Fluid properties across a throttle

[Andy330hp]

We got a little off-topic today, and a discussion arose about what happens to the temperature of air as it crosses an engine's throttle. For now, assume it's a MPFI, so fuel is not involved. Heywood uses compressible flow equations with a discharge coefficient to calculate mass flow. This suggests that compressible flow equations could also be used to predict the temperature. However, if you look at a fluids textbook, these equations were originally derived using the assumtion of ISENTROPIC flow, which I would think that a throttle is not. Of course, this is why the discharge coefficient term is needed in the first place. But, is it usable to figure out the temperature?
From another standpoint, my thermo book says that "throttling" processes (not really defined) are constant enthalpy, which for air means that the temperature across a throttle is CONSTANT! Thus, unless the discharge coefficient is zero, it doesn't seem like you could EVER use compressible flow equations to calculate the downstream temperature correctly. What am I not understanding?

 

[TZSiR]

I was curious about this about three weeks ago when I was porting a friends TB. I figured, by looking at it only, that the flow on the bottom side of the plate would be high velocity with low pressure, and the flow over the top would be lower velocity with higher pressure. I didn't think about temp seeing how there's two equal portions of air entering through the same size openings and entering a larger chamber then the TB diameter or the intake tubing. So yeah I think the temperature for the most part remains the same, but I know that the velocity slows down a bit.

 

[GregLocock]

Flow across a throttle is usually idealised as isenthalpic, adiabatic and non reversible.

Check out Rogers and Mayhew sec 10.3 for a worked example. No discharge coefficients there, and frankly I can't see compressible flow applying in an engine intake. I don't see the flow exceeding M0.3

Cheers

Greg Locock

 

[Andy330hp]

One clarification/addition: I am also (actually mostly) confused by what happens during part throttle, when the velocity will most definitely be at or near choking conditions. What happens then?

 

[Chumley]

Andy330hp: maybe I'm missing something here but in the same way that the cylinder temperature must go up when
you create compression in the combustion chamber of a diesel
engine, the air temperature must drop when the pressure drops after it passes the throttle body and enters the intake manifold plenum in a naturally aspirated spark ignition engine at all throttle positions other than wide open throttle, assuming the throttle body is sized correctly.

My aluminum intake manifolds are always colder than ambient
temperature after driving several miles at part throttle.
MPFI (hvap of the fuel) has something to do with that but so does the pressure drop.

Carburetor icing in the olden days comes to mind.

Chumley

 

[strokersix]

Chumley,

Re-read Greg's post "Flow across a throttle is usually idealised as isenthalpic, adiabatic and non reversible."

For an idealized perfect gas system, which air at normal temperatures and pressures is not far from, the temperature on either side of a throttle is identical. The extent the temperatures deviate is the Joule-Thomson effect and is probably on the order of only a couple degrees.

Compression in a combustion chamber is an entirely different process. It is more or less reversible. In other words adding work to compress and heat the gas can be reversed.

Expanding a gas across a throttle does not produce any work. Expanding a gas in a cylinder with a moving piston does produce work and that is the difference in the most basic terms.

Your cold intake manifold and carburetor icing is due to evaporative cooling of the fuel, not expansion of the air.

Now my Thermodynamics is a bit rusty so I welcome any corrections to my statements.

Mike

 

[ProgressiveRacing]

Greg wrote,

"Check out Rogers and Mayhew sec 10.3 for a worked example. No discharge coefficients there, and frankly I can't see compressible flow applying in an engine intake. I don't see the flow exceeding M0.3"

Please excuse my ignorance Greg but who are Rogers and Mayhew? I'm assuming that you are referring to a book? If so, do you have an ISBN# or perhaps the title? I'd like to read more on the subject... thanks.

WRT, aluminum MPFI manifolds. I've never seen ambient temperatures in one of these manifolds. Wet manifolds (carbureted etc.), I can see there being a lower manifold temp due to fuel vaporization, but I can't see the comparison here.
Many guys looking for hp above everything else use a thermal barrier coating to try and lower manifold temperatures and increase charge density.

On the subject of throttle bodies and throttle sizes... how does one go about sizing a throttle body for an application. I have hardly ever seen a case where going to a larger-than-stock throttle body on a V8 hurt performance... so it has made me wonder.
Thanks,
Allen

 

[GregLocock]

Rogers and Mayhew, "Engineering Thermodynamics", 4th Edition,
Longman, 1992.

Not my favourite book, but a widely used thermodynamic reference. Beware, there is only really one or two pages on throttles in there.

Large throttle bodies are a disadvantage for part throttle performance, ie fuel consumption, emissions, and transients. A race car won't need to worry about that lot much. Obviously it was more important in the days of carbs.



Cheers

Greg Locock

 

[Andy330hp]

Chumley and greg are citing something very similar to what I found in my thermo book. But if flow across an engine's throttle really is isenthalpic and non-reversible, then why can Heywood use compressible (read: reversible or ISENTROPIC) equations to calculate flow through it? And if he can, then why won't these equations predict zero temperature drop? Please note that no work is done on a fluid as it passes through a converging-diverging nozzle, but temperature changes are definitely found. We don't have a ton of thermocouples, or else we'd test this. bummer.

As an answer to Progressive Racing, what we have been using is a peak average mach no of somewhere between 0.2 and 0.3. By average, I mean that we don't look at individual spikes of velocity or anything like that. Just a quick calc, engine disp/VE/RPM thing (or whatever data you have available to figure flow rate), then find the area (don't forget the effect of shaft diameter) such that the velocity will be correct. It's a diminishing returns thing: going from M=0.3 to M=0.2 may still give you a 2-3% power increase (MAYBE, others things are at stake as well...), but going from M=0.2 to M=0.1 will give you almost nothing. That's my limited experience talking, others here may be able to provide more insight.

 

[ProgressiveRacing]

Andy and Greg,

Thanks for the info guys. Andy, your explanation makes a great deal of sense to me.

Thanks agin,
Allen

 

[Chumley]

Strokersix: may be my thinking is upside down but isn't the engine performing work as it creates the partial vacuum
in the intake manifold at partial throttle which leads to the term "pumping losses?" As far as my limited understanding goes, one of the reasons diesel engines are more efficient than gas engines is because
they limit pumping losses (no throttle body) so that at all power settings diesel engines don't have to do additional work that gas engines do at all power settings other than at WOT.

Try breathing through a straw. It's a substantial amount of work that isn't a thermodynamic free lunch nor is it idealized or perfect when it comes to an engine!

Chumley

 

[strokersix]

Chumley:

You are absolutely correct about pumping losses for throttled (spark ignition) versus unthrottled (compression ignition) engines.

Interesting argument about the work associated with pumping loss. I'll think about that.

I'll stick by my statements if we are just considering flow past an orifice or throttling process. Isolating the throttle body and taking temperature measurements above and below the throttle should show very little difference, assuming it is dry flow as in a port injected engine. Now, I must admit, I have not done this. My statements are based on thermodynamic principles and not actual experience. Perhaps in an actual engine system there is more going on than a simple throttled flow approximation can describe?

Mike

 

[ivymike]

air flowing through an orifice is doing work - you have a pressure difference, thus a force on each molecule; each molecule travels as it moves through the orifice; forceXdistance=work, etc.

I don't remember how the calcs go, and I don't have a thermo book handy, but if I remember correctly -
* a C02 bottle get cold as it exhausts the gas inside to atmosphere. The gas coming out would be hotter than ambient if you caught it
* An evacuated bottle gets hot as it fills

 

[racer1]

Racer1

I recommend looking at cengel and boles, about Joule-Thomson coefficient. For a constant enthalpy process the the temperature can cool, stay the same or increase in temperature through a throttle. Depending whether its is left or right of the inversion line on Temperature pressure diagram.

 

[facty]

Greg Locock - "and frankly I can't see compressible flow applying in an engine intake"

What about intake ramming/tuning. I recommend reading chapter 2 of this book on the design of engines by G.P. Blair:

http://www.amazon.com/exec/obidos/A...49874/sr=11-1/ref=sr_11_1/104-2368532-0342369

http://www.sae.org/servlets/productDetail?PROD_TYP=BOOK&PROD_CD=R-186

This chapter describes all the gas dynamics and thermodynamics relevant to air flow in engine ducting.