peejay

Those ratios have air mass on both sides, which really IS the only way to do things when you are calculating from measured data.

However "volumetric efficiency" only takes air volume into account.

And, an aside, I was taught to set up GM computers by disabling the mass airflow, then fixing the VE table so that there was as close to zero fuel trim correction as possible, and then re-enabling the MAF and adjusting the MAF transfer table based off of the VE table that I'd just made. The hardest part was the simple act of wrapping my head around the new (to me) paradigm of air mass per cylinder per stroke, which really DOES seem like the best way to do things because if the mass is X then you know exactly how much fuel it will need, period, and you will also know the best timing because it doesn't matter if it is because of high VE and low manifold pressure or low VE and high manifold pressure, the cylinder pressure will be Y any time the mass is X...

arghx

iTrader: (3)

I looked back at those formulas and I realized that I put hydrocarbon fraction when I meant humidity.

yeah but you have pressure, temperature, humidity in the formula so you can move back and forth between mass and volume units. That's the basis of a modern ECU with a gas flow model for control rather than a few lookup tables.

Yes, that's how you do it when you are tuning in a car. But do you know how that kind of system is put together originally? They have engine dynos running automated testing for hours and hours on end with a gazillion lab grade sensors, spitting out data to develop a physical model inside the ECU. Then the software on board the ECU is smart enough to use learning to compensate.

milligrams of air per charge is just another way to represent charge air into the cylinder. it's all a function of airflow/rpm pretty much. Subaru uses grams of air per revolution. The old FB and FC ECUs use a load term similar to airflow by rpm. The Rx-8 ECU uses a charging efficiency term closer to the atmospheric VE formula I posted.

In a basic feed-forward sense, yes. However in the ECU you have errors in the on board sensors and you have some accuracy of the model. You also have to make assumptions about the fuel and you have to calculate a target air fuel ratio. That's why you end up having closed loop fuel control. I'm sure you know that though.

I can see why you would think that. Keep in mind that those most advanced ECU systems have a full torque and combustion model inside based on burn rates and/or predicted peak pressure location. The spark is tuned based on the combustion data found in the lab.

What ends up happening is that, at part loads, you have a bunch of different conditions that affect the combustion speed such as the amount of residual gas in the chamber. The spark needs to be advanced when you have slow combustion during low load overlap for example. A torque model and combustion model based ECU can understand that based on tons of data and maps that are not well understood in the aftermarket.

The peak combustion pressure is greatly influenced by engine speed & burn rates, and also burn timing (combustion phasing/ location of peak pressure). If my location of peak pressure is around 11 degrees ATDC on a piston engine, I'm usually at MBT spark timing. On a rotary it seems to be much later. I haven't seen any published studies but from the little bit of work that's been done around here it seems to be about 45 degrees ATDC.

peejay

I'm well aware that, of the parameters I can tweak, there are probably 10x more that are not available to me

What I am getting at is that it is a far cry from the manifold pressure based systems where you tune to an empirical assumption, which works well enough if you can tailor the tune to each individual engine.

I do like how utterly detailed the fuel injector modeling is.

I just wish that HPT would crack the port injected 4-cylinder computers as thoroughly as they did the V8 stuff. I'd love to put a GM PCM on my GTI. But the tables that they give you access to are extremely rudimentary, even for the Ecotec stuff. I do respect that there is a lot more demand for tuning V8 computers than Cavaliers so they have to prioritize

zak rabbit

Here's a wrench to throw in here:
As it's been explained to me, the reason the old F1 cars were able to make such high hp was due tio the fact fast they achieved higher intake pressure than exhaust manifold pressure; I've heard them described as "a turbine engine with a reciprocating internal combustor."

Vicoor

Is that anything like a perpetual motion machine?

Hp is simply a function of torque and rpm. Those old F1 cars were running really high boost to get the torque, and lots of revs to turn that into work(hp) I'm sure every aspect of them were as engineered as you could get for back then. But if there were some silver bullet to do with back pressure to boost ratio, I figure every manufacturer(and us) would be doing it that way today.

j9fd3s

iTrader: (3)

i've got no first hand data on this, but the simulation says that most turbo setups have more boost than backpressure at some rpm points. the big difference with F1 is that they set it up to have more boost than backpressure at peak revs. i think if you did this on a street car, power would be laggy at best.

its not magic, its just sizing the turbo for max HP