Diesels: volumetric efficiency versus boost 1

[JoeFrickinFriday]

I'm trying to develop a coarse model of volumetric efficiency (VE) as a function of intake manifold pressure for a heavy-duty diesel engine (~15L). For this discussion, I'm defining VE as the in-cylinder density at intake-valve closure divided by the density in the intake manifold.

I have data from motoring tests that I can use to develop a relationship between RPM and VE, and this is consistent with what I see in numerous technical sources (Heywood, etc.), i.e. the peak value is 86% @1300RPM, and falls off at lower or higher RPM. But I can't get data to help me understand how boost affects VE: when we run the engine with load/boost, EGR is active, and so if I use the air flow measured by the intake air flow meter for my calculations, I get very low values for VE.

So far the only useful info I've been able to find on boost vs VE is from this thread on the Ford Truck Enthusiasts discussion board. Check out the third plot, "Scale Factor for Volumetric Efficiency VE(%) vs BP (psi)". AIUI, this plot means (for example) that for any given RPM, if you run with intake MAP at 3 bar absolute (30 psi on that plot), you won't get 3X as much air into the cylinders as you would at 1 bar absolute (0 psi on that plot); you'll only get (3 x 90.5% =) 2.715 times as much air.

Seems plausible, but I'm having a difficult time finding a citable reference that backs this up. I've got Heywood, but didn't see anything relevant in there Can someone point me toward a published source (book, SAE paper, whatever) that would confirm this behavior?

Thanks...

 

[turbomotor]

Hi Joe, It sounds like you are doing a lot of literature investigation to get where you want to be. My suggestions for supporting documents include some relatively old work, but the actual filling and emptying of a cylinder hasn't changed much over the years. For an SAE paper, try Dave Alfano's 1986 paper "Turbocharger Applications", SAE 862051. For textbook material, the best source that I can recommend is Watson & Janota "Turbocharging the Internal Combustion Engine" MacMillan 1982 and reprinted 1984. (You will find this book to be a bit pricey these days.) And an absolute must is page 163 of Taylor Volume 1 "The Internal Combustion Engine in Theory and Practice" MIT Press 1977 (for paperback edition).

In my understanding (this may be much less helpful than the works noted above), the V.E. is impacted more by the delta pressure across the engine than by the manifold absolute pressure. Highly waste-gated turbochargers may improve transient response, but they negatively impact the engine delta P and the V.E. and the pumping losses, thus negatively impacting the steady state performance. A high degree of intercooling may negatively impact V.E. because the additional heating of the intake charge as it flows through the valve ports will reduce the density of the air trapped in the cylinder versus the air in the intake manifold. Anything that changes the air temperature during the fill cycle, such as fuel evaporation, will impact V.E. And finally, valve size, timing and event overlap are all going to play a large role in determining the speed at which the fill process is most effective.

Best of luck with your project!

Dick V.

 

[Lou Scannon]

By your definition, if you put a cork in the exhaust, there'd be zero net flow into the cylinder (from the intake manifold), but the VE would be close to 1.
A standard definition considers the cylinder to be a pump, and the VE is the ratio of the actual net mass flow through the cylinder in one cycle, divided by the product of the intake manifold charge density and the cylinder's swept volume.

"Schiefgehen wird, was schiefgehen kann" - das Murphygesetz

 

[JoeFrickinFriday]

By your definition, if you put a cork in the exhaust, there'd be zero net flow into the cylinder (from the intake manifold), but the VE would be close to 1.
A standard definition considers the cylinder to be a pump, and the VE is the ratio of the actual net mass flow through the cylinder in one cycle, divided by the product of the intake manifold charge density and the cylinder's swept volume.

You're quite right. I expressed myself poorly. I'm actually calculating it exactly the way you describe; I just wanted to differentiate this definition from the other definition, which always uses STP density instead of intake manifold density (which may be quite a bit higher due to boost).

Turbomotor, thanks for the cites. I will dig up the SAE paper you mentioned, and I will look up a colleague who has Taylor in his collection.

 

[SomptingGuy]

That's an interesting metric to want to measure. (The standard definition of) VE is normally the holy grail for N/A engines, kinda irrelevant when you're pumping the intake up to whatever pressure you want.

You could easily compute your metric from any of the 1D engine simulations out there. It might even be possible to nominate some point in the high pressure part of the intake as your reference point for VE, so it comes out by default.

Steve

 

[JoeFrickinFriday]

That's an interesting metric to want to measure. (The standard definition of) VE is normally the holy grail for N/A engines, kinda irrelevant when you're pumping the intake up to whatever pressure you want.

You could easily compute your metric from any of the 1D engine simulations out there. It might even be possible to nominate some point in the high pressure part of the intake as your reference point for VE, so it comes out by default.

The whole thing is weird, I know. The end goal is to be able to estimate cylinder air mass at any speed/load point in the operating map. The problem is that the data I have available right now is limited. I have intake MAP for the entire operating map, but I don't have exhaust MAP data, so I can't use Pe/Pi as an input parameter. That's why I'm trying to come up with a way to estimate VE as a function of only intake MAP: so I can estimate the mass of air trapped in-cylinder. VE is just a waypoint on that path.

 

[BrianPetersen]

Exhaust manifold pressure, and therefore exhaust residual remaining in the cylinder after the exhaust stroke, can be estimated based on the operating conditions of the turbocharger. Knowing engine RPM and load would allow creation of not only a compressor pressure-ratio map but also an exhaust pressure-ratio map. It will not be accurate during transient conditions; whether that's meaningful depends on what exactly you are trying to do. The turbo manufacturer should have pressure-ratio maps, but these won't account for operation of wastegates or variable exhaust vane geometry or whatever other method of boost pressure regulation is in effect.

Intercooling, and related "thermal inertia" effects, will also have a time-based influence on this.

In this day and age of stringent emission controls, the OEMs routinely do what you are attempting to do, because accurate knowledge of airflow through the engine is crucial to achieving emissions targets. But they don't do it the way it appears that you are attempting to do it.

What is "the big picture" of what you are attempting to do, and why?

 

[Lou Scannon]

VE is normally the holy grail for N/A engines, kinda irrelevant when you're pumping the intake up to whatever pressure you want

I don't agree that it's irrelevant. Certainly boost trumps VE in the first instance, but a boosted engine benefits from improved VE in the same way a normally aspirated engine does.
As BrianPetersen hints, the turbocharger match can play a large role in the gas exchange process, i.e., how much work the piston has to do during gas exchange, and concomitantly, the amount of residual gases that remain in the cylinder at exhaust valve closing.

"Schiefgehen wird, was schiefgehen kann" - das Murphygesetz

 

[Marquis]

Defining VE in this manner (referencing In Man conditions) means you cant do meaningful development work on the intake system upstream of the intake manifold. I've seen folks use this method (usually very academic types who have no real world experience developing real engines) and then pat themselves on the back doing so.

To me it tells me very little.

I always reference ambient conditons-usually in the test cell. Yes you can get big VE numbers of 140-180 % but what of it?

http://www.auto-scape.com

Sideways To Victory!