Once a semi-known idea of the friction coefficient of drag is determined, it can be input as a dynamic function of loss. Might be hard to do with a basic modeling prospect used in excel but it can be included as a factor.

Some interesting reading...

http://www.ewp.rpi.edu/hartford/~ern...2-Bearings.pdf

 

Not even attempting to solve a problem just because you don't even know what tools to use is self-defeatingly ignorant. Let me study my lightning bolts if I want to. Part of the joy of all of this is figuring out what tools to use, what to look for, it's a puzzle, it's how we learn and improve.

People who thought the world was flat told sailors to be safe and only study what they could see, too. Going against convention definitely paid off in that case.

 

If you told me the land was flat, and I Wanted to find out, I would build a flying machine so I could find out for sure. I wouldn't use excel

I'd like to point something out.
A given V8 engine from 2005-2012 of any displacement would not mind 5psi of boost from a turobcharger or anything just about.
Around 10psi is when things get uncertain, but lets say I have methanol/ethanol at this point. So I am still good to 10psi in this scenario on nearly ANY V8 produced from 05-12. Nearly.

Now, how much power is that? I Want to say its around 500bhp for many random V8 (5L-6L), more with just a camshaft swap and mild maintenance improvements. 500bhp also happens to be around the limit to which I would expect any random V8 engine produced from 05-12 to also last for 100,000 miles on the OEM longblock. So here is where I make my comparison: We already know that any V8 produced from 05-12 can be used for 500bhp at 10psi and would still likely give us 100,000 miles of driving. No need to know pressures or coefficients or do anything except weld some metal together and stuff the motor into a 2800lb chassis for my brisk, wide example scenario. And this is where the usability of the OEM longblock ends for most of these engines- 500bhp is starting to get into the "build me forged" bracket if you goal is daily drivers (I am always using daily drivers as a base of operations). This also implies everything we suggested through the years about using OEM V8 engines for traditional swaps in vehicles to gain performance boost for minimal cash, the whole point was to use the OEM longblock for a $500 engine with 500bhp, and a couple spare engines for the inevitable. This is the end result, or inescapable fact of using this viewpoint, that if you could get 2-4 years out of a $500 engine at 500bhp your $15000 in swap components makes sense.


As to your RPM problem, I will point out that if we expect 100k from an OEM engine with an OEM bearing etc... then we will want to limit RPM to a minimal number. Perhaps a VERY limited number, say 6400 or 6800rpm for most daily drivers. Sure you can adjust it for a dyno pass or the track, I just mean to romp around town keeping that number quite sane will go a long way to keeping the engine alive. Tuning is a given, if you are this far in to saving money and gaining performance you need to be tuning your own engines. With such a limited RPM we also know the camshaft will be very reasonable, and turbocharging is the right take advantage of what we do have.

I am not trying to rain or muddy up anybody's ideas, I was only trying to steer the ship towards simpler waters with productive outcomes. I will avoid pointing out the mayhap truths of chasing mystical beasts.

 

It's all in good fun, but the people who figured out the world was round, didn't build flying machines to make that observation. They used the tools available to them. We don't have flying machines, or engine dynos, or the fancy electric motor engine tester... We have excel, and we are at least attempting to understand something inarguably complex to gain a better understanding of how it works. Yes, our tools are rudimentary, probably entirely inadequate for the task at hand, but there's absolutely no harm in trying. Why would you, or anyone else for that matter, find it necessary to be discouraging when all we are trying to do is learn for the sake of learning?

 

Guys, good call on stopping with air and looking at fueling instead. Also, Martin, thanks for that video. That helped redirect as well and simplified a lot of things I was looking into. The energy balance is starting to make a lot more sense now. I had to throw away a lot of the internet HP / Air / Fuel calculator equations and start over.

I'm going to use an example at 4890 RPM, since that is peak TQ and numbers most of use are used to seeing.

At Peak TQ, the fueling is 23.23 g/sec. The average heat of combustion of 91 octane (averaged from a bunch of sources) is 45.7 KJ/g. Simple math on multiplying Fuel (g/sec) x Hc (KJ/g) = KJ/s, which is coincidentally KW. so the total possible energy input into the system (assuming perfect combustion, etc, which isn't going to happen) is in this case 23.23 x 45.7 = 1061 KW of raw energy

I added Hc to the constants list and left 45.7 in for now.

Next is the Otto cycle efficiency - it's the maximum efficiency a reciprocating engine can possibly attain. The equation is:

Otto Eff = 1 - (1 / (r^0.4)) where 'r' is the static compression ratio.

In my case, since SCR is 11.45, this makes the otto efficiency 62.3%. The engine will never achieve this, but it gives a reference point. All your losses start from this point and go forward, NOT from the total energy input. SO, that has made a big difference in the approach.

Multiply the Otto efficiency in, and the total possible output for the engine at peak TQ is 886.7 HP. And the overall efficiency (work out / energy potential in) is 27.3%, which is much more realistic of an efficiency number.

Well, anyway, Starting from there, and converting everything into Mean Effective Pressures (which REALLY makes this easier, because you do the integration up front vs after the fact), It is getting close.

Brake MEP (Using net TQ measured) = 1254.4
Pumping MEP (Using my derived equation) = 595.2 at PkTq
Friction MEP (Using GM's coefficients) = 209.8
Piston Acceleration MEP (using my derived equation) = 506.5

All this adds up to 2565.8 Mean Effective Pressure, which calculates back out to 851 ft-lbs of torque. The otto cycle max torque for the engine at this RPM is 887 ft-lbs. So, with everything figured in, the error in the model is about 4%.

This is sort of what I was expecting. There would be a few major contributors and after that, the remaining contributors would be relatively minor. This is where I suspect your fuel and ignition timing losses come into play, because you have to fire the plug before TDC, so as the crank rotates up to TDC, it is fighting against an increasing pressure gradient until it passes TDC, and then the pressure becomes helpful. Ideally, you'd fire at 1 degree after TDC and get and instantaneous max pressure, but this isn't real world.

I want to go back through the equations to make sure they are real numbers and I can repeat the derivations. Then, I'll post up what they are and how I got them.

 

That is awesome! I honestly expected the percentage of error to be a little higher than 4%, but I find this very encouraging.

Have you ran the model at different rpms?

Do you have access to the data required to run this model with someone else's engine/dyno results?

And you did all this with excel and no flying machines?

 

I just realized I compared a TQ number to a HP number. The error is more like 12%, but still within the realm of reality, because you have imperfect combustion, etc. actually the more I think about it the more I realize 12% is closer to the truth.

One thing that shows up is that the low engine speed high overlap is coming out very inefficient at the low end, which is exactly right. I'm showing very high BSFC at idle - like 2.3 vs 0.5 in the sweet spot. I'm also showing the engine is only 5% efficient at idle vs 27% at peak TQ.

So it's making sense. I think that if I took someone else's VEs and ran the model, it would find peak rpm values but may not accurately predict outputs. I don't have a way to account for valve events yet. With the right amount of data from other motors as Tal0n was suggesting, I might be able to use some Cam calculations to generalize the model, mainly calculating DCR, overlap, and EVO to dynamically calculate pumping losses vs my static pumping loss calcs.

 

You are starting to get it.

I mentioned energy balancing a couple pages back. It deals with valve curtain but the ideal is energy retention after the exhaust stroke. The real issue is, & always will be, that balancing requirement will largly depend on compression & the fueling of the engine.

You only need to largely Factor in pumping losses, due to the inability to have very low ignition advance timing. That is where your losses are mainly coming from. If you derive a calc for pumping losses over the hypothetical , with an increase in compression of 4-5points you will get closer in HP/ TQ production & have a shift in RPM as well. A very large shift by several thousand RPM with a large decrease in required exhaust duration.

As a result, Your BMEP will rise dramatically because the engine has compression non-reliant on cam timing. It will have the compression to compensate for it. But at the same time because it has the compression it will not need as much overlap. Because it requires less overlap it won't waste as much fuel down low or up top. The overall efficency will rise from start to finish.

One thing you need to look at in these pumping losses is the micron size of the fuel droplets, on an average, & the saturation, or homogenization event. That is your pulse width vs intake duration time. That is why I wanted to see the number. With 50-65% duty cycle it will be not so great.

Yeah yeah j know some will say put the pulse width in time with valve events but if you go back to the compression & exhaust valve timing cue you will begin to see it isn't supposed to be the way you see it as needing to be in order to not waste the fuel & decrease pumping losses.

 

I made this for you foxy


Yes is intended to be a joke, but also there is some truth in it as well.

 

I see the humor. The real joke is GDI is best... Diesels showed it to not be the case already so why the manufacturers are fixated on it baffles me.

I owned one a couple years ago, for about 3yrs. It didn't deliver static stoichs nor was it overly energy efficient.

 

I'm not sure what is meant by "static stoich" can you please explain it.

 

"stat·ic
ˈstadik/
adjective
1.
lacking in movement, action, or change, especially in a way viewed as undesirable or uninteresting.
"demand has grown in what was a fairly static market"
synonyms: unchanged, fixed, stable, steady, unchanging, changeless, unvarying, invariable, constant, consistent
"static prices"
2.
PHYSICS
concerned with bodies at rest or forces in equilibrium."

Static stoichiometric fueling: an operational condition meaning that the fuel delivery method is calibrated to deliver a stoichiometric A/F ratio & it remains the same regardless of engine RPM or load.

The ability to do so is most often seen in easily vaporized fuels such as propane & natural gas. Gasoline can do the same thing if it is in the form of a vapor as well. The ability to do so is an indicator of an operational stasis. It doesn't, necessarily, mean it has to be, but a static A/F ratio is still desirable even if not specifically in stoichiometric ratios.

 

I'm not following something in your post. Are you saying that fuel atomization is better with higher duty cycle vs lower duty cycle? Not being argumentative, I just haven't thought of that or read about it before. Is it a function of pressure vs nozzle/orifice size?

Also, pulse width vs valve events paragraph. I think there is a thought behind that paragraph that isn't conveying, so the statement left me confused. Again, not making an argument, I just didn't follow the statements.

Thanks

 

I, too, am curious about the atomization of fuel in relation to the duty cycle of the injectors.

And I think v8r said it best, quite eloquently I might add, about the message behind the pulse width/valve events paragraph getting lost in translation. As it left me mildly confused, as well.

 

Fuel usually sits in a puddle on the intake valve, waiting for it to open. You normally do not want to inject fuel into an open intake valve. It is therefore the temperature (of the valve, intake, piston, cylinder walls, etc), velocity of incoming air, and characteristics of the combustion chamber which mostly affect fuel atomization.


I believe there are several reasons for this behavior. I always thought one was to cool the intake valve (a good thing) and heat the fuel (a good thing for vaporization which is about to ensue). Another was that the chances of a thin vaporized stream of fuel coming out of a fuel injector straight past a hot valve and into a hot combustion chamber is much more likely to start burning on it's way in, without first being able to collect on the valve. The act of collecting the fuel into a small puddle ensures that it wont start burning at it's fringes where the fuel molecules are most likely to be spread far enough part to allow such reactions to occur.

 

See, my understanding of the sequence is obviously incorrect.

I would imagine that you don't fire the injectors whilst the intake valve is open in case of detonation, right? Fire in the intake manifold could result. But my intuition says that you would want to spray the fuel into an open intake valve to avoid the very pooling of fuel against the valve that you just mentioned. I guess in a perfect world, we would not have to worry about detonation and would spray the fuel into an open intake valve, after the exhaust valve closes, to avoid wasting fuel during the overlap period.

 

At WOT fueling the droplet size should be fairly uniform at a given pressure. There is certainly a variance in droplet size but that variance should be uniform. The design of the injector nozzle impacts this but a very low pulse width doesn't allow the spray pattern to develop.

The ideal is to have a homogenization event that lasts for the entire intake event. You can't get a good mixture if the injector fires for 50-60% of the available time. Only a portion of the incoming air is being subjected to mixing.

When I worked with Myron Cotrell when he owned TPiS we discussed these things at quite some length. We would go to his injector flow bench & play with pressures & pulse widths so I could visually see what was going on so I could understand what he was saying.

 

The most important part of vaporization is Kenetic energy input into the fuel. This can be done by heat energy.

A normal injector vaporizes very little of the fuel. If there is puddling at the valve the fuel is not, nor was, completely in a vaporous state to begin with.

 

A homogenization event that lasts for the entire duration of the intake valve?

Would this mean you actually start to spray the injectors BEFORE the intake valve opens, so the homogenization event has already begun as the intake valve opens? And stop after the intake valve closes? This is the only way to ensure all the air going past the valve is exposed to fuel. It also allows the fuel to start atomization/homogenization in the column of air above the closed intake valve, before the valve even opens. And if it doesn't stop until the valve fully closes, you will end up with fuel sitting on the closed valve, doing all the good stuff that v8r mentioned about cooling the valve/adding energy to that pool of fuel.

Am I understanding this correctly?

 

I define intake event as the entire revolution minus a moment of reversion. The air doesn't stop flowing, entirely, so why stop the homogenization?