Turbo Efficiency Discussion

[NateD4]

When you talk turbos and engines, everything is computed by mass flow not volume. Volume is kind of a weird way to address it.

When your car measures intake temp, pressure, manifold pressure and manifold temp (and in some cases MAF)... it is really computing mass flow (I believe they used to call this a speed density system).

For non turbo flow mass flow = density * speed * area. You can solve in any way you want using Algebra.

The speed * area actually give you a volume/second. While the density gives you the mass flow of that volume.

Of course there isn't an easy way to measure density of moving air... so it is calculated by a number of sensors.


What is actually done (and much simplified for this discussion) is to consider the mass of air constant.


We then measure the mass of air coming in and record its temperature.

Then we calculate its energy with some more equations.

Then we add the fuel energy to it.

We then add the fuel mass and air mass and do some special math to figure out how they relate the pressure to temperature based on the chemistry of the burned mixture.

Then we subtract the power we extract from the crankshaft (including friction losses).

Then we expand the gas volume and divide by the ratio. This gives us the final pressure prior to opening the exhaust valve.

THEN we can calculate the density of the exhaust valve.


Once we do that we can figure out the speed of sound of the gas and calculate the maximum speed we can push it through the exhaust valve openings (which is exactly the speed of sound).

What is important to remember is that speed of sound is a function of density, pressure and temperature. Alter any of those and the speed of sound changes and thus the maximum flow rate through the exhaust.

You are correct in saying that the velocity drops as the gas contracts. However... piping size is really more a function of scavenging effects then anything. If you want scavenging in an exaust system you want a high velocity so the wake after the pressure pulse goes by creates a lower pressure then otherwise would be present.


So in general small tubes = fast velocity, large tubes = slow velocity and to an extent less energy.

The real tough part is picking the right size as the Mach number changes based on the pressure in the manifold. The Mach number sets the speed limit of the gas in the tube and therefore its flow rate. So when you hear people talk about hotter exhaust it is usually to give them a higher speed of sound, thus higher flow through the tube. But ALSO scavenging effects from higher velocity are important too.

In the turbine inlet there is a tapering shape that converts pressure energy into velocity energy by reducing the area. This is called a nozzle. It is basically a funnel.

The less area you have the higher the velocity is up until the speed through the smalls area of the nozzle is Mach 1 (at whatever the Mach value is for the gas at that point).

Interestingly enough a you can only move so much air so fast through a nozzle before it refuses to move more air through with more pressure. So what this means is that there is a maximum flow rate through the turbo. The only way to increase this is to increase the exhaust pressure when the area is further reduced. This is where my discussion of efficiency comes in. As the increased exhaust pressure needed to push the air through the nozzle faster causes an increase on exhaust pressure which eventually means a tougher time moving air in and out of the cylinder. At some point the turbo is pushing air just to overcome the exhaust pressure requirements. This decreases the output power unless sufficient additional air and fuel are thrown into the cylinder.

You are somewhat right with the diameters... except we area really talking about velocity to the turbine wheel and flow. With flow comes the exhaust energy that the turbine wheel removes and uses to power the compressor wheel.

Turbine wheels don't make a lot of torque, but they do turn at several hundred thousand RPM. What this means is that using the above formula HP=torque*RPM/525 you can estimate the turbine torque. Say the compressor puts out 40 HP worth of air compressing at 170,000 RPM. You can see that the torque is 1.23 ftlbs. The turbine wheel spins faster for higher velocity and slower for slower velocities. For the most part the turbine extracts more power at higher RPMs with more exhaust energy.

Turbo size is effected by the max flow. A big turbo requires a lot of mass flow and energy to spin, while a small turbo requires less and there fore produces less compressor power...

The differences is one will max out sooner than the other.

Interestingly enough in the 800-1000 HP engine classes the turbos area almost always about the same size whether you are talking about a small 4 cylinder drag engine or a large 550 CI big block. The difference really comes down to how much air each engine flows and how much it takes to power the turbo up.

In fact I'd be willing to bet in a competition drag 4 cylinder that the turbo spools itself up as it delivers more and more air to the engine. Which would give it a slower lag time then the bigger 550 CI... which based on mass flow could spool it up at a much lower engine RPM....


Hope this makes some sense. Any other thoughts?

[TGP37]

That is impressive, I understood everything. It is molecular velocity that determines max impulse speed, which is directly dependant on temperature.

So if we found a way to increase molecular velocity w/o increasing the temps, we can see greater efficiencies? I think I remember reading about longitudinal standing waves altering molecular velocities. The standing waves (which do not move) would need to be adjested so they travel one direction (exhaust flow) at a speed higher then the mach limit. I.e. the standing waves work towards one direction.

Not an easy task BUT, the turbo gives off a unique sound. If the blades were shaped to emit the right frequency, a standing wave may be induced so the negative pressure region can "pull" the exhaust gasses along.

For those who don't understand, imagine looking at a wheel spinning faster and faster. Eventually the wheel can appear to spin backwards slowly. A longitudinal standing wave is almost the same concept, visually.

[AleroB888]

............. Instead of simply absorbing it however, the heat added to the water injection in the intake manifold helps get the water closer to its boiling point (which could be up to 20 or 30 degrees above normal due to the boos(t) pressure.

Assuming the water injection performed as well as an intercooler, the resultant intake charge temperature would be only 20-30 degrees above ambient, so I don't see how it could get closer to it's boiling point, if the boiling point is raised 20-30 degrees.

But even if the water in the charge is just below the boiling point, I'm not convinced it has any benefit except to avoid knock or reduce emissions.

The combustion 'event' then pushes the water past the boiling point and turns it into steam. The steam adds pressure to the cylinder....

In that case, fuel energy in the form of heat would have been consumed to make the steam, so that effect seems to cancel itself out.

[NateD4]

I'm not sure you are understanding the basics of a Rankine, Stirling and Otto cycle (thermodynamically).

A more clear (however more abstract) way to see this would be:

http://upload.wikimedia.org/wikipedia/commons/thumb/a/af/Stirling_Cycle_color.png/220px-Stirling_Cycle_color.png

Or:

http://www.acr-news.com/news/images/663_2.jpg

When you boil water you have to first hear it to the boiling point. (which are functions of pressure) . Then when you have it at it's boiling point it requires additional energy to finally become a gas.

What I'm saying is that you've added additional heat from the intake. This is less heat that the combustion process needs to put into it. Since you are not throwing that heat away with an intercooler, it gets re-used to 'help' boil the water and therefore increases the overall system efficiency & reduces knock.

[NateD4]

http://www.tunersgroup.com/images/image1629.html

http://www.tunersgroup.com/Products/aquamist.html

That might help. There is a graph with the pressure vs time.

I'll have to go into more detail later.

[AleroB888]

I'm not sure you are understanding the basics of a Rankine, Stirling and Otto cycle (thermodynamically)........

In that case, I'm not sure you understand what is happening in the combustion chamber to unboiled water after spark.......

What I'm saying is that you've added additional heat from the intake. This is less heat that the combustion process needs to put into it. Since you are not throwing that heat away with an intercooler, it gets re-used to 'help' boil the water and therefore increases the overall system efficiency & reduces knock.

I'm saying that any unboiled water after spark will take extra fuel energy (heat) to boil it while it cools the chamber slightly.

I have a source that quotes research done on water injection by GM saying little or no change in efficiency resulted, only effective octane.

On a normally aspirated engine, the water injection allowed the compression ratio to be increased 1 point, only then did the system efficiency increase, by 3 per cent.

[TGP37]

The presence of water vapor will also increase the over all compression ratio. So the mass of water in the cylinder can be subtracted from the over all volume.

And there is a point of compression in the power stroke where the vapor must precipitate out as the dew point drops as the pressure rises. And when the water vapor precipitates, heat is released. Heat that was absorbed in the compressor.

So it seems the released heat could keep more fuel in vapor form, increasing efficiency of burn.

And then the cooling of the aircharge can be seen as the same effect as increasing the intake pipe diameter, volumetric efficiency increase. So the trade off of energy temporarily is simulating a slightly better VE.

I'm just tossing this out there, more brainstorming.

[NateD4]

Well, just to be clear, I didn't mean that previous post to be at all insulting. I replied from my phone and typing all this would have taken a long long time...

If you work through the pure thermodynamics of it you'll see that you're changing the shape of the pressure curve, not really changing its overall area (total energy). (though there may exist a set of conditions that increases the area slightly).

You have to be very careful when discussing boost vs NA applications.

In an NA application water injection would allow you to run 12 to 14:1 compression on 87 or 89 octane. Now the efficiency increase would come from the compression increase (power or work more accurately is directly linked to the expansion ratio of the device in question. A higher compression is quite well documented to increase the power output of many different types of heat engines).

Another thing about water injection on NA cars (and turbo for that matter). You can absorb some of the heat conducted in the intake manifold and put it back into the combustion chamber without greatly altering the air density in a negative way.

Again as discussed before... any energy absorbed by the water and put back into the combustion chamber is going to increase efficiency (if only slightly). The reason is basic thermodynamics. In any heat cycle the total pressure available for expansion determines the power output.

The pressure is determined by the enthalpy of the mass inside the combustion chamber. So, in the case of a water injected engine you have: Internal energy of air + internal energy of fuel + internal energy of water being injected. These would all be measured in BTUs or KiloJoules (the metric equivalent of BTUs). Now in the case of air, the internal energy is a function of the ideal gas law (or one of its algebraic forms). the case of both water and fuel you have latent heat capacity and heat of vaporization. And in the case of fuel only you have the chemical energy stored. When you do the math for these you end up with the total energy available for doing work. However due to the net inefficiencies of the internal combustion engine you are lucky to get more then 25 or 30% of it as crankshaft power. The rest goes to friction and heat losses through cooling etc...

Back to water injection:

However the downside to water injection is that you are placing molecules of water in the line of combustion whereby they deter flame front propagation. So, there is a limited degree to which water injection is helpful.

I'm sure if we dug deep enough we'd find dyno evidence that supports engines running better on humid days than non humid days.....



Now, for forced induction especially at high boost levels we are talking something totally different since the intake charge is heated by mostly 'wasted' exhaust gas energy converted by the turbine wheel and transmitted to the compressor wheel where the power does work on the air, compressing it and heating it (following the ideal gas laws).

The heat in the air is what creates detonation, as the mixture is compressed its temperature and thus enthalpy go up. At some critical enthalpy there is enough energy and sufficiently close distance between molecules to spontaneously combust.

The 'easy' way around this is to cool the air with an intercooler. In which case you are literally 'throwing away' some of the power the turbine extracted from the exhaust and putting it into the surrounding air. On the other hand water injection simply 'absorbs' the energy to raise the temperature of water. At the same time the air is cooled and the density is increased (or maintained), allowing more fuel to be added to the charge and in some applications more boost to be run before detonation occurs. Thus, allowing even more power to be made.

Now from a pure efficiency standpoint the energy absorbed by the water injection goes somewhere. In this case it is sucked into the combustion chamber and compressed (at which case the boiling point of the water is increased quite a bit according to the table I posted earlier). When the combustion event happens the total energy in the water is increased (while the combustion gas energy is transferred and slightly decreased) and the water will find an equilibrium point (a point under the 'dome' where the quality of the steam is something less then 100%). At this point we've successfully converted SOME of the water to steam pressure and a small amount remains as liquid. The steam pressure adds to the total pressure in the combustion chamber and contributes to the expansion cycle's shaft power.

The graph from aqua mist shows a small change in pressure where the water injected cylinder pressure is slightly lower in peak magnitude, but slightly longer in duration. If I were to capture the data from those graphs I'd probably find they are nearly identical in total area (and thus total energy).

So you are to some extent correct that it may be a wash. However the gain you get is in either running an increased compression ratio with a turbo or using more boost without higher octane fuel. To me thats the gain in efficiency (perhaps mis-labeled): running higher compression or boost making more power with the same or less fuel or... cheaper fuel.

after all most of us are only interested in efficiency for the sake of operating cost. Otherwise we'd all be running race fuel on the street with high compressions (of course I neglect the EPA NO2 requirements that high compression alters....).

I'm not sure if that makes more sense or not... I'll try to get some math posted, however I'm not sure when I'll have some free time to type it all up.

As to water injection, its been around since WWII when bombers used it to help avoid detonation at high altitudes.

As to GM's reports.... I'd take anything GM publishes (short of a very well documented white paper) with a grain of salt. The initial Volt advertisements and press releases claimed 100+ MPG numbers... which really were 100 MPG (effective)... and in my opinion mislead consumers who didn't understand that they had to pay for the electricity that is used to help achieve that 100 MPG number. GM makes great cars, but their marketing (and many other companies) sometimes make claims that aren't really substantiated.

------
One reason no one uses water injection today is the hassle; imagine having to put a gallon or two of distilled or demineralized water in a separate tank a month.

The other primary reason is the requirement to mix alcohol with it in extreme cold weather to prevent freezing. Until recently (with Urea diesel injection fluids) it is probably perceived by 'normal people' to be more hassle then its worth in fuel savings... but there may come a day just as with urea in diesels that the EPA requirement will force water injection on us.... In which case I have no real issue adding water to my tank. The trick to it will be de-mineralizers to allow tap water that won't clog the injectors up with mineral deposits after years of running.

Sorry this post was so long....

[NateD4]

TGP37,

Water injection does not effect the compression ratio (as defined by the inlet area vs the outlet area). If anything it increases the compression ratio slightly (though I seriously doubt it) as the total volume of compressible gas is decreased.

You are somewhat correct that the water state could change during the entire compression, ignition, expansion stroke. On compression the water vapor will find a new equilibrium point and any evaporated gas could possibly condense back into water. Once the combustion event happens it stores up energy and as expansion happens it turns back into gas as a function of cylinder pressure and the steam-quality equilibrium points.

[NateD4]

Water precipitating may not be quite the right term. In general precipitation is the chemistry term used after a chemical reaction takes places. What you are probably trying to get at is condensation which can happen at high temperatures and pressures.

In the intake tract I'm not sure you could think of the intake as larger... nor do I think the water injected temperature change would significantly alter the fuel vapor state...

Water injections primary gain is in preventing detonation and all the wonderful benefits that come with it.

It does increase the air density in the intake and allow a slightly greater flow, however there are some losses that occur with the change in effective density...

One big gain is absorbing heat from the intake manifold.

The other gain is in re-shaping the pressure curve in a more favorable way.

I'd be very interested to see water injection on a modern direct injected engine where the available fuel/air mixture is maintained more evenly as the piston travels down its expansion stroke. In such a case the extra fuel could be pumped into super-heating the steam in addition to making the gas volume greater and therefore could potentially make even more power then a non water injected engine....

In fact I could see a water injected DI engine running 13 or 14:1 on 87 octane... that would mean the current Camaro/CTS 3.6 DOHC engine would make almost 350 HP maybe more with perhaps a 1 or 2 MPG increase in cruise MPG from the compression change.

[TGP37]

So the water is sort of buffering the combustion process? I see how the vapor would absorb heat energy during combustion and that probably happens rapidly in the first phases of combustion. So the energy absorbed is the initial flame front, allowing a softer, more productive flame front.

But any added water is also going to absorb more heat from the intake valve. Heat that is used to help vaporize the fuel.

[NateD4]

Yes it buffers in a sense... more like it buffers detonation....


Yes you are right about the intake vaporization... however I've never thought of it that way.

The ideal situation (and the reason aqua mist sells a special controller for their system that resembles a stand a lone engine management system) is to inject just the right amount of water injection so that the water is heated to about 98% of boiling temperature. In which case it won't absorb TOO much extra energy from the valve... and if it does it'll turn to vapor.

Again direct injection and water injection could prove to be a very very interesting combination....

As a reference alcohol has different flame properties and (I think) requires more energy to vaporize then gas...

http://www.sae.org/mags/aei/MANUF/7656

Very interesting article about what you can do with a high octane fuel and the right engineering.....

[TGP37]

Yes it buffers in a sense... more like it buffers detonation....


Yes you are right about the intake vaporization... however I've never thought of it that way.

The ideal situation (and the reason aqua mist sells a special controller for their system that resembles a stand a lone engine management system) is to inject just the right amount of water injection so that the water is heated to about 98% of boiling temperature. In which case it won't absorb TOO much extra energy from the valve... and if it does it'll turn to vapor.

Again direct injection and water injection could prove to be a very very interesting combination....

As a reference alcohol has different flame properties and (I think) requires more energy to vaporize then gas...

http://www.sae.org/mags/aei/MANUF/7656

Very interesting article about what you can do with a high octane fuel and the right engineering.....

I'll read that. I reminds me of the B-2 Stealth bomber....as it could never fly properly w/o constant computer adjustments.


One concern I have with water/methanol injection is the methanol. It is corrosive to aluminum and I would definately want to plumb a recirculation system from the BOV to prevent water/meth from being sprayed all over the engine bay.

But I am reading methanol can increase power output even more so when combined with gasoline and nitrous oxide. So the concept of gaining even yet more efficiency by adding small amounts of nitrous oxide is an idea.


"One reason no one uses water injection today is the hassle; imagine having to put a gallon or two of distilled or demineralized water in a separate tank a month." - Yet some people rather add a fluid once a day to avoid a costly repair. People are funny like that.


EDIT: Just read the article. Using aluminum components with steel lining. That answers my question about corrosion, lol. I like the idea of Primary Alcohols used as a fuel source. They degrade in nature rapidly which can prevent the stuff from polluting water tables, rivers, etc.

[NateD4]

I'm pretty sure the steel lining refers to the cylinders only.

Also you can use ethanol rather then methanol. It is a bit gentler on aluminum.

But really if you anodize your parts I think it becomes a non issue (though I have to do a bit more digging to be 100% sure).

The ONLY reason you put anything in the water is to prevent freezing. In fact some people use blue washer fluid as it is water with alcohol and some blue coloring in it. Of course it has the ability to leave behind the blue coloring and clog jets etc....

[TGP37]

Yeah, I figured that part out preventing the water from freezing.

Methanol itself will also produce steam (and carbon Monoxide) from combustion. If anything else it is steam energy that hasn't been converted to steam from existing heat. So a 50/50 solution will absorb heat as normal, but the steam output is doubled with only having to put heat energy into half of it (the water half).

So there is a bit of energy extracted from the use of methanol. I'm just not 100% on the energy required for methanol to produce steam + carbon monoxide.

Plus, it seems the use of copper plugs and methanol will create formaldehyde in the combustion chamber. Mix copper and methanol with 300+ degrees F = formaldehyde. Though it should only be in small amounts. Called dehydrogenation of Methanol.

What are your thoughts on Methanol combustion introducing yet more steam into the combustion process? As it is steam that hasn't absorbed residual heat in the intake system and the intake valve. Is it possible to use 100% methanol instead?