Many people ask me why most of the Subaru engines on this page are turbocharged, why not just use a larger naturally aspirated engine? There are 4 primary reasons:
1. Torque- Turbocharging is the most effective method to boost torque. Torque is what drives your prop. A naturally aspirated engine cannot hope to match the torque output of even a much smaller turbocharged engine.
2. Weight- With the addition of around 25 extra pounds for the turbo, intercooler and plumbing, the turbo system can deliver an easy and reliable 50 to 60% gain in torque and hp. This cannot be matched with a larger engine.
3. Altitude performance. Even if you had a much larger and heavier naturally aspirated engine, it would start losing power as soon as you left the ground. The turbo has far superior hot and high performance as well as superior climb and cruise power at altitude. Speed above 15,000 feet is improved considerably.
4. Reliability. Because the turbo engine develops more torque everywhere in the powerband, you can use much lower rpm to achieve the same power. This reduces the inertial stresses on the engine considerably. High rpm causes far more wear than the extra cylinder pressure of a low boost turbo system. It is important to remember that inertial stresses vary as the square of the rpm. Heat stress is of little concern with a properly designed water and oil cooling system. Stock camshafts can be used with low valve spring pressures, soft ramp rates and factory reliability. The high lift, rapid opening aftermarket camshafts used by certain Subaru converters are a proven reliability concern.
It is best to take some manufacturers claims with a grain of salt. Simple math can often tell you if their claims are BS.
Torque X RPM divided by 5252 equals hp. Hp X 5252 divided by RPM equals torque. Reduction ratio times torque equals prop torque.
One company claims 180 hp at 5900 rpm out of a naturally aspirated EJ22. This is highly unlikely as the engine would have to develop 160 ft/lbs of torque to achieve this figure. Even with extensive head work and a camshaft change, this is not going to happen at this high rpm. The peak torque might be 160, but not at 5900 rpm. The stock engine develops maximum torque at 4400 rpm and only develops 126 ft/lbs at power peak (5400 rpm). Typically, head work and performance cams will develop slightly more torque than stock at a somewhat higher rpm. 34 ft/lbs at 500 more rpm just won't happen in real life. Performance cams almost always have more valve overlap which usually has a negative impact on fuel flows. A well matched turbo system can and does recover some energy from the exhaust and reduces pumping losses on the intake stroke substantially to achieve better fuel flows.
Another company provides all the figures which you can work back and forth, but the prop torque does not come out right and their figures are just plain hard to swallow with an admitted 8% loss in the belt type redrive (their figures). Just be careful about what you get for all this money. You may be disappointed.
09/04/04
Lately Chev aluminum V8 engines are being offered to power various aircraft. One company claims their package including prop, all accessories and coolant is the same weight as an IO-360 Lycoming with C/S prop. I figure the Lycoming at about 455 lbs., theirs is 538. Claimed hp is 375 at 4400 rpm yet they use the identical IVO propeller that we use on our 180hp Subaru which is not enough to absorb the full hp of our engine at flight altitude and speeds. Granted we turn ours to 2200 rpm and they turn theirs to 2600 but calculated maximum absorbtion is on the order of less than 300 hp at this rpm. How can the engine put out 375? Torque at the prop is claimed to be 525 ft./lbs. yet with a 1.7 redrive ratio, this should be 760. Fuel consumption is claimed to be 15 gal./hr. at max power which represents a SFC of .24 lbs./hp/hr., better than virtually every diesel engine built today. How? Finally the TBO is set at 3500 hours yet the engine has never accumulated even 1/5 of this time. These values are all highly suspect even at a glance.
07/12/04
Yet another Subaru converter publishes what are said to be actual test results of hp, torque and fuel flow on the EJ22 and EJ25 engines. These figures have no basis in reality in my opinion. For instance if we take the figures for their EJ25 engine at 5200 rpm, they claim 189 hp, 396 ft./lbs. of torque at the prop flange with a 2.12 to 1 reduction ratio and a fuel burn of only 8.9US gallons/hr. How do they get a SFC of .282 lbs./hp/hr.? The volumetric efficiency comes out at 133%. Not bloody likely! Their EJ22 at the same rpm develops only 114hp (75 less hp with only 300cc less displacement!!!!) yet somehow, mysteriously, still develops 321 ft./lbs. at the prop flange with the same redrive ratio. How? It should have only about 242 ft./lbs. The EJ25 somehow develops 189 ft./lbs. at the crank at 5200 rpm, naturally aspirated from 2.5 liters. Pretty impressive. Boy, Subaru should fire their engineeers and hire these guys!!! Can you believe anything from a company which publishes information like this?
12/08/99
I was looking at many of the Subaru aircraft engine sites a couple of days ago. One company offers their engines with a mechanical fuel injection system, extolling the virtues of this over EFI and the fact that there is no computer or sensors to fail yet they use a microprocessor based ignition system with a failure prone magnetic crank sensor. This is simply illogical nonsense.
Another company advertises that their Subaru conversions will cut fuel flow in half compared to conventional aircraft engines. Again, this is utter nonsense. This implies that their Subaru conversions achieve specific fuel consumption figures of something on the order of .23 lbs./hp/hr. This is far lower than the best direct injection diesel engines available today which have 23 to 1 compression ratios and much higher thermal efficiencies. This is a thermodynamic fantasy on any Otto Cycle engine. How these figures are arrived at is a mystery and casts doubt on any other claims made. Subaru's own engineering data quotes a best BMSFC of .46lbs./hp/hr. for their EJ series engines so unless these guys can run their engines at 30 to 1 air/fuel ratios, this claim is downright false. They also claim 180hp at 5800 from their 2000cc engine. This represents a VE of 142%. Again, the factory Subaru engineers must all be idiots since they can't achieve figures like this.
12/23/99
I was looking at a site on Corvair engines for aircraft use recently. I found some more numbers that didn't add up. They claim 100 hp at 3200 rpm and 160 ft./lbs of torque at 2800. These figures were supposedly obtained on a dyno. Working these through our formulae, I get 164 ft./lbs at 3200 rpm. If the torque peak is at 2800, then the engine will likely be producing less torque at 3200. Since the engine makes 85 hp at 2800 rpm, it is probably only making around 90 hp at 3200. This "expert" also claims that the bottom end assemblies of Corvair engines are the same. Not true. 140, 150 and 180 hp engines featured heat treated, nitrided, 5140 alloy steel cranks with the lower hp engines having plain carbon steel cranks. The special cranks were developed for the higher stresses on the turbo engines. It is simply foolish to me that anyone would use a non-nitrided crank in an aircraft application- especially direct drive. Nitrided cranks are known to have 4-5 times the fatigue resistance of non-nitrided cranks. Exhaust valves on the turbos were also changed to a Nimonic alloy. The writer also seems confused about hp ratings. SAE net hp on the 140hp model engine was only 115 hp in reality at 5200 rpm. These engines from the '60s never put out their advertised gross hp. This fact is confirmed by acceleration tests by magazines of the era. The Corvair engine is not a bad design and has some good engineering and metallurgy behind it but they do have some mechanical problems such as dropped intake valve seats on 140s, distributor failures and ring land failures. These are usually caused by many miles or overtemping but should be checked.
You can roughly calculate the VE of an engine if hp is known with:
VE = 5600 X HP divided by (RPM X Displacement in cubic inches)
If we take the values above at torque peak, this engine is achieving a VE of 103.7% at 2800 rpm and 106.7% at 3200. This is virtually impossible on a Corvair engine which has relatively poor intake and exhaust flow as proven on my flow bench and the Corvair engine's poor specific output of only 0.70 hp/cubic inch. The stock EJ22 achieves nearly 1 hp/cubic inch by comparison. This is usually a good indication of the relative breathing efficiencies of various engines.
09/10/04
The biggest load of BS I see on most alternative engine sites is probably on fuel flows. Below are some representative SFC (specific fuel consumption) figures for various engine types. SFC describes how much fuel is burned in pounds per hour to produce 1 hp. One US gallon of Avgas weighs about 6 lbs. SFC is one measure on how efficient an engine is. The lower the number, the more efficient.
Automotive engines .40-.50
Small Lycoming/ Continental engines .42-.47
Large injected Lycoming/ Continental engines .37-.45
Turbo compound engines .34-.39
Light Diesel engines .27- .30
Industrial diesel engines .23-.27
The thermal efficiency of an engine is largely determined by its compression ratio and the type of fuel it burns. Diesels use about double to triple the CR of most gasoline engines and burn fuel that has a higher energy content per pound, therefore they have much higher thermal efficiencies than gasoline Otto cycle engines. It is IMPOSSIBLE for conventional Otto cycle engines available today to approach the SFCs that diesels offer. If you read of some gasoline engine making 200 hp that only burns 10 US gallons/hr., this is an SFC of .30 and complete BS. Best not to believe any other claims that this company makes about their engines.
For more reading:Direct Drive vs. Geared