergeant
lack of cooling for the third row.
You would be better off with an inline liquid cooled motor.
John
You would be better off with an inline liquid cooled motor.
John
Such as??????
Shortround6
Major General
With Readie do you really have to ask? 
As far as the three row engine goes it has a lot of disadvantages and few real advantages.
Engines can be looked at several ways. One is to try and build on engine of a certain power class. Like a 1000hp engine, or a 1500hp engine or a 2000hp engine.
Another way is to say you have developed a cylinder of a certain size/power, now how can you arrange multiples of that cylinder to get the power you want? This last is especial true of air-cooled cylinders because of the cooling problem.
Let's say you have a 70hp cylinder (in 1937) and you are making both a 630hp nine and a 980hp fourteen. Were do you go from there? A 1260hp eighteen or a 1470hp twenty-one?
The 21 will be longer and heavier than the 18 with it's longer crankshaft and crankcase. Cooling the rear row will be harder. The frontal area will be almost the same if not exactly the same. Arranging the valve gear for the rear row is going to be a headache. Many two row radials used a common cam ring at the front and routed the push rods fo rthe rear cylinders in-between the front cylinders. Some used a cam ring both in front and in back. The 3 row engine is going to require two cam rings at a minimum. Or, as was done on the Deerhound, lining up the cylinders and using overhead cams with 7 different cam drives.
A two row 18 does weigh more than a two row 14 but the 21 only has 16.7% more displacement/power than the 18. Is it enough to make up for the extra weight, cost and complication?
As far as the three row engine goes it has a lot of disadvantages and few real advantages.
Engines can be looked at several ways. One is to try and build on engine of a certain power class. Like a 1000hp engine, or a 1500hp engine or a 2000hp engine.
Another way is to say you have developed a cylinder of a certain size/power, now how can you arrange multiples of that cylinder to get the power you want? This last is especial true of air-cooled cylinders because of the cooling problem.
Let's say you have a 70hp cylinder (in 1937) and you are making both a 630hp nine and a 980hp fourteen. Were do you go from there? A 1260hp eighteen or a 1470hp twenty-one?
The 21 will be longer and heavier than the 18 with it's longer crankshaft and crankcase. Cooling the rear row will be harder. The frontal area will be almost the same if not exactly the same. Arranging the valve gear for the rear row is going to be a headache. Many two row radials used a common cam ring at the front and routed the push rods fo rthe rear cylinders in-between the front cylinders. Some used a cam ring both in front and in back. The 3 row engine is going to require two cam rings at a minimum. Or, as was done on the Deerhound, lining up the cylinders and using overhead cams with 7 different cam drives.
A two row 18 does weigh more than a two row 14 but the 21 only has 16.7% more displacement/power than the 18. Is it enough to make up for the extra weight, cost and complication?
wuzak
Captain
The Deerhound used vertical shaft drive to its overhead cams - like those employed in the Merlin, V-1710, etc. Not a great issue, I wouldn't have thought. Each of the cam drives was taken off a common gear ring - so in that sense is no different to the radial cam rings, in thet both needed to be driven by reduction gearing.
The Deerhound I was only 38l (2260ci), the II was increased in size to 41l (2510ci)and up to 1500hp. The redesigned III stayed at 41l and the same 44" in diameter as the II. The only on ewhich was built was tested at 1800hp. So, similar in capacity and performance to an R-2600 of the time, but much smaller diameter. It was only very early in the testing stages, however, so much work still had to be done.
Armstrong Siddeley engineers apparently originally wanted to make the engine liquid cooled but the MAP wanted competition for Bristol in air cooled engines.
The Deerhound was re-endineered to use reverse cooling early in the program. With hindsight, it may have been possible to provide the necessary cooling had they elected to use a system of baffles to guide the cooling air around the cylinders and heads - as was used in the R-4360.
The Deerhound I was only 38l (2260ci), the II was increased in size to 41l (2510ci)and up to 1500hp. The redesigned III stayed at 41l and the same 44" in diameter as the II. The only on ewhich was built was tested at 1800hp. So, similar in capacity and performance to an R-2600 of the time, but much smaller diameter. It was only very early in the testing stages, however, so much work still had to be done.
Armstrong Siddeley engineers apparently originally wanted to make the engine liquid cooled but the MAP wanted competition for Bristol in air cooled engines.
The Deerhound was re-endineered to use reverse cooling early in the program. With hindsight, it may have been possible to provide the necessary cooling had they elected to use a system of baffles to guide the cooling air around the cylinders and heads - as was used in the R-4360.
wuzak
Captain
Just in case I didn't make the point clear, for a given capacity (and presumably power level) the diameter of a 3 row radial could be smaller than for a 2 row radial.
The Deerhound II/III had a 5" (127mm) stroke as compared to a Wright R-2600's 6.312" (160.2mm), the Hercules' 6.5" stroke (165.1mm) and the P&W R-2800's 6" (152.4mm). So, it may have been possible later in development to use more rpm than those - as it was the III tested at 1800hp at 2600rpm.
Also, comapre the diameters:
Deerhound 44"
R-2600 55"
R-2800 52.8"
Hercules 55"
So, 80% of the diameter of the R-2600 and Hercules, meaning 64% of the frontal area. 83% of the diameter of teh R-2800, 69% of frontal area.
The Deerhound II/III had a 5" (127mm) stroke as compared to a Wright R-2600's 6.312" (160.2mm), the Hercules' 6.5" stroke (165.1mm) and the P&W R-2800's 6" (152.4mm). So, it may have been possible later in development to use more rpm than those - as it was the III tested at 1800hp at 2600rpm.
Also, comapre the diameters:
Deerhound 44"
R-2600 55"
R-2800 52.8"
Hercules 55"
So, 80% of the diameter of the R-2600 and Hercules, meaning 64% of the frontal area. 83% of the diameter of teh R-2800, 69% of frontal area.
Shortround6
Major General
You are quite right Wuzak. the 3 row radial will be smaller in diameter for the same power, that is about the only advantage.
The Deerhounds use of overhead cams was one way around the multiple rows of pushrod problems but introduced a few of it's own. A V engine uses two cam drives, a W engine uses 3 and the Deerhound used 7 cam drive units. While the parts are identical it is that many more parts that have to be made and fitted.
Overhead cam boxes also tend to limit some of the area that could be used for cooling fins. Perhaps not a big deal for an engine using 87 octane fuel but as the octane went up it becomes more important.
Does anybody have any information of the weight of the Deerhounds?
The Deerhounds use of overhead cams was one way around the multiple rows of pushrod problems but introduced a few of it's own. A V engine uses two cam drives, a W engine uses 3 and the Deerhound used 7 cam drive units. While the parts are identical it is that many more parts that have to be made and fitted.
Overhead cam boxes also tend to limit some of the area that could be used for cooling fins. Perhaps not a big deal for an engine using 87 octane fuel but as the octane went up it becomes more important.
Does anybody have any information of the weight of the Deerhounds?
wuzak
Captain
No, I can't find any weight data.
- Thread starter
- #9
Shortround6
Major General
The three row should be smaller than the Centaurus for the same power. Wither you can get it as small as the Sabre is a different question. Then you have the problem of the sleeve drives. easiest is to put the cylinders in line and run the shafts that control the sleeves between the rows, the inline configuration makes it harder to cool though, compared to staggered rows.
The question is if the 3 row is worth it. For a given power level it is smaller than a 2 row as Wuzak has shown. The engine will be longer, the power section will be at least 50% longer although the gear reduction section and the supercharger/accessories section will stay the same length. It will be heavier for the same power. It is more complicated and expensive to manufacture, It will be more difficult to develop. Look at the problems with mixture distribution in the R-3350. The R-4360 was no picnic either. Sneaking intake pipes forward to the front row can be done, it just takes a bit more work than a 2 row engine and the same for exhaust.
Is the extra work and weight worth the smaller diameter?
When designers reached a plateau in engine power per cylinder they fiddled a bit more with odd ball combinations to use more cylinders, Wright tried making a 22 cylinder two row radial but didn't get very far. an R-4090 or 4100 of pretty much the same diameter as the R-3350.
The question is if the 3 row is worth it. For a given power level it is smaller than a 2 row as Wuzak has shown. The engine will be longer, the power section will be at least 50% longer although the gear reduction section and the supercharger/accessories section will stay the same length. It will be heavier for the same power. It is more complicated and expensive to manufacture, It will be more difficult to develop. Look at the problems with mixture distribution in the R-3350. The R-4360 was no picnic either. Sneaking intake pipes forward to the front row can be done, it just takes a bit more work than a 2 row engine and the same for exhaust.
Is the extra work and weight worth the smaller diameter?
When designers reached a plateau in engine power per cylinder they fiddled a bit more with odd ball combinations to use more cylinders, Wright tried making a 22 cylinder two row radial but didn't get very far. an R-4090 or 4100 of pretty much the same diameter as the R-3350.
- Thread starter
- #11
I was looking at the in-line cylinders and poppet valves, ie. a layout of the Deerhound, but circa 3000 cu in displacement (for mid-war and after that). The 3-row is bound to be longer than a 2-row, but also shorter than an in-line; the complete powerplant section (from supercharger end to tip of spinner) shuld be somewhere in between.
About the extra work: the all mighty 'depends' is the keyword. If RAF (or LW) can have 2000-2500 HP engine of the layout size like I've proposed in 1943-44, than it compares rather favorably with what they had historically at the disposal at same time frame.
About the extra work: the all mighty 'depends' is the keyword. If RAF (or LW) can have 2000-2500 HP engine of the layout size like I've proposed in 1943-44, than it compares rather favorably with what they had historically at the disposal at same time frame.
wuzak
Captain
A 3 row radial will be more difficult to cool than a 2 row radial, just as the two row radial is harder to cool than the single row, and the 4 row harder than the 3 row. Not sure that the in-line configuration makes it any harder to cool than a staggered arrangement. Without the baffles the R-4360 wouldn't have worked at all, and I am sure the 3 row radial, be it staggered or lined up, could be made to work with similar arrangements.
I figure the Sabre could fit in a round cowling of approximately 48-50" in diameter, similar to that required by the Vulture. The Vulture installations, however, could be even more tightly packaged with bulges covering the front of the cam covers. Not sure if that could be done or was desirable for the Sabre.
Tomo never specified sleeve valves, but having the cylinders in line should simplify the sleeve drive arrangement.
For the same capacity I'm sure the power section will not be 50% longer, since the bore will be smaller, and each cylinder will have less heat rejection requirements and thus require smaller fins - so the bore centre distance could possibly be reduced.
Mixture distribution could very well prove to be problematic. It was for the engines you mentioned, as well as many others, including at least one V12 (the V-1710 had this problem before the "ram's horns" intae manifolds were developed).
Certainly it will be more expensive to build and develop, and probably heavier.
That's a question for aerodynamcists everywhere.
Certainly there is no benefit in using an engine as an alternative in some circumstances - like the Tempest, for example. It was deisgned for use with the Centaurus, so its fuselage was probably bigger than it needed to be for the Sabre, and using the smaller diameter Deerhound would probably give negligible aeordynamic benefit.
However, an aircraft designed around the more compact engine may gain more benefit from the drag reduction than the cost of the extra weight.
Perhaps a liquid cooled 3 row radial would be more viable? It would allow for single piece blocks, strengthening the whole unit. Cooling would be less of an issue, the engine will be shorter, though its installed weight would go up.
wuzak
Captain
That may have happened had the Deerhound been less troublesome or Armstrong-Siddeley hadn't been advised by the MAP to concentrate on axial flow gas turbines.
Isn't there cooling drag with a air cooled engine ? The more complicated the cooling airflow path is under the cowling, the more cooling drag there'll be.
While the smaller diameter of the engine will mean less drag, it would increase the cooling drag. You've got to get air thru there someway, a engine driven fan, like the Fw190, takes engine HP, or cowling flaps, increase drag.
Reducing the diameter would decrease drag, but it's not a free ride, it introduces it's own problems with drag also.
While the smaller diameter of the engine will mean less drag, it would increase the cooling drag. You've got to get air thru there someway, a engine driven fan, like the Fw190, takes engine HP, or cowling flaps, increase drag.
Reducing the diameter would decrease drag, but it's not a free ride, it introduces it's own problems with drag also.
- Thread starter
- #15
Indeed, we could use some data re. cooling drag.
Anyway, I'd really love to see a Spitfire with a 2000 HP in small diameter radial in 1943*, or, some early, yet not wide Tempest II/Sea fury.
*a plane somewhere in between Yak-3U and F2G (the Super Corsair)
Anyway, I'd really love to see a Spitfire with a 2000 HP in small diameter radial in 1943*, or, some early, yet not wide Tempest II/Sea fury.
*a plane somewhere in between Yak-3U and F2G (the Super Corsair)
wuzak
Captain
No, don't give the Spitfire a radial - it isn't right!
GregP
Major
Wuzak,
In the U.S.A., one of our best radials of WWII was the Pratt Whitney R-2800. Basically it is a winner. But it didn't start uout that way.
The crankshaft went through about 10 itterations before the higher-order harmonics were eventually eliminated or damped to where they were acceptable. After that, the R-2800 was pretty smooth and reliable. If you design a 4-row engine, you can basically run 2 two-row radials if you design the main bearings correctly to eliminate coupling between the two sets of two-rows ... so you have a good crankshaft design with little develoment time and effort. But designing a 3-row radial would entail another protracted crankshaft development since the radial will NEVER be in perfect balance. I bet the prospect of creating a 3-row crankshaft is what prevented the Americans from trying it. Wright was certainly aware of the difficuties.
In other countries, they stuck with one and two-row radials, so my bet is that everyone found the crankshaft design to be a bit daunting, especially without computers for the calculations.
They could probably make one now, but there is no market for it these days now that turbines are known to be so smooth and reliable.
A Spitfire with a radial would be interesting and somewhat shorter than the standard Spitfire.
In the U.S.A., one of our best radials of WWII was the Pratt Whitney R-2800. Basically it is a winner. But it didn't start uout that way.
The crankshaft went through about 10 itterations before the higher-order harmonics were eventually eliminated or damped to where they were acceptable. After that, the R-2800 was pretty smooth and reliable. If you design a 4-row engine, you can basically run 2 two-row radials if you design the main bearings correctly to eliminate coupling between the two sets of two-rows ... so you have a good crankshaft design with little develoment time and effort. But designing a 3-row radial would entail another protracted crankshaft development since the radial will NEVER be in perfect balance. I bet the prospect of creating a 3-row crankshaft is what prevented the Americans from trying it. Wright was certainly aware of the difficuties.
In other countries, they stuck with one and two-row radials, so my bet is that everyone found the crankshaft design to be a bit daunting, especially without computers for the calculations.
They could probably make one now, but there is no market for it these days now that turbines are known to be so smooth and reliable.
A Spitfire with a radial would be interesting and somewhat shorter than the standard Spitfire.
Last edited:
Readie
Chief Master Sergeant
An American radial in our wonderful and undefeated Spitfire?
There are some things that are simply not on old boy....
Your radials had there place to be sure but, if you really wanted brute power and streamlining then only a inline will do.
John
wuzak
Captain
Greg, if you design a 4 row radial with two independent carnks you need to find some way of transferring torque from both crankshafts to the propellor reduction gear and accesories.
The Wright R-2160 Tornado is an example of this - 6 rows of 7 cylinders, with 3 separate crankshafts and 7 lay shafts to transfer torque between the cranks and the propellor reduction gear.
A similar situation existed with the BMW 803. In that case the rear half of the engine drove the supercharger and accesories.
The Pratt Whitney R-4360 didn't have two separate twin throw crankshafts - it had a single crankshaft.
The R-2800 used, as did Wirght and Bristol radials, a built up crank. That is, the 2 throw crankshaft was built of 3 parts joined together. That allowed the use of a single piece master rod. For the R-4360 I think (going from memory) that they reverted to the one piece crank and two piece master rod.
In terms of balancing I think the AW Deerhound had some advnatge in that its banks were aligned, not shifted through a small angle (20° for 9/row, 25.7° for 7 per row - for a 2 row radial), and that the 3 throw crankshaft was probably better for balancing than the two throw crankshaft. In any case, teh Deerhound didn't seem to have problems with the crank - perhaps because they were trying to sort other issues first!
GregP
Major
John,
You might recall the current world speed record is held by ... a radial. Rare Bear and the mighty R-3350 went 528.31 mph in FAI supervised tests in 1989 and currently holds the world piston air speed record. No other radial or inline has come very close as yet. The fastest piston speed over a long range (1,000 km) is held by a Boeing B-29 at 410.43 mph in 1946. Please note both a powered by radials and both weer powered by the Wright R-3350. In WWII, they were right there but Spitfires, Fw 190's, Me 109's, etc. haven't been in the hunt for the world speed record for a LONG time.
Maybe the inlines aren't quite what you think they are? Although I DO like them, both engine varities have their places and uses. In the real world, Strega could probably take the world piston speed record right now, but the money is an obstacle. It probably costs about $350,000 or more to get a world speed record, and there is no payback.
After WWII, rather sensibly I might add, the various government ceased piston development. Private people in the U.S.A kept at it and set world speed records. It seems people in Europe are less interested in that arena, and I understand that sentiment and reasoning. The money will never come back to you. But the speed records have all been set by Mustangs, Bearcats in the last 20+ years. It doeasn;t look like a change either as I'm told Rare Bear may go for a new record (won from itself) sometime in the next couple of years. I suppose we'll see, won't we?
You might recall the current world speed record is held by ... a radial. Rare Bear and the mighty R-3350 went 528.31 mph in FAI supervised tests in 1989 and currently holds the world piston air speed record. No other radial or inline has come very close as yet. The fastest piston speed over a long range (1,000 km) is held by a Boeing B-29 at 410.43 mph in 1946. Please note both a powered by radials and both weer powered by the Wright R-3350. In WWII, they were right there but Spitfires, Fw 190's, Me 109's, etc. haven't been in the hunt for the world speed record for a LONG time.
Maybe the inlines aren't quite what you think they are? Although I DO like them, both engine varities have their places and uses. In the real world, Strega could probably take the world piston speed record right now, but the money is an obstacle. It probably costs about $350,000 or more to get a world speed record, and there is no payback.
After WWII, rather sensibly I might add, the various government ceased piston development. Private people in the U.S.A kept at it and set world speed records. It seems people in Europe are less interested in that arena, and I understand that sentiment and reasoning. The money will never come back to you. But the speed records have all been set by Mustangs, Bearcats in the last 20+ years. It doeasn;t look like a change either as I'm told Rare Bear may go for a new record (won from itself) sometime in the next couple of years. I suppose we'll see, won't we?