Engine efficiency
 

What makes one engine more efficient then the next?

First, lets define efficiency. For my purposes I define efficiency as making the most miles per gallon of fuel.

Given that definition, Diesel will win, always. Why? First, because there is more energy in a gallon of diesel then a gallon of gasoline. 10 to 20 percent more depending on how you define diesel, gasoline, and the ethanol content.

Second, even given the exact same amount of BTU's of fuel, a diesel is inherently more efficient then a Gasoline engine. Diesels do not have pumping losses (no throttle plate restriction). And diesels have PLENTY of air to burn all the diesel making them slightly more efficient in burning.

Diesel engines are usually heavier then gasoline engines, and we all agree less weight makes a more efficient vehicle.

From there, picking on a gasoline engine, some engines are more efficient then others.

Pumping losses are a function of size of engine (displacement), RPM, and throttle position.

The bigger the engine, the more pumping losses, and to some extent, the more friction.

The more RPM, the more losses due to pumping and friction.

Throttle position closed means higher vacuum, which means more pumping losses.

So, in a perfect world, at cruising speed, we would have a small engine running at low RPM, at WOT. Again, the diesel wins this war over the gasoline engine as they have no throttle.

Which is more efficient - a 500 cc engine running at 4000 rpm or a 2 liter engine running at 1000 rpm)? If both vehicles have the same intake vacuum (measure of throttle and engine efficiency), then I would bet on the 2 liter.

I believe the trick for a gasoline engine is to have the manifold vacuum to be as minimal as possible, and the engine displacement times the RPM to be as small as possible.

You've missed out expansion ratio - a major reason why diesels are more efficient than gasoline engines.

how does expansion ratio relate to anything????

I agree. I think a lot of this has to do with stroke and RPM. If you look at any of the super-efficient industrial diesels out there, they run at extremely low RPM and have an insane stroke. They are getting absolutely every last ounce of expansion energy out of the gas before the exhaust stroke begins. Running a short stroke and/or high RPM will allow you to make more power, but at the expense of efficiency. The other downside of the low-RPM long-stroke engine is that its specific output (HP/LB) is extremely low.

Here's an article on the most powerful engine in the world... over 50% thermodynamic efficiency!

Most powerful diesel engine in the world

compression ratio and expansion ratio are pretty much unrelated other then they have the word "ratio" in both.

Unless I am missing something? Which is why I posted this thread in the first place.

A higher compression ratio results in an increase of thermal efficiency.

Don't be fooled about "engine speed" by only looking at crankshaft rpms:

Most powerful diesel in the world: 98" stroke at 102 rpm = 1,666 ft/mn piston speed

Honda 50cc Super Cub at max fuel efficiency speed: 1.63" stroke at 4300 rpm = 1,168 ft/mn piston speed.

Hmmmm.

Yes one is a supercharged two-stroke diesel and one is a four stroke gasser. Different engine types seem to have different piston speed ranges for max fuel efficiency.

You can easily drop off the bottom end of the engine's efficiency zone by doing that.

Language issue. Apparently in some places Expansion Ratio is used instead of Compression Ratio.

I have always thought of Expansion Ratio as the ratio things expand as they go from Liquid to Gas - like Water to Steam.

Wrong. Ex: Atkinson cycle - Wikipedia, the free encyclopedia

FWIW, that is AVERAGE piston speed. If you look at the PEAK piston speeds, you might find some interesting results.

But to do that, you must do a bit of math and know the Rod length and stroke.

I think as concerning fuel efficiency, we want to look at average piston speeds.

If we are getting into the nuances of intake and exhaust tuning, then peak piston speeds get more relevant.

For an Otto-cycle engine they are interchangeable, for Atkinson, ER is often significantly higher than CR.

Edit: Damn you Frank!

Is there a more accurate predictor than piston speed? Assuming relatively constant VE/RPM. I don't quite understand why piston speed alone is a factor, but I could understand that there is an optimal rate of expansion for the combustion gasses, and that the combustion chamber should not overly retard, or outrun the expanding gasses, or mimize the discrepancy anyway. So that it would be a function of bore*stroke or something.

^Yeah.

I think it has to do with how much combustion heat energy pushes on the piston vs dissipating into the cooling system.

Too slow rpm = more time for heat to go into coolant
Too fast rpm = less push on piston; more internal engine friction

The larger bore sizes have lower cylinder wall surface area as a percentage of displacement, so they will also have higher efficiency. Optimal piston speed is, as Frank already posted, the ideal point between the excess losses of the two extremes.

Other factors are the harmonics of induction and exhaust pulses and how they can reduce pressure waves which would affect in cylinder compression and exhaust scavenging. Also the amount of residual exhaust remaining in the combustion chamber which would tend to increase efficiency to a point, one reason why highly scavenged exhausts usually produce lower mileage.

Combining gas and diesel engines could yield best of both worlds

regards
Mech

I've read that a long stroke relative to rod length increases stress on certain engine components. Does this mean that it increases friction?

Are we talking rod/stroke ratio, or absolute stroke length? Can you have a good rod/stroke ratio and still approach an "ideal" stroke length simply by building a bigger engine?

Also, why does increased stroke length increase efficiency anyway?

longer rod = more better. more efficient, more power, more everything, but slower maximum piston speed, which again, is better.

Sigh.

Expansion ratio - Wikipedia, the free encyclopedia

if we want to get really picky, the Otto by definition has different compression then what you are calling expansion because of exhaust valve timing differences then intake.

So Frank. How does this affect efficiency for mileage? (hint: variable cam timing).

Always a trade off in reciprocating engines. Longer rods and longer stroke generally give you more torque at lower RPM, but also increase reciprocating mass which will reduce efficiency.

I think the 225 Chrysler slant 6 had a 3.5 inch bore and a 4.5 inch stroke.

regards
Mech

After some reading, I'm guessing the reason the Slant 6 was a... well, Slant 6, was so they could make it extremely undersquare to develop a lot of torque at low RPM, efficiently, but also have long enough rods that the rod:stroke ratio wasn't too poor (and thus causing excess friction, and bending rods all the time), yet still fit under the hood?

A longer rod also reduces the side forces of the piston so there's less friction between the rings and the cylinder walls.

yeah, there is kind of two conversations going on at once here.

there is bore to stroke ratio. perception has always been longer stroke means more power at lower rpm, bigger bore means you can fit bigger valves for higher RPM horsepower.

Also, there is stroke to rod length. after doing a bunch of reading, smokey yunich used to believe longer rods are "free" power, but more modern theory is there is a perfect ratio. too long of a rod can lead to issues also, but too short of a rod really sucks.

it seems to me retarding the exhaust timing leads to better bottom end power and better gas mileage.

That's not entirely true. You are correct that diesels typically don't have throttling losses, however, there still are pumping losses. This includes all the losses associated with getting the air & exhaust through the manifolds & ports. In addition, in a turbodiesel, there are turbo inefficienies that are considered pumping losses as well. In some cases, however, if the turbo is being run in an efficient manner (in the sweet spot of the turbo maps), there are some speeds and loads where the tubo actually results in "positive" pumping work (negative pumping losses). The intake manifold pressure can actually be higher than the exhaust manifold pressure. I imagine this might also be possible on a gasoline turbocharged engine, but that's not really my area.

interesting.
i did not know you could have higher intake pressures then exhaust pressures. thinking about it, theory says it sure is possible. though.

thank you.

One of my favorites, no reciprocation, no masses changing direction, kind of like a piston in cylinder turbine.

Animated Engines - Gnome Rotary

A true rotary engine. Even a centrifugal supercharger.

regards
Mech

there is another reason a long rod helps. (insert obligatory tasteless joke here) it decreases the side load on the piston when the crank is at 3 and 9 o'clock. These side loads play hell on the piston meaning it has to be bigger with short stroke engines.

so, when power to weight ratio is concern number 1 as with a MC, sports car or aircraft, light compact short strokers are the way to go, but, they will be less efficient.

stationary engines should have very over square engines as weight is of no concern.

high efficiency vehicles will likely be something somewhere in the middle, probably a little oversquare, ideally.

looks kind of like a hydraulic motor / pump.

there still is some reciprocatin' going on as best I can tell. it sure is a cool engine to watch run though. radial engines really are works of art.

how do they fall in the efficiency area? is there a reason they are no longer made other than prohibitive production costs?

you confusing rod length and stroke. different things.

Efficiency was not their primary objective. Cooling and power to weight for performance were the priorities. 160 HP out of 8 liters at 1300 RPM and 325 pounds weight, in a era when compression ratios were 5 to 1 and fuel was probably 70 octane.

Vicious handling characteristics when you have 300 pounds spinning around in fron of an airframe that was basically a glorified kite. Compared to the same year Mercedes grand prix engine at twice the displacement and 2300 RPM producing 200 HP and you get an idea of the power in the context of the era when they were made and used.

As a fighter plane engine they were practically invincible, using engine torque to turn twice a quickly in one direction as the other.

With rods spinning around the fixed crank journal and combustion pressure actually pushing the cylinder head away from the fixed journal, there is very little reciprocation except amybe in the valve train. In the Gnome the exhaust valve was actually 2/3 of what you would normally call the cylinder head, and the valve spring was so weak you can actually open the exhaust valve with your finger. It depended on centrifugal force for "spring tension".

http://www.youtube.com/watch?v=-UBAukXPD-0

regards
Mech

Why leave out electric motors? They are 2-3X more efficient than internal combustion engines. It's not even close.

start your own thread. feel free.

Frank's point is that when talking about engine efficiency, expansion ratio and compression ratio are two different concepts, both of which affect efficiency. (Expansion ratio as explained in the Wiki article you provided the link to is a different, largely unrelated, concept.) The Atkinson cycle engine is the classic example of an engine that is not symmetric in compression and expansion ratios.

Ordinarily, the efficiency quoted for a given engine is based on energy out / energy in (which is derived from BSFC). Miles per gallon is usually not used as a measure of engine efficiency because it has as much or more to do with the car (and car/engine fit) as it does with the engine.

So, we have to make some assumptions. You mention "both vehicles", suggesting that the vehicles might be different. In that case there is no way to answer the question. Let's assume that it's the same vehicle with an engine swap, and that we have ballasted the car to the same weight in both cases.

We'll also assume that the engines are of the same general specific output class: say 60 bhp per liter.

In that case the 500 cc engine will be at half throttle and the 2000 cc engine will be at 1/8 throttle. RPM will not have a great effect on BSFC

For simplicity, we'll assume a single test speed, at which 11 hp is required (a small car cruising at 60 mph).

The two engines could be covered by the same BSFC map if one was a single cylinder version of the other (and had all other details the same).

Here's a BSFC map for a Saturn, a pretty typical engine:

http://ecomodder.com/wiki/images/3/3..._dohc_bsfc.jpg

For each engine, you'd want to pick the engine speed that gets you closest to the sweet spot. The torque figures along the right of the chart would be divided by 4 for the 500cc engine. For the 500cc engine I'll guess at 2500 rpm, which would require 23 lb-ft torque (31 nm) (which will show up as 124 on the chart). So BSFC would be about 230 gr/kWh.

For the 2000cc engine, at 1000 rpm 78 nm would be required, so consumption would be about 290 gr/kWh.

So the larger engine would consume 290/230 or 1.26 times as much fuel.

At lower cruise speeds, the difference would be much greater.

Um... OK, with that link we now have an example of "talking different languages".

Along the lines of what Ken Fry posted. The University of Maryland had a dyno setup. They were testing a GM Iron Duke 4 cylinder push rod engine, old tech.

At 1200 RPM (may have been slightly higher) they loaded the engine to 20 HP. I don't remember the exact quantity of fuel consumed but lets just call it 1 unit.

Increasing the load up to 50 HP the fuel consumption increased by 50%.

Basically you had 20 HP per unit of fuel, versus 50 HP for 1.5 units of fuel, which closely follows Ken's figures. The extra 30 HP only took half again as much fuel as the first 20.
That's the secret of P&G.

AS long as you know the RPM range for best BSFC and the manifold vacuum reading at 90% load (about 2.8 inches) you know you are in best BSFC as long as you keep the RPM in the range on the graph and the manifold vacuum precisely at 90% of atmospheric pressure.

In Neil's post on this thread he stated that electric motors are 2-3 times as efficient as IC engines. He did not specify which motor or engine, but made a general statement.

In fact if you take specific engines and loads his statement is far from factual. Best BSFC on the Wiki site for is 54.4% (admittedly not a practical passenger car engine, but such limitations were not demanded). No electric motor is over 100% efficient much less over 150% efficient so without limitations his statement has issues.

With the best passenger car engine as an imposed limitation best BSFC is right at 43% and there certainly are electric motors that are better than 86% so using that logic he can claim credibility. The problem with that is when you understand there is a 15% loss in charging a Nissan Leaf, the only commercially available all electric car today, even if your vehicle was perfect it could still never be better than 85%. When you buy liquid fuel, you don't have to pay 15% of the volume in loss to get it in your tank. You must also limit your electric motor choice to the same passenger car limitation, not even considering the weight of the energy storage system.

Electric motors also have BSFC maps, with a peak efficiency figure. Claiming that as the efficiency of the motor is just as wrong as claiming the peak BSFC for an IC engine is the best it can produce. Agendas that ignore facts are fallacious. The cost of electricity is what it cost at your meter. The cost of fuel is what it costs at the pump. If that cost is lower because it is subsidised by taxpayers then it should not be subsidised, because it puts my money in their pocket, which to me is simply wrong.

Pale Melanasian can almost equal the fuel consumption of a Nissan Leaf in his 96 Civic, using his considerable skills and route choice and low average speed. If he drove a Leaf and used the same techniques, the Leaf's fuel consumption would considerably exceed EPA figures, so that is a fair apples to apples comparison, but the Leaf would not be twice as efficient, so his rationale about electric motors is not quite correct. Also consider the generally accepted practice of not charging the battery to 100% or discharging it to below 20% and the fuel tank on the Leaf just shrunk by 40%.

Without engine design optimization, the power train is the secret to higher efficiency. 80% mileage improvement is a practical figure for a power train that allows only best BSFC engine operation. It's also possible with current technology to make an engine better than 40% efficient on gas and closer to 50% on diesel if you optimize the engine for only operating at best BSFC instead of using throttling to control power output, or in the case of the diesel super lean burn AF ratios.

regards
Mech

Fuel efficiency and fuel economy are two different animals. I get the best MPGs in the least efficient engine rpm/torque range.

The best efficiency is when you are coasting with the engine shut down.

regards
Mech

Hello drmiller100,
Well I don't know what make any give design choice better and under which conditions - more swirl, hemi head, indexed spark plugs etc. - but in so far as which engine gets better mpg based on being right-sized I have an opinion.
I'm going to describe an imaginary Otto cycle gasoline engine but other cycles behave similarly.
You emphasized pumping losses in your post, so I will look at that first. From the mechanical point of view, the exhaust pumping losses, and impact on efficiency, are straight forward. Any back pressure is subtracted from the average expansion stroke pressure to find the effective power stroke pressure. So if the average expansion pressure is 200 psi and the back pressure is 5 psi then the engine produces 2.5% less power than it could. Intake pumping losses are approximated the same way, with a couple of constraints. Intake partial vacuum is also subtracted from average expansion pressure. So for an engine with a closed throttle (about 10 psi), but still high in the power band - expansion pressure of 200 psi - the engine produces about 5% less power. The actual pumping power is easy to compute. It's mass flow times suction force. For an average engine, throttle pumping power is always less than about 5% of shaft power. The constraints on intake pumping are that the pressure can't go below a complete vacuum. In the case of exhaust back pressure, that can go as high as the average expansion pressure and the engine still chug along, if barely. The exhaust gasses are compressed to the point that the tail pipe will be shrieking with leaks, but almost all of the exhaust mass will be ejected. The intake gets rarefied until there is nothing there. And the maximum pressure across the throttle is 1 atm, about 14.7 psi.

However, the most important figure determining efficiency is compression ratio. For the Otto cycle the expansion ratio (ER) equals the compression ratio (CR), except that the fuel does take time to burn. In the Diesel cycle the effective ER depends on how far down the expansion stroke fuel continues to be injected. Modern Diesel engines must meet emission limits that limit maximum power, but also allow for high efficiency at high power output.

In the Atkinson cycle, the ideal expansion ratio is just enough more than the compression ratio by the amount of heat added by the burning fuel. Conventional designs allow for about 8.5:1 effective CR and about 11:1 ER, based the fuel burn adding about 25% more than the heat of compression.

Compression ratio determines efficiency because the way a heat engine works is to take your working fluid, raise its temperature, add some more heat by burning fuel and then allow the whole blob to cool off through a machine that turns the heat in to motion. The amount of heat you can turn into usable work is some portion of that flow from the highest temperature to the temperature you end at. The end temperature can't be below the room temp. where your engine is. In fact, for conventional materials, the ejection temperature must be well above the boiling point of water.

Imagine an engine with a CR of 10:1, at some reasonable rpm the compression temperature will be about 650 K. Assume, and this is purely fictional to make the numbers easy, that burning the fuel adds 350 K so the peak temp. is 1000K. The ejection temp is 300K. The maximum thermal efficiency is is the highest temp. minus the lowest temp. all divided by the highest temp. So (1000-300)/1000 x 100 = 70%
Now imagine the CR is reduced to 5:1. Now the compression temp. is about 500K, if you add the same amount of heat the temperature rises by the same proportion. The 350K we added in the 10:1 engine is a rise of 350/650 about 54%. The same rise in the 5:1 engine is about 270K. So the maximum efficiency is (770-300)/770 X 100 = 61%

Why does intake throttling reduce efficiency? It's like reducing the compression ratio. If you reduce the amount of mixture going into the cylinder, that reduces the compression pressure and temperature. Consider reducing the mixture (and amount of fuel) to 1/5th. Now the compression temp. is about 360K the amount of fuel added is 1/5 and the amount of heat added is 360 * .54 /5 = 38. So the max. efficiency is (398 - 300)/398 X 100 = 25%

Here's a crazy exaggeration. If your engine has a very wide efficiency band you could get 1/5 the maximum power by either running at 1/5 the maximum rpm. Say your maximum power occurs at 5000 rpm, you can gear down to run at 1/5th power at 1000 rpm and use about 1/5th as much gas as full power. Or you can throttle down to 1/5th power at 5000 rpm and use about half as much fuel as at full power. Or put another way if you could get 30 mpg at cruising 1000 rpm, you'd get about 10 mpg throttled down but in first gear at 5000 rpm but at the same cruise speed.

Ideally you run your engine at the highest compression pressures you can for the amount of power required. While many gas savers will tell you to limit your use of brakes. It's clear that the big efficiency killer is allowing the throttle to close. Pedal to the metal at all times!

-mort