Lets talk turbos

[eHoward]

Kind of like this:



I know how to measure VE. I just want to cheat.

This is how you do it.
Take your engine, or car to the appropriate dyno.
Not an inertial dyno, a real dyno, that holds the engine at a fixed RPM under load.
Hook up the intake side to a mass airflow meter.
Run the engine up to the specific RPM point that you're interested in, at full throttle.
Start the MAF readings and a stopwatch at the same time.
Stop the MAF readings and a stopwatch at the same time.
-----------------
Since you now know the volume of air ingested by the engine, and you should know the swept volume of the cylinders, and the amount of time, which corresponds to the RPM (revolutions per minute)..
You can now calculate the VE of the engine knowing that you don't have to worry about internal engine drag, drivetrain losses, etc.
Ain't I smart?
BTW, I think this is the way the OEMs do it.....Ask Cali Kid, the doorknob guy.

[Gixxer]

That's EXACTLY what I was talking about!
Even down to the dyno operator ready to pull the plug!

Here's a pic of my portable dyno at home-- :bootyshake:
-----------------
Merry Christmas!

[The Dark Side of Will]

Kind of like this:



I know how to measure VE. I just want to cheat.

Don't we all.
What's under all that mess anyway?

I'd like to find a relationship between VE and torque because a lot more people know their torque than have datalogged their MAF sensors.
Even people who have the facilities to datalog their MAF sensors still have to go through several mental backflips to come up with VE, as MAF reads lbs/min (or g/s) and has to go into a calculation with ambient pressure and ambient temperature in order to come out with volume, and then has to be further calculated with RPM and displacement to yield VE.

For a given engine, the relationship for VE and torque MUST be linear... however, the other factors that affect it are ignition advance, compression ratio and rod length, to name a few...

Speaking of which... there's an idea for engine controls that I've been bouncing around my empty head...
High temps in a forced induction engine make it especially prone to detonation, while cold temps allow noticeably more boost. What if, in a setup with ECM controlled boost, the algorithm for boost control was based on the product of manifold pressure and intake temp, rather than just manifold pressure. That way, on a cold day the computer will automatically turn up the boost and on a hot day will automatically turn it down.

[Gixxer]

For a given engine, the relationship for VE and torque MUST be linear...

Why does it have to be linear? Nothing else in the universe is.

however, the other factors that affect it are ignition advance, compression ratio and rod length, to name a few...

And there you are. Include also variability in combustion chamber burn (turbulence), friction rise...

Speaking of which... there's an idea for engine controls that I've been bouncing around my empty head...
High temps in a forced induction engine make it especially prone to detonation, while cold temps allow noticeably more boost. What if, in a setup with ECM controlled boost, the algorithm for boost control was based on the product of manifold pressure and intake temp, rather than just manifold pressure. That way, on a cold day the computer will automatically turn up the boost and on a hot day will automatically turn it down.

This already exists. Running around with this already in my Lotus. It uses a GM P4 series ECM, 1228708, 1989 vintage. You just need to enable the desired functionality with a suitably programmed Mem-cal. Unfortunately I think this particular unit only works with a 4 cyl. Only downside is that the MAT sensor mounted in the chargecooler outlet is slow reacting (IMO), so the total system has to be de-tuned a bit (only 18.5 lbs boost).
Cheers, I'm outa here for a while.

[The Dark Side of Will]

IIRC, 8708 is a 7730 with lowZ injector drivers. The 730 and 749 have been controlling boost for over a decade in Tubro Grands Prix, Syclones, Typhoons and Sunbirds. Lotus is really good with other people's engine management, though. They got 190 HP out of Toyota's 2ZZ-GE, from which Toyota was only able to extract 175ish.

However, rewriting the code to control boost via the product of MAP and IAT is a whole different beast than remapping an existing calibration to work with ported heads, bigger cams, bigger turbos, etc.

For a given engine, torque vs VE has to be linear because in a given engine, torque is directly proportional to cylinder pressure, which is in turn directly proportional to cylinder filling, which is nothing other than VE.

Now when comparing different engines, the torque/VE relationship is affected by the other factors I mentioned.

[The Dark Side of Will]

This already exists. Running around with this already in my Lotus. It uses a GM P4 series ECM, 1228708, 1989 vintage. You just need to enable the desired functionality with a suitably programmed Mem-cal. Unfortunately I think this particular unit only works with a 4 cyl. Only downside is that the MAT sensor mounted in the chargecooler outlet is slow reacting (IMO), so the total system has to be de-tuned a bit (only 18.5 lbs boost).
Cheers, I'm outa here for a while.

What mask are you using?

[eHoward]

I'll explain or edit later.

No. You are thinking about ideal compression, as if the compressor was 100% efficient. This is not possible. The highest you can get is about 80% today.

At 80F with a turbo out of its efficiency range like the GN turbo@55%, you're talking about a 212.782F raise in temp vs a 156.04F raise for something in its range like the garrett would be @75%. And that's not factoring in that the garrett would be having less ideal outlet temp because you could run less boost to make the same power.

T2 = T1(P2/P1)^.283

where T is temp R and P is pressure psia. That is ideal. What you were thinking.

To get actual, you divide what the rise in temp in degrees F by the compressor efficiency.

[Kohburn]

I'll explain or edit later.

No. You are thinking about ideal compression, as if the compressor was 100% efficient. This is not possible. The highest you can get is about 80% today.

At 80F with a turbo out of its efficiency range like the GN turbo@55%, you're talking about a 212.782F raise in temp vs a 156.04F raise for something in its range like the garrett would be @75%. And that's not factoring in that the garrett would be having less ideal outlet temp because you could run less boost to make the same power.

T2 = T1(P2/P1)^.283

where T is temp R and P is pressure psia. That is ideal. What you were thinking.

To get actual, you divide what the rise in temp in degrees F by the compressor efficiency.

now what about an upgraded GN turbo using the .83 exhaust side with an upgraded compressor side.

[eHoward]

Then it's not a GN turbo. and you probably don't have a compressor map so you can't quantify it.

now what about an upgraded GN turbo using the .83 exhaust side with an upgraded compressor side.

[bryson]

The TEC3 has an enrichment for manifold air temperature, and I'm sure the Motec does as well. I bet several stock ECUs have something similar.

[Kohburn]

Then it's not a GN turbo. and you probably don't have a compressor map so you can't quantify it.

now what about an upgraded GN turbo using the .83 exhaust side with an upgraded compressor side.

just curious since there are a lot of upgrades available for the GN turbo that replace the compressor side

[The Dark Side of Will]

The TEC3 has an enrichment for manifold air temperature, and I'm sure the Motec does as well. I bet several stock ECUs have something similar.

*Usually* factory algorithms are much more advanced than aftermarket algorithms. Yes, GM computers have fuel/ignition modifiers for air temp.

[eHoward]

I have some neat datalogs from a turbo engine, but it's a subaru engine.

Who wants to see MAF readings vs horsepower and injectory duty cycles?

[The Dark Side of Will]

show me the money

[eHoward]

Say it like you mean it!

show me the money

[Shaun41178(2)]

Say it like you mean it!

I LOVE BLACK PEOPLE!

[The Dark Side of Will]

This is very interesting...
The first graph shows that thermodynamic efficiency is higher between 3.8 and 6K than it is at lower RPM... however, the airflow/HP ratio starts to nose over about 3.5K.

The second graph shows basically the same phenomenon wrt BSFC... the engine becomes more efficient above 3.8K and doesn't require as much fuel to make HP. This graph also starts to nose over at 3.5K

The third graph links the first two together by showing that the turbo has just spooled at 3.5K. This shows that the engine is more efficient at higher initial cylinder pressures. One can also infer that below 3.5K the extra energy that isn't making it to the dyno is going into spooling the turbo.

[whipped]

I'm sorry, are you guys saying something? All I see is a glaring 95% duty on the injectors. :scratch:

[Kohburn]

The third graph links the first two together by showing that the turbo has just spooled at 3.5K. This shows that the engine is more efficient at higher initial cylinder pressures. One can also infer that below 3.5K the extra energy that isn't making it to the dyno is going into spooling the turbo.

yeah I was looking at that - during spool up is when its least efficient

[Shaun41178(2)]

I am going to go ahead and bump this back to the top. I need to do some more calculations.