Large Radial Engines Were About As Good As Can Be?

[Shortround6]

We've established the reliable HP per cubic inch capacity of modern diesels is in the area of 1 hp/cu.

Can we take modern diesel tech and apply it to an aircraft radial engine to achieve the desired hp to weight ratio?

Two different things isn't it?

What is overlooked or glossed over is the peak cylinder pressures in a diesel are higher than in a gasoline engine so the engine has to be built stronger. Stronger engine means more weight.

BTW the Junkers Jumo 205-207 series that both the British Chieftain engine and the Ukrainian engine are based on seems to have topped out at 1000hp for 1014 cu in.
However it was a two stroke engine, opposed pistons and the 1000hp version used a two stage supercharger, one engine driven and one turbo.
It also weighed 1430lbs. and was liquid cooled, just like the tank engines.

A big problem comparing the aircraft and tank engines is that the tank engine weights often include all kinds of stuff (like generators) that aircraft engines do not include.

The Caterpillar RD-1820 is listed at weighing 3900lbs but that seems to include a step up gear to the drive shaft, plus the oil cooler and engine fan/drive mechanism.
I would also note that the diesel version is down rated by about 300rpm from the gasoline engine version (at least the diesel manual makes reference to a -100 series engine for certain procedures) but this was common for aircraft engines used in tanks.

However a problem with using tank or marine engine specifications as a point of comparison for aircraft engines is that they are sea level engines,
The long nose Allison in the early P-40s was rated at 1040hp at altitude. However the engine/supercharger combination was capable of making around 1700hp at sea level, the engine just wouldn't give you that kind of power for very long without breaking although a few pilots reported using it for 16-20 minutes.
Trying ot make sea level power at 20-22,000ft means the air intake system has to handle twice the cubic feet of air as it does at sea level.

 

[m37b1]

"Two different things isn't it"?

• Yup - But, without establishing the first, the conversation was moot. A 600hp RD-1820 would be a non-starter. However, a 1400hp RD-1820 would be something worth investigating.

"What is overlooked or glossed over is the peak cylinder pressures in a diesel are higher than in a gasoline engine so the engine has to be built stronger. Stronger engine means more weight".

• Cylinder pressure issues with diesels has been largely mitigated with staged injection events. That's one of the reasons we can triple the hp of the Cummins B/ISB series with little increase in mass.

"BTW the Junkers Jumo 205-207 series that both the British Chieftain engine and the Ukrainian engine are based on seems to have topped out at 1000hp for 1014 cu in.
However it was a two stroke engine, opposed pistons and the 1000hp version used a two stage supercharger, one engine driven and one turbo.
It also weighed 1430lbs. and was liquid cooled, just like the tank engines".

• I've never worked on the Chieftain engine, so I can't speak to it. However, I've done a lot of work with Ukraine on the 6TD series. That engine is now at 1500 hp in -3 form and 1800hp in -4. My work with them involved retrofit of T-72's with the 6TD-2 and finding other markets for production. My recommendation was marine use which was embraced. However they had serious QC issues due to antiquated manufacturing methods. The factory looked like something out of 1942, producing Sherman tanks. We redesigned the manufacturing process of the engines, employing modern CNC methods. Result: 90% reduction in scrap and reduction in failures in end use. This all in time for the Russian invasion. So, no marine engines, as they're using every engine they make for T-84 manufacture and T-80 rebuild/re-power.
• If no turboprops existed, this is the route I would take with an aviation diesel.

"A big problem comparing the aircraft and tank engines is that the tank engine weights often include all kinds of stuff (like generators) that aircraft engines do not include".

• Agreed

"The Caterpillar RD-1820 is listed at weighing 3900lbs but that seems to include a step up gear to the drive shaft, plus the oil cooler and engine fan/drive mechanism.
I would also note that the diesel version is down rated by about 300rpm from the gasoline engine version (at least the diesel manual makes reference to a -100 series engine for certain procedures) but this was common for aircraft engines used in tanks".

• Agreed

"However a problem with using tank or marine engine specifications as a point of comparison for aircraft engines is that they are sea level engines,
The long nose Allison in the early P-40s was rated at 1040hp at altitude. However the engine/supercharger combination was capable of making around 1700hp at sea level, the engine just wouldn't give you that kind of power for very long without breaking although a few pilots reported using it for 16-20 minutes.
Trying ot make sea level power at 20-22,000ft means the air intake system has to handle twice the cubic feet of air as it does at sea level".

• Agreed, maintaining power at altitude is more than just adding boost to the intake charge. The beauty of diesels is that they eliminate the issues resulting from the fine balance between: Boost, Mixture, Ignition, that altitude impacts. Basically, the diesel takes a full intake charge on every stroke, meaning all you need is a waste-gate to maintain constant intake pressure. Injection timing and charge is all that controls power output.
• Not saying this idea isn't without hurdles. The Russians tried with the Charmskiy M-40 during WWII and gave up. But they didn't have the tech we have today.
• If I ever gain my investment values back, maybe I'll try to find an RD-1820 and give it a shot.

 

[swampyankee]

Successful -- or at least moderately successful -- diesel aircraft engines existed before WW2, with the best and most successful being those of Junkers, but Packard and Guiberson in the US also made serviceable, albeit not commercially successful, aircraft diesels. Post-WW2, there were attempts in the USSR, UK (Napier Nomad, which was a ported, valveless, 2-stroke), and, in the US, McCulloch (TRAD-4180). Currently, there are production aviation diesels, many of which seem to be targeted to the UAV market.

 

[Admiral Beez]

There are still radial engines produced today, such as the Vedeneyev M14P. I wonder if these have benefited from 2000's tech vs. 1940s tech.

 

[pinehilljoe]

I think material and manufacturing technology improvements for aircraft IC engines reached diminishing returns. In the end it comes down to the volume of air into the engine. The amount of air sucked into a piston engine is limited by, among other factors, the size of the cylinders, the intake manifold or carburetor. In a gas turbine there is almost no limit: a gigantic fan at the opening of a gas turbine can swallow vastly more air than can be taken into a piston engine, - as a rule of thumb seventy times as much, Whittle's invention was truly a paradigm shift. * paraphrasing Simon Winchester from his Book Precision.

 

[GregP]

Not too sure if there are any large diesel radials in production, but the smaller Zoche series are in production. They have an even number of cylinders. The 4-cylinder ZO 01A has a displacement of 162.6 cubic inches (2660 cc) and makes 150 HP, for .92 HP/su in. The 8-cylinder ZO 02A is 325.3 (5330 cc) cubic inches and makes 300 HP for 0.92 HP/cu in. So, these small diesel radials are right there in the ballpark of 1 HP per cubic inch.

Modern gasoline radials aren't even close. The Rotec 9-cylinder radial is 220 cubic inches and makes 150 HP for 0.68 HP/cubic inch. The Vedeneyev M-14-P has 9 cylinders, 621 cubic inches, and makes 400 HP for .644 HP/cubic inch. These are very reliable and desirable engines for aerobatic aircraft. They now come with either air-start or electric start as an option. One reason they aren't as close to the magic 1 hp/cubic inch is we are limited to about 100-Octane LL fuel today. There isn't any 145 / 150 PN leaded fuel around except in special batches sometimes made for the Reno air races and whatever brews are made up in private hangars.

 

[swampyankee]

For aircraft engines, I would argue that hp/in^3 isn't a really important metric; hp/lb would be far more important. Current technology permits production of aircraft diesels with power/weight ratios competitive with spark-ignition engines and, at least potentially, lower operating costs, although many don't make it because of quite high maintenance costs.

Had large turbines not panned out, big diesels may have taken over from spark-ignition engines as there are ignition issues that start being more pronounced as engine bores get larger.

 

[GregP]

Personally, I like Hp / pound, too.

 

[Simon Thomas]

The SI bores are pretty much on the limit now. CFD and thermal FEA may eek out another 1/4" or so, although only having 100LL may not even allow that.
Wright was ok with 6 1/8", which I credit to Sam Heron being fastidious about cooling the 1750/1820. When P&W tried to go above 6" it was not a success - they went back to 5 3/4" for most engines after the Hornet. Bristol could have gone beyond their 5 3/4" with the sleeve valve, but not without Fedden in charge.

 

[Koopernic]

Two different things isn't it?

What is overlooked or glossed over is the peak cylinder pressures in a diesel are higher than in a gasoline engine so the engine has to be built stronger. Stronger engine means more weight.

BTW the Junkers Jumo 205-207 series that both the British Chieftain engine and the Ukrainian engine are based on seems to have topped out at 1000hp for 1014 cu in.
However it was a two stroke engine, opposed pistons and the 1000hp version used a two stage supercharger, one engine driven and one turbo.
It also weighed 1430lbs. and was liquid cooled, just like the tank engines.

A big problem comparing the aircraft and tank engines is that the tank engine weights often include all kinds of stuff (like generators) that aircraft engines do not include.

The Caterpillar RD-1820 is listed at weighing 3900lbs but that seems to include a step up gear to the drive shaft, plus the oil cooler and engine fan/drive mechanism.
I would also note that the diesel version is down rated by about 300rpm from the gasoline engine version (at least the diesel manual makes reference to a -100 series engine for certain procedures) but this was common for aircraft engines used in tanks.

However a problem with using tank or marine engine specifications as a point of comparison for aircraft engines is that they are sea level engines,
The long nose Allison in the early P-40s was rated at 1040hp at altitude. However the engine/supercharger combination was capable of making around 1700hp at sea level, the engine just wouldn't give you that kind of power for very long without breaking although a few pilots reported using it for 16-20 minutes.
Trying ot make sea level power at 20-22,000ft means the air intake system has to handle twice the cubic feet of air as it does at sea level.

Junkers Jumo 224 Aircraft Engine
"The Jumo 224 used the same bore and stroke as the Jumo 207C. While the Jumo 224 was being designed, a Jumo 207C was tested to its limits to better understand exactly what output could be expected from the Jumo 224. Tests conducted in late 1944 found that with a 200 rpm overspeed (3,200 rpm), intercooling, modified fuel injectors, and 80% methanol-water injection, the Jumo 207C was capable of a 10 minute output at 2,210 hp (1,645 kW)—twice its standard rating of 1,100 hp (820 kW)."

The development of better alloys has overcome much of the power to weight ratio issue that diesels had compared to petrol and supercharging has increased power levels greatly. The Leopard 2 powerplant is a hyperbaric diesel with a manifold pressure of 4.5 atmospheres.


The Modern version of this engine again has a supercharger combined with the turbosupercharger. Two stroke diesels need a supercharger to scavenge and therefore to start. In the modern version a infinitely variable drive is used which allows the manipulation of the mixture:

Some interesting information comes out of this interview:

(take a moment to warm up)

Please note the "German Bomber" is the Dornier Do 26K Seefalke (sea falken). It is not a bomber, merely a mail plane. With a range of 9000km/5400 miles it could certainly do the mission. The mission was a humanitarian one, not bombing.

"One notable Do 26 civilian mission was carried out by V2 Seefalke, when on 14 February 1939 the veteran Lufthansa pilot Flight Captain Siegfried Graf Schack von Wittenau embarked on a mercy flight to Chile, taking 580 kg (1,279 lb) of medical supplies for earthquake victims in Chile. The 10,700 km (6,600 mi) flight between Lisbon and Rio de Janeiro lasted 36 hours."

 

[xylstra]

Suppose I gave you year 2020 design technology and a large pile of money for the design process. But when it comes time to build the engines, you are limited to what's available in 1944. So no electronic ignition, no computer controls, only things that could be built in 1944. Would you be able to produce a large radial engine that was significantly better than what they had in 1944?

Think you can make one that's 10% lighter, 10% more fuel efficient, and 50% more reliable?

P.S. Yes, I know the answer is the gas turbine, but I'm not asking about gas turbines. I'm asking about how close-to-optimal large radials from 1944 were.

Nonsense! Bosch (Germany) did have electronic CDI ignition. Test-flown in a fighter at Rechlin. Collaborative development with JUNKERS.

 

[tommayer]

Not too sure if there are any large diesel radials in production, but the smaller Zoche series are in production. They have an even number of cylinders. The 4-cylinder ZO 01A has a displacement of 162.6 cubic inches (2660 cc) and makes 150 HP, for .92 HP/su in. The 8-cylinder ZO 02A is 325.3 (5330 cc) cubic inches and makes 300 HP for 0.92 HP/cu in. So, these small diesel radials are right there in the ballpark of 1 HP per cubic inch.

Modern gasoline radials aren't even close. The Rotec 9-cylinder radial is 220 cubic inches and makes 150 HP for 0.68 HP/cubic inch. The Vedeneyev M-14-P has 9 cylinders, 621 cubic inches, and makes 400 HP for .644 HP/cubic inch. These are very reliable and desirable engines for aerobatic aircraft. They now come with either air-start or electric start as an option. One reason they aren't as close to the magic 1 hp/cubic inch is we are limited to about 100-Octane LL fuel today. There isn't any 145 / 150 PN leaded fuel around except in special batches sometimes made for the Reno air races and whatever brews are made up in private hangars.

Are Zorches actually in production? Not according to their web site. Do you know of any that have flown? Is either the 4 or 8 cylinder engine certificated in U.S. or Europe? The specs for the 8 sound almost too good to be true--271 lbs including accessories and super charger and turbo charger. Please tell me where to look for info that goes beyond the Zorche web site that comes up on Google when you put in Zorche aircraft engine. Thanks in advance.

 

[davparlr]

Two different things isn't it?

What is overlooked or glossed over is the peak cylinder pressures in a diesel are higher than in a gasoline engine so the engine has to be built stronger. Stronger engine means more weight.

BTW the Junkers Jumo 205-207 series that both the British Chieftain engine and the Ukrainian engine are based on seems to have topped out at 1000hp for 1014 cu in.
However it was a two stroke engine, opposed pistons and the 1000hp version used a two stage supercharger, one engine driven and one turbo.
It also weighed 1430lbs. and was liquid cooled, just like the tank engines.

A big problem comparing the aircraft and tank engines is that the tank engine weights often include all kinds of stuff (like generators) that aircraft engines do not include.

The Caterpillar RD-1820 is listed at weighing 3900lbs but that seems to include a step up gear to the drive shaft, plus the oil cooler and engine fan/drive mechanism.
I would also note that the diesel version is down rated by about 300rpm from the gasoline engine version (at least the diesel manual makes reference to a -100 series engine for certain procedures) but this was common for aircraft engines used in tanks.

However a problem with using tank or marine engine specifications as a point of comparison for aircraft engines is that they are sea level engines,
The long nose Allison in the early P-40s was rated at 1040hp at altitude. However the engine/supercharger combination was capable of making around 1700hp at sea level, the engine just wouldn't give you that kind of power for very long without breaking although a few pilots reported using it for 16-20 minutes.
Trying ot make sea level power at 20-22,000ft means the air intake system has to handle twice the cubic feet of air as it does at sea level.

Remember when Oldsmobile tried to make a diesel engine out of their 350 gas engine? They strengthened the block but apparently forgot to strengthen the head bolt design. It didn't work out too well.

 

[Builder 2010]

You can't fool mother nature. When you try to do something on the cheap, it usually works out poorly. It was one of many blunders GM made in their car design world. The GM Saturn comes to mind. Great concept, nice cars until the corporate guys got their hands on it.

 

[Conslaw]

The original Saturn S-series was competitive when it first came out. When Honda redesigned the Civic for 1993 model year, the Saturn was a generation behind.

 

[swampyankee]

Are Zorches actually in production? Not according to their web site. Do you know of any that have flown? Is either the 4 or 8 cylinder engine certificated in U.S. or Europe? The specs for the 8 sound almost too good to be true--271 lbs including accessories and super charger and turbo charger. Please tell me where to look for info that goes beyond the Zorche web site that comes up on Google when you put in Zorche aircraft engine. Thanks in advance.

Zoche seems to be currently in the state of "vaporware." There are diesel aircraft engines currently in production; these are sold by Continental, Safran, Austro Engine, and perhaps one or two other makers.

 

[D3801]

Don't forget, the WW II engines were rather underrated. To qualify for wep rating the engine had to run 7 1/2 hours at that rating PLUS the rest of the 150 hour type test. Car engines seldom run for more than a minute at a time at full power. Even at Le Mans with that long straightaway.
Type test might be equal to 5-6 24 hour endurance races without ever replacing a major part in the engine.
Modern knowledge will improve things. But those old engines had another design criteria. That was power to weight. A lot of the high power engines were running more than 1hp per pound of engine weight.
What are the weights of those Ford car engines?
The other thing is altitude. Most of those engines were running at part throttle at sea level or take-off.
How much power do the car engines make at 15,000ft? Or pick critical altitude for a specific aircraft engine.
See how much power some of those engines made when set up for hydroplane racing.
Again, modern knowledge will improve things but those engines were pretty high tech not only in their day but compared to race car engines 10-20 years newer. Granted it wasn't supercharged but the engine in the 917 Porsche was the first Porsche engine to make more than 1hp per pound.

 

[RW Mk. III]

I think perhaps there is a bit of underestimation of what gains 2020 computer modelling would yield against millions of man hours in the 1940's, particularly as it applies to optimizing superchargers and turbochargers. BUT I am not really qualified in that area so many grains of salt. I'll ask my dad next time I talk with him he may have something to contribute, I'll report if there is.

Thinking a hair laterally though...

I have two thoughts to the OP's question:

1. We can DESIGN on 2020 computers and build with 1940's abilities and materials? Okay that's grand. We want to improve performance? Forget about power to weight of the engine assembly for a moment. An aero engine by itself isn't very useful. Let's look at how it's wed to the airframe. With 2020 computer design, 2020 reference materials/textbooks/engineers we can install an aircooled engine better. Imagine if Kurt Tank could have designed his originally intended BMW 801 installation on the FW190 with those resources. Extremely complex aerodynamic modelling including all conceivable variables can be plotted now. I cannot believe the drag numbers on an aircooled radial engine couldn't be appreciably lowered. Pneumatic/hydraulic/electric actuated aerodynamic elements can be introduced to govern and direct airflow productively. And because we had the computers at the front end, we can generate the programming for a --mechanical-- computer to control it all in a 1940's application.

2. Continuing that thought. Engine management, we can't have computers right. Well within the rules of this scenario we can at least somewhat end-run that restriction; we can definitely have mechanical computers. Those very much exist in the 1940's and we can design and program THOSE using 2020 computers and technology and understanding. That's huge. The kommandogerat is a tinkertoy compared to the mechanical control systems you can rustle up now.

 

[swampyankee]

RW, (Large Radial Engines Were About As Good As Can Be?), No. 1 is probably the most applicable. Largely because of the requirements of emissions control, there has been a great deal of research on combustion and flows within engine. A program like Kiva and the computers needed to run it (preferably a Beowulf cluster) would let an engineering group to replicate something like the entire US Army Hyper program in about a month. Something like ANSYS or NASTRAN would permit accurate stress calculations throughout the engine, CFD would permit modeling of the airflows through the engine.

Let's give this to the design staffs in Bristol, Grumman, Lockheed, and Hawker. Defeating the Germans quicker in WW2 would be vastly better for humanity.

 

[Shortround6]

1. We can DESIGN on 2020 computers and build with 1940's abilities and materials? Okay that's grand. We want to improve performance? Forget about power to weight of the engine assembly for a moment. An aero engine by itself isn't very useful. Let's look at how it's wed to the airframe. With 2020 computer design, 2020 reference materials/textbooks/engineers we can install an aircooled engine better. Imagine if Kurt Tank could have designed his originally intended BMW 801 installation on the FW190 with those resources

The thing is the state of the art didn't stay stationary during the period from 1940 to 1945. Not to say that things can't be improved but the engineers of 1945 would have laughed at Kurt Tanks installation of the 801 if it was suggested it was state of the art.
The P & W R-2800 C series need 10% less airflow to cool at 2000hp than the R-2800B series did due to more and deeper fins and better baffling. Which means close to 10% less drag. P & W built the 4 row R-4360 using essentially R-2800 C cylinders. They did require careful handling (and forced the development of electric engine analyzers) but getting a four row radial to work was by no means easy. We could certainly do things faster and easier today. P & W used 23 engines to get to 15,000 hours of ground running on the R-4360 while the original R-2800 only had 3500hrs before it passed a type test. Computer design and simulation could sure make the design and development go faster and cheaper.
By the late 40s the R-4360 was rated at 3500hp wet and 3250hp dry for take-off from an engine that weighed 3520lbs and it could make 2650hp at 6,000ft for pretty much as long as the fuel lasted if any one cared to run it that way. (pay for the fuel)
The R-2800 in the F4U-5 was rated at 1800hp Military and 1500hp max continuous, both at 30,000ft with it's rather complicated supercharger system.
Somehow, without the aid of either mechanical or electronic computers they managed to keep the engine near the proper temperatures at both high and low altitudes and under different operating conditions. Yes we could do better now but were they really off by that much?
P & W did have trouble with the R-4360 sagging or bending when mounted like a two row radial (engine mounts at rear of crankcase. but that was solved by moving the engine mounts out until they nearly reached the center of gravity of the engine.

Kurt Tank did help show the way, but there was a constantly evolving area of knowledge about cowling internal air flow and exhaust thrust in addition to just building engines.

What is truly amazing is not that we can do better now, but that they were building hundreds or thousands of engines per month that were equivalent in HP to weight and fit and finish to the race car engine of the 1960s which were never built in anywhere near those quantities ( 25 engines a year might be high production for a race car engine)