Compression, turbos, and efficiency
 

Reading Christ's thread about a turbo install, I realized I know next to nothing about compression ratios, turbo chargers, and how it affects fuel economy. In this thread, I want to consolodate all of my questions and hopefully begin to understand the physics behind it all.

I have read before that increasing the compression ratio increases fuel efficiency, but what is the reason behind this?

If increasing cylinder pressure increases fuel efficiency, why not direct-inject gasoline into the cylinder at just the right moment and not worry about detonation (like a diesel engine)?

This leads me to the question of why it's important to have a stoichiometric air/fuel ratio in a gasoline car when it doesn't matter with a diesel?

I imagine a turbo increases the effective compression ratio, and this explains the increase in fuel efficiency. Why not just engineer a higher ratio into the design of the engine?

The turbo has a dual benefit - it reduces pumping losses, and makes use of excess heat from the combustion cycle, which would otherwise have been wasted as noise from the exhaust.

The compression issue, I don't fully grasp, myself. I know that there is a limit to comp enhancement, where it just takes more energy than it's worth. Diminishing returns and such. Basically, its a mater of heat and ratios.

Heat, in that you have enough heat for total combustion of the mixture AT THE RIGHT TIME.

RATIOS: mechanical advantage, pure and simple. The more leverage produced, the more energy is transfered to the crank shaft. Increasing comp can achieve this, because it changes the ratio of induced volume to compressed volume. However, it takes energy to compress air and fuel, and there is a point of diminishing returns.

This is the most interesting aspect of a college level course on thermodynamics. The 'dirt simple' answer is the expansion ratio is effectively increased as the fuel-air charge at ignition is increased. But the detailed answer is a little longer ... you might check the Wiki or ask Mr. Google. I'd have to reconstruct my college course notes (from 40 years ago!)
That is the theory of direct injection engines and promises to make a significant improvement in ordinary ICE engine operation. Everyone is working on their versions.
It has to do with the 3-way, catalytic converter that needs an oscillation between lean and rich. A balance is needed so the oxides and hydrocarbons combine to complete the hydrocarbon combustion and convert the oxides from reactive NO{x} to N{2} and CO{2}.
Weight as a turbo-charger lets a physically smaller engine perform at power levels a substantially heavier engine would require. It breathes like a big bore and makes power like a big bore but it is much lighter so there is less rolling drag and inertial losses (aka., when the brakes have to slow or stop the car.)

Bob Wilson

It comes down, to a great part, to the chemistry of the fuel that is being used.

Increasing the compression ratio, which increases the pressure and heat inside the cylinder creates, for lack of a better term, a bigger "bang" that should allow for more complete combustion and less loss in terms of radiated heat.

Gasoline direct injection allows higher compression ratios, but once again, if the fuel quality is such that it does not react completely, you will have pre-ignition/detonation problems. This has been dealt with through higher injection pressures, better atomization and knock sensing ignition systems by and large.

Every fuel has an ideal fuel to air ratio chemically and diesel fuel is no exception.

It's all about control. Diesel engines are designed to withstand high cylinder pressures and abnormal combustion events, along with very lean mixtures. The fuel allows this.
That rattle that comes from a diesel? Detonation / pre ignition.

As long as the fuel / air mix is prepared properly before introduction into the combustion chamber for reaction, and the harder you compress it, the greater and more thorough the release on the power stroke.

Gases burn completely, vapors burn completely, droplets do not. Because most fuel systems take things down to droplet size, whether carburetor or fuel injection, combustion is never complete and hence aftertreatment is needed to deal with emissions.
If the reaction was complete inside the engine, then it would not produce emissions and the fuel economy would go up. The engine would run cooler.

If I missed something, I apologize in advance; there is so much involved.

Ok, my brain re engaged; here's some more. When you look at modern engine designs on the gasoline side, you'll hear about high swirl combustion chambers, tumbleports, etc. This is in reference to things designed into cylinder heads to induce motion inside the combustion chamber to help make the fuel / air mixture more reactive, and therefore more complete. In other words, better vaporize everything mechanically so that it reacts more completely. An example of this is the May "Fireball" combustion chamber design that was used on the 5.3L V12 Jaguar engine as installed in the XJS back in the day. This allowed the compression ratio to be raised to 11.5 to 1 on the pump junk we have here, and lowered the fuel consumption quite a bit. Your Acura is another example; when you look at the cylinder head and the piston crown shape, along with the port shape, they all conspire to create a lot of motion in the cylinder, breaking the droplets down even further and making things more homogenous for combustion. This is the reason Hondas run so clean, and can run low octane fuel on a high compression ratio.

The problem with engineering the compression in with the current fuel systems and current fuel chemistries is ( I think ) NOx production as a consequence of higher combustion temps. Now, this is just me, I find it debateable, but who am I. I believe there's a way around everything, just need the time.

Sorry about the segue..........mind drifts. Anyway, the fuel we have is the limiter along with.....oh better stop now before I rant.

Fuel atomization, by the way, is the reason you don't polish intake runners after the injectors. Turbulence helps break up the drops. You can polish them up to within an inch or so, though, and it will still help.

Why don't manufacturers turbo-charge all cars? It seems a sure bet it would more than pay for itself by avoiding the need to build a bigger engine, and in fuel savings.

My point is, it would actually be desirable for fuel to instantly combust if it is directly injected into the cylinder just at the right moment in the engine cycle. It shouldn't then matter what octane the fuel is since we want it to burn right as it gets squirted in. Diesels do this.

Surely gasoline engines can be designed to withstand the increased pressures, heat and such. Diesel engines do this while burning a fuel that has 30% more energy density.

Gasoline burns alot faster than diesel, which means much higher cylinder pressure. Technically, the engine would have to be even beefier to withstand the violence of a gasoline explosion.

This is addressed by the "pee bottles" (urea injection) that many modern diesels employ. Could this not be used by gasoline engines?

I'm not sure a faster burn equates to greater pressure, but perhaps the pressure builds faster (explosion vs rapid burn)? Afterall, diesel has more energy density, so that would imply it generates more pressure, right?

Common-rail diesel engines inject a small amount of fuel just prior to main injection, and this results in a quieter engine. Perhaps something similar can be done with gasoline that would more gradually load the pistons, rods, etc, and reduce the violent forces they encounter.

You got it. Faster pressure rise.

It gets back to the chemistry of the fuel. Even though gasoline direct injection and diesel direct injection do the same thing, the fuel reacts differently.

Look at what happens between seasons with the blends of gasoline and diesel. Mileage usually drops. Gasoline with ethanol versus alcohol free gasoline, there is a mileage difference. It's like that because the fuels are designed to burn instead of instantly combust like you mentioned, and because of this, no engine running on what is accessible at the pump will run as good as it could.

The engineers and chemists can do a lot more in terms of efficiency, but are only going to be allowed to go so far with it because of business. It's unfortunate, but that's the reality.

That is happening but technology often moves forward in fits and spurts. Every part in a car adds to the cost sometimes and the accountants who run companies choose a less than optimum solution for one that is "good enough." This penny-wise but pound foolish attitude (aka., their own pay and benefits) almost killed two of three USA car makers and for Chrysler, this is their second time at bat.

Bob Wilson

The new Mazda skyactive system uses 14 to 1 compression with fuel injected directly into the combustion chamber in stages at very high pressures. Fuel is initially introduced to get ignition, then fuel is added during combustion to eliminate the shock wave of pre ignition.

Ultra high pressure (like 2800+ PSI) actually exceeds combustion pressure. Direct injection allows the fuel to absorb heat directly from the combustion chamber, and the multiple injections prevent pre ignition by not allowing a sufficient quantity of fuel to preignite in the first place.

All of these developments are in the direction of the goal of true HCCI combustion (homogeneous charge compression ignition) where mixture ratios of 25 to 1 can be ignited without knock since the fuel air charge would be completely homogeneous at the point of combustion.

It may actually be done, and when it is with whatever combination actually gets the results, we may see IC engines that get 60% fuel conversion efficiency. Argonne labs is already working on that premise, and when it is done and capacitive storage and application of energy are combined, you will see 2000 pound 5 passenger sedans regularly getting over 100 MPG in normal operation, up to speeds of 70 + MPH as long as the aero is good enough to keep power requirements low at higher speeds.

regards
Mech

Here is another pathway that will eventually be used, probably together with other developments.

Green Car Congress: SwRI simulations show turbocharged Scuderi Engine could reduce a standard 2011 Nissan Sentras fuel consumption by up to 35% on FTP cycle

regards
Mech

So far I'm disappointed at what Google has to say on the subject. From Wiki:

"A high compression ratio is desirable because it allows an engine to extract more mechanical energy from a given mass of air-fuel mixture due to its higher thermal efficiency."

In other words, a high compression ratio is more efficient because it is more efficient.

From answers.com:

"Engine efficiency is increased though compression ratio by allowing a more thermodynamic energy to be converted into mechanical energy. Energy transfer is the key to efficiency."

I believe all things can be distilled into a simple concept, but it's improper to define something by using the subject as the definition. Those that cannot explain something in lay terms often do not have a firm grasp of the concept themselves.

Now, I have a hard time understanding how a higher compression ratio equates to higher thermal efficiency in light of the 2nd law of thermal dynamics. In my mind, increasing pressure also increases the heat density of the fluid. Since energy always flows from higher concentrations to lower concentrations, this should mean that the highly compressed fluid would give up more of its heat energy to the surrounding cylinder walls, piston, and head. Fluid that is not compressed as much would not give up as much heat since the temperature differential is not as great.

So, something about the higher compression does cause greater thermal efficiency despite the heat that is lost due to the increased thermal density. The principle behind this is still a mystery to me.

:thumbup:Let's see if this helps.

Compression-ratio encyclopedia topics | Reference.com

The link added this to the explanation:

"High ratios place the available oxygen and fuel molecules into a reduced space along with the adiabatic heat of compression - causing better mixing and evaporation of the fuel droplets. "

This statement implies that lower compression ratios do not adequately atomize the fuel/air charge. I'm wondering if there is something being left out of the explanation. Perhaps the increased pressure and heat also causes a more rapid burn, and therefore the expanding force has a longer stroke (duration) to perform work? After all, burning gas the moment before the exhaust ports open would allow only a very short duration for useful work to be extracted.

Setting this topic aside for the moment, I'm wondering what kind of real world efficiency gains I can expect from installing a turbo and swapping in a taller final drive. Of course the answer is that it depends, but I'm also considering a front grill block, and I wouldn't run a grill block if I had a turbo. So would I likely get more gains from a grill block, or from a turbo/gear install?

I'm sorry, I should have been more specific. So let's start with this URL:

Internal combustion engine - Wikipedia, the free encyclopedia

They don't really discuss compression ratio but this is the take away:
Practical, commonly sold engines have abysmal thermal dynamic efficiencies. To be perfectly blunt, they are rubbish compared to what they could (and should) achieve.

So lets switch to two university web pages that go into details addressing the effect of compression ratio:

3.5 The Internal combustion engine (Otto Cycle)
Design of an Otto Cycle

Both papers show the math to derive the same efficiency formula for a 'perfect' Otto cycle engine:

efficiency = 1 - ( 1 / (r ** (k-1) ) )

r - compression ratio
k - specific heat ratio, a measure of energy in the fuel

For simplicity, here is the MIT chart:
http://web.mit.edu/16.unified/www/SP...nRatio_web.jpg

Here is the Northwestern chart:
http://www.qrg.northwestern.edu/ther...efficiency.gif
Note they have have different scales with the MIT showing the full range and the Northwestern a more practical range we find today.

Now everything in these classical, Otto cycle lessons derives from the P-v diagram:
http://www.qrg.northwestern.edu/ther...Pv-diagram.gif

The top curve coming from point 3 towards point 4 is the power stroke curve and the longer it is, the more energy extracted. The x-axis is the "v" expansion ratio.

Bob Wilson

ps. I don't know if you are interested but the Prius Atkinson cycle takes the line from 1 to 2 and breaks it into two sections:

1 to 1.5 - this is a flat line as the intake valve is kept open and part of fuel-air mix goes back into the intake manifold to be sucked in the next cylinder.

1.5 to 2 - this is the shortened compression stroke which being smaller, means less compression losses.

Qs - the energy added is less because there is less fuel-air to burn

Qout - is the same

4 - is moved to the right about 50% further (aka., 13 to 1 expansion versus 9 to 1 typical compression ratio.) So the Prius has a much longer power stroke to extract more energy.

I do not like this P-v chart but it is 'close enough:'
http://upload.wikimedia.org/wikipedi...nsonMiller.png
I do not like it because the segment 3-4 implies a substantial increase in volume without work being done and that doesn't happen. If you stretch 4 over to combine it with 3, you'll have an accurate, Atkinson cycle P-v diagram.

Bob Wilson

I'll tell you what I would do if I had a TSX. With the type of powerband it has, I would either put an Eaton TVS style supercharger, a Lysholm screw compressor or a Rotrex on it and regear it. With the modern designs, the parasitic losses are not as bad ( they have a bypass when not in boost ), and the instant boost beefs up the bottom end torque really nice, which is what Honda engines need. The added air plus some careful tuning boosts the mileage, and the bottom end torque allows lower shift points. Not a cheap way to go, but what the heck, it's only money.

That's a nice writeup Bob. Just to clarify there are lots of cars running a "modern" atkinson cycle aside from the prius:
Atkinson cycle - Wikipedia, the free encyclopedia

+1 Bob, excellent:thumbup:

If they get the variable intake and duration valve systems worked out, Atlkinson will be the norm, not the exception ... and I can't wait to see it happen.

When I was offered a chance to buy an early 2010 Prius and learned they only have variable angle valves, not variable duration, I almost walked away from the deal. Even now, our 1.5L, 03 Prius and the 1.8L, 10 Prius are getting 52 MPG. Of course the 1.8L has more space and does it 5 mph faster BUT I wanted 60 MPG. <sigh!>

Bob Wilson

Of course that is where adjusting the nut behind the wheel comes in mighty handy :)
EcoModder Fleet list - EcoModder.com

I cant wait for better valve operation either.

Are the Prius' sold here different than the overseas units? I notice that they mpg better than our units; or am I looking at Imperial versus US units of measure.

To the best of my knowledge, the same but it really takes friends overseas to check. For example, the North American NHW11 has a hydrocarbon converter not found overseas. However, they had an optional fold-down rear seat. The North American NHW20 has a thermos for hot coolant that is missing in the EU and Japanese NHW20s. Apparently the ZVW30 can have a 'heads up' display that is not available in North America. But the fundamental vehicles and drive trains are the same.

Now the EU and Japanese mileage testing protocols are different from ours and more different after the EPA 'adjusted' them in 2008. So even after adjusting for gasoline units, the respective government tests are different and ours are more conservative. Funny, the modified EPA tests have not really impacted the actual mileage. <wink>

Go to Fuel Economy and compare the EPA numbers versus user mileage for:

• 50(EPA) vs 48.8(139 vehicles) - 2010 Prius
• 41(EPA) vs 49.1(16 vehicles) - 2010 Insight
• 34(EPA) vs 44.8(13 vehicles) - 2010 Jetta TDI

So the 2008 'adjustment' fixed the Prius numbers but the Honda Insight and Jetta TDI still look bad. What is really sad is what happened to the NHW11 Prius:

• 48(old EPA) vs 41(new EPA) vs 45.4(24 vehicles) - 2003 Prius

The EPA error has flipped to the other side.

Bob Wilson

Found an interesting post in USENET to complement this thread. In particular, the first link
Some of the articles are a little dated such as the 1.5L Prius engine analysis. Still, it is a good collection of URLs . . . Thanks to Bruce Richmond.

Bob Wilson

Getting back to the direct injection discussion (if you don't mind me rewinding the thread a bit here), the reason that diesels can run lean mixtures more easily is because the intake air charge is compressed to a temperature that is above the autoignition temperature of the fuel.

Since diesels are essentially unthrottled, a full air charge enters the combustion chamber with every intake stroke, no matter what the power setting is. In this sense, where spark ignited engines vary the intake charge (and compression ratio) - maintaining a constant air/fuel ratio, compression ignited engines vary the air/fuel ratio - not the intake charge.

Listen carefully next time you hear a diesel vehicle accelerate, then coast. They are quite loud with that traditional diesel "knock" during acceleration, but the combustion becomes almost inaudible during deceleration. This is because diesels do not need to maintain a constant air/fuel ratio in the cylinders, and can basically cut the fuel entirely when none is needed. DFCO (deceleration fuel cutoff) is an integral part of compression ignition engine operation. If you gun a diesel engine from idle, you'll hear it rev up loud, then wind down very quietly until it reaches the set idle speed, then sort of "whoosh" back into steady-state idle.

There is a limit to how lean of a mixture spark-ignited engines can run because beyond certain air/fuel ratio (somewhere around 18:1 for gasoline, I think) the fuel molecules are simply too sparse in the air charge to fully combust from being ignited by a spark. This is not a problem in the compression-ignition engine because the intake charge is always hot enough to ignite any amount of fuel.

So, from what I gather, it's the spark-ignition which is the limiting factor preventing gasoline engines from being able to direct-inject like diesels do. A diesel engine is actually closer in operation to a steam engine, partly because the higher energy content of fuel requires more heat and burns slower than gasoline does. In regards to diesel fuel injection, the duration of injection is one of the main factors that determines the amount of power generated in the cylinder, very similar to a steam engine.

As far as the compression discussion is concerned, I'm no expert, but my normally aspirated gasoline engine gives me more miles per gallon accelerating at only 75% load than at 100% load. Perhaps this is simply due to the design of the engine and transmission. But since the compression ratio is lower at lower throttle settings, due to smaller intake air charge, this would mean that my engine is more efficient at a lower-than-maximum compression ratio. Just my thoughts...

Old Mechanic brought it up first, here's an article on the Mazda Sky Activ engine (s).


Mazda SKYACTIV-G 1.3 Engine Details | Motorward

The gasser is direct injected.

Hi abogart,

Excellent post, but a clarification:

I don't know if you mean 75% of full throttle at high rpm or wot at 75% of red line or something else. It's very likely that the engine is most efficient near wot and very near the rpm for peak torque.

-mort

Yes, it's direct injected. But it is still spark-ignited, and therefore subject to the limits of how lean a mixture can actually be ignited by a spark. Although, if it is truly the lack of a homogeneous mixture - such as Old_Mechanic mentioned - which is the limiting factor, wouldn't spark-ignited engines running on gaseous fuels (CNG, propane, etc.) be able to run very lean mixtures without compromising ignition?

Also, I notice that they market the engine as having 14.0:1 compression. Is this geometric compression ratio derived from engine design, or actual maximum WOT compression of the intake charge? If the variable valve timing allows the intake valve to remain open during some of the compression stroke (not sure if this happens at WOT), the compression ratio would be lower due to reduction of the amount of cylinder area compressed. Sometimes I think that automakers are just developing extensive new technologies to compensate for the flaws in one type of engine, while basically ignoring that another type is just more efficient. :rolleyes:

I'm not trying to trash the new, innovative engine; I'm simply being skeptical and objective. :p

Sorry about that Mort! :o

Allow me to clarify... My definition of "load" was the manifold pressure (observed MAP as in. Hg on Scangauge) relative to atmospheric pressure (standard 29.9 in Hg). Regardless of throttle position, this most closely represents the percentage of maximum engine power being used at a given RPM. I rarely use more than 35% throttle (as observed TPS on SG) at any time. Under certain conditions (1400 RPM, TC locked), 29" MAP (100% load) can be obtained with as little at 25% throttle in my car.

Through actual experimentation, I found that accelerating at 22" MAP from 2000 to 2500 RPM gives less of a drop in average MPG than accelerating at any higher throttle setting in the same RPM range. I initially derived the 75% load value by figuring 22 / 29. However, I just realized that this would be inaccurate because the MAP does not go all the way to 0 at closed throttle, but closer to 7", depending on RPM. So I guess that I could reduce the scale to 22 and figure 15 / 22 which would be about 68% load... But I think there is some math involved here which is way beyond my current level of understanding :o.

Anyway, I found 22" to be the "sweet spot" which yields the best MPG for the way that I drive. Note that the car does have an automatic transmission (doh!) and these numbers may be construed by torque converter losses or any number of other factors. Not to mention that I really cannot do any RPM vs. throttle experimentation outside what the PCM will allow :(. You might very well be right though Mort, the auto trans. really complicates things.

A reasonable explanation why WOT is not more efficient at making power than %75 throttle on a gasser may be that the cars computer goes into enrichment mode at WOT (the obd port MAY indicate open loop operation under these conditions as well).

I have observed this using Scangauge. Not sure of the actual TPS value, but it was definitely somewhere between 75% and 100% throttle opening. Loop goes to OPEN and AFR (air/fuel ratio) drops to 12.5 from 14.7.

Another reason that higher throttle settings may not necessarily be more efficient is fuel's resistance to detonation. I have observed while running regular 87 E10 fuel that the engine's knock sensor will detect detonation and start retarding ignition timing under as low as 23" Hg manifold pressure in certain cases.

Check out this thread for more on this subject.

Higher compression adds mechanical advantage to the power stroke, independant of other also desirable factors.
Consider that an engines effective power stroke is mostly done barely 1/3 way down the stroke, and think about it this way:

Imagine a small firecracker being used to push a cannon ball out of a cannon. Would you get the best push on the cannon ball if the ball was right down on tight on top of the firecracker, ( high compression)or if it was held a bit away, giving some volume around the firecracker? ( lower compression)

Having a fixed and limited amount of firecracker produced gas available means that with more chamber volume the peak pressure realized will be reduced, reducing energy transfered to the cannonball even though you used the exact same firecracker, and the cannonball will go less far.

Hope this helped.

Dean in Milwaukee

That's a nice analogy, Dean :thumbup:.

But isn't the added mechanical advantage of higher compression counteracted by increased energy loss during the compression stroke? It seems to me that if it takes more energy to compress the fuel/air charge further, but you get more energy out of it, shouldn't the two factors just cancel each other out?

"But isn't the added mechanical advantage of higher compression counteracted by increased energy loss during the compression stroke? It seems to me that if it takes more energy to compress the fuel/air charge further, but you get more energy out of it, shouldn't the two factors just cancel each other out?"


Air is a perfect spring. 100% of the energy used to compress it is directly returned when it reexpands, at least thats how it would be in a perfect world.

The reality is that higher compression creates more piston friction, and part of the heat of compression and therefore its energy is lost to the cylinder. If this was'nt true, youd'e have a perpetual motion machine. :)

The way it works out though, a large % of the energy needed for compression is in fact simply returned to the piston on the downward stroke, and the extra friction and heat losses from bumping compression from say 9:1 to 11:1 is a fairly small increase, and far offset by the extra mechanical advantage gained by the increase.

This is especially true when an engine is throttled way back in cruise operation. The net effective cylinder pressure is reduced by the throttling, hurting system efficiency. An engine can actually handle much more compression at low throttle settings without knocking, but has compression ratio's chosen to handle heavy throttle application to optimize max output.

I've often wondered what sort of mpg results you could get by going sky high on compression to max out efficiency at cruise settings and just greatly retard timing for heavy throttle, ( making efficiency and max power at heavy throttle suffer) since the engine really spends very little time at heavy throttle anyway.

Dean in Milwaukee

I have wondered the same thing, yet what would it take to get something like 30 to 1 or more static compression with what's out there now?

That's where cooled-EGR comes in.
Adding a big chunk of cEGR at FOT and high loads can hold off knock quite effectively, without the need for traditional enrichment-charge-cooling.
Clearly you are going to loose a few % of peak HP but for a much better top-end FE.
http://www.swri.org/3pubs/ttoday/Sum...n-and-Cool.pdf

Thanks to ConnClark for link

In this way max. power suffers a little but not efficiency.