dan83590

I am wondering....

This is just a hypothetical question. What happens when a turbo is used beyond its efficiency? Why is it not possible just to keep pushing a turbocharger? What happens to airflow? Why does a turbo become inefficient at very high RPM?

I'm not interested in the answer 'cus you'l blow yer turbo up mate innit'.

wayno

iTrader: (3)

You'll find an optimal point of boost depending on your turbo and the cooling capabilities of your intercooler, but it will always be the more boost or the harder you work your turbo then the more superheated the air will be.

although i'm sure someone can better this

dan83590

So then say I had the kit to cool the super heated air... What then?

DeanF

iTrader: (1)

There is a limit to the amount of physical air the compressor/Turbo can produce.
So if you could cool the air super efficiently, The Turbo would work more efficiently, But it would still not remove it's physical limitations..

What about the engine ?? Now getting this super cool air & producing a bigger Bang per buck !! Thats now getting Hotter

As im sure you well know.Movement/force = energy, which in turn Generates that Ole fella called heat

Dean

Alan Jeffery

A normally aspirated engine will ingest and expel a specific amount of air relative to swept volume and design factors controlling air flow. To be efficient, an exhaust gas driven supercharger (turbo) has to be able to exceed the amount of intake air and exhaust gas flow that the engine can flow on its own at peak revs to be able to produce boost. It's the reason why some 2.5 engines can give less actual power using the same turbo than some 2.0 ones.
Turbos are a bit like Love, somewhere out there is the right one for you.

dan83590

So it's the CC of a block together with the size of the turbo that determines where inefficiency is reached?

Alan Jeffery

Close.. Bear in mind a turbo can be inefficient at both ends!

dan83590

Think I'm with you...!

An block can only expel a certain amount of air. The exhaust wheel can only turn as fast as the expelled air allows. At the same time the block is sucking air. The compressor wheel has to turn fast enough to push air and exceed the amount of air being taken in by the block and pistons.

At a given point the exhaust gas cannot spin the turbo exhaust wheel fast enough to exceed the air going in at the other side.

How's that???

joz8968

iTrader: (13)

How to choose the right turbo - DSM Forums

http://www.ztechz.net/id7.html

joz8968

iTrader: (13)

This turbo choice calc is interesting. If nothing else, it seems to pretty accurately predict BHP for the known rpm limit/boost level when you remap your car...

Ray Hall Turbocharging - Java turbo matching calculator

dunx

iTrader: (3)

I have a ancient book from the USofA that gives all the calcs., it's all down to the "turbo efficiency" map which manufacturers published.

On large (competition) diesels they would run two turbos one to heat the air up, then it gets intercooled, then compress it again !

dunx

P.S. that's a cool link, it says mine should make 500 bhp at 2 bar....

joz8968

iTrader: (13)

Yeah, the turbo's A/R ratio determines what sized engine any given turbo is suitable for, with respect to working within the turbo's most effecient "sweet-spot" part of the map.

pat

Where to start, LOL

Right. Let's first define efficiency....

A turbocharger's compressor is a radial flow (centrifugal) compressor. It works by imparting kinetic energy to a flow of fluid (in our case air), and then slowing the fluid back down through a diffuser, thereby raising its static pressure. If all of this went perfectly, we would have isentropically compressed the working fluid and we'de have a temperature rise. This temperature rise can be predicted theoretically for a perfect working fluid : Tout=Tin*(pout/pin)^(gamma-1/gamma)...

OK, an example. If we start with air at atmospheric pressure (1013mbar) and at 20 degrees C, then compress it to 2026mbar, ie double it's pressure, what would its temperature be ? Running the numbers through the equation, and using gamma=1.4 for air, we get Tout=(20+273)(2026/1013)^(0.4/1.4) = 357.17 Kelvin or 84 degrees C. The air has heated up 64 degrees C. So far so good...

Now spin up a real compressor and get it to discharge air at 2026mbar, and measure the temperature. We find that it's discharge temperature is 104 degrees. This is more than predicted, the rise is 84 degrees rather than the predicted 64. Obviously something has gone a little "wrong" along the way! This is merely reminding you that a turbo is not a perfect machine working on a perfect working fluid. It has heated the real air up more than it should; this means that a) the process is not reversible and b) it requires more power to be delivered to the compressor by the turbine in order to add this additional heat to the air. A compressor's efficiency is the reversible compressor work divided by the actual work; since the flow is identical for the two terms the flow rate cancels and leaves just the temperature rises, which are 64 and 84, respectively. This compressor, operating at those conditions, is (64/84) * 100% = 76.2% efficient.

Turbo manufacturers will run their compressors at various different pressure ratios and flow rates, measuring the efficiency as they go, and build up a 3 dimensional map of compressor efficiency as a function of normalised mass flow rate and pressure ratio. We often see these compressor maps and articles linked to above explain how to interpret and use them. What they don't say is why the efficiency changes....

Remember that a turbocharger's compressor is a radial flow compressor and that kinetic energy is imparted to the fluid as it passes through it, ie the air is accelerated. The inbound air will have some amount of axial speed (it is moving already) and it may have some radial speed or angular velocity. Whatever it is, it won't be the same as the blade tips that it's about to encounter! This leads us to an important ratio, U/C, which compares these velocities. If the air is moving really slowly then a) it's going to get an almighty bashing by the leading edge of the blade (try stepping out in front of a plane travelling at mach 0.6 and see how much that hurts!), furthermore, there will be severe turbulence of the trailing side of the blade (it is quite literally trying to outrun the air). This is obviously not a good combination, and it should come as no surprise that the compressor isn't very efficient like this! It spends most of its time bashing into "stationary" air. The air tearing away from the back edge of the blade narrows the apparent passage size and causes flow perturbations which can (do) lead to compressor surge. This is why making inlet pipes smaller helps reduce surge.

Let's speed the air up a little, but just axially. Air is rushing into the inlet quickly now. U/C becomes more favourable because although the air isn't spinning round trying to match the blades at the inlet, the blades are swept back, looking into the wheel, so as the air strikes the blade, it's already wanting to continue, and it's like it just got onto a mini-slide. There's a little impact, and then off it goes... much nicer, much less turbulence etc. Also, where previously the blades were literally trying to outrun the air on the back face, now there's more air coming in to fill the void just left. No big flow stagnations and a nice wide flow passage means it's making good use of the available passageways.

To make things even more pleasant for the air, let's give it a little bit of swirl too, so it's already rotating (albeit more slowly) in the same direction as the compressor wheel. Once again, the "shock" of it encountering the blades is reduced and suddenly we're actually spending most of the time just coaxing the air down the air mini-slide, rather than mashing it up and generally not doing anything useful. This point represents something close to peak efficiency. It's as good as it's likely to get!

As we spin the compressor faster, the inbound air needs to be faster to maintain an acceptably efficient U/C, and that gets progressively more difficult, until you encounter boundary conditions where some of the flow goes marginally supersonic (the local mach number will exceed 1.0 - note this does not mean it exceeds 344m/s, it could be less, or more, since the working fluid is no longer at standard temperature and pressure) and Bad Things (TM) start happening.

Now that we understand what efficiency is, and why it changes with operating conditions, we can look more closely at the question at hand... what happens when we get beyond the peak efficiency island.

The first thing that happens is that, errr, it gets less efficient The airflow velocity is no longer well matched to the compressor angular velocity and the flow is somewhat sub-optimal. This means that a) the air comes out hotter (but we can cool it again if we have a good intercooler) and b) the exhaust back pressure goes up because that extra heat has to come from somewhere... that somewhere is the turbine wheel and it gets its energy from the exhaust gas. Only an increase in turbine mass flow rate or an increase in enthalpy change across that stage would increase its power output, and our only real control is pressure, via the wastegate (although that will also increase mass flow at the same time). With this the volumetric efficiency of the engine falls, the piston is having to work against ever increasing exhaust pressure and the chamber still has finite volume, we can never expel all the exhaust gas, there will be some left, and the higher the exhaust back pressure, the higher the remaining exhaust gas in the cylinder at the beginning of the next cycle, the less room there is for fresh air. All in all then, things are getting pretty uncomfortable. And we've not reached choke flow yet.

As indicated above, there will come a point where the compressor cannot flow any more air. This is typically a mach limit problem, where the air just cannot flow through a given passage any faster, the only way to increase the mass flow rate is to make the passage bigger (bigger turbo, hmmmm, yeah that normally works) or to increase air density (cold damp days, yep, turbo cars go really well on cold damp days!). When the flow chokes that's it, all the air you're gonna get, until you improve the density or the size of the passages. You can spin the compressor faster but you get the same airflow. This is where it gets really nasty. The efficiency is falling like a lead balloon, the power requirement of the compressor is going up drastically to cope with the inefficiency, and all this is doing is heating up the air more. So. Very very hot air out of the compressor, with enormous exhaust back pressure needed to provide the power. In some cases you may even see a reduction in airflow because the engine is so choked up it's almost going backwards (exhaust pressure far higher than intake). A very bad situation to be in, and best avoided!

Typically a turbo will tell you that it's getting closer to its flow limit as the discharge pressure falls. This is normally the result of the back pressure coming up and blowing the wastegate open against the force of the spring. It's a little hint, ignored at one's own peril. You can ramp the wastegate duty up to try to get the full flow from the compressor, but it is wise to do so cautiously! If it doesn't hold boost easily, it's because something's not quite right...

Hopefully this answers most of the questions the OP raised... if not, feel free to ask more detailed questions

Cheers,

Pat.

joz8968

iTrader: (13)

LOL, I was hoping this thread would 'bait and hook' a 'white paper'-like explanation from one of the "big boys"... and Pat invariably never fails to deliver.

I LOVE reading these.

But phew, er... exhausting stuff! (Sorry lol).

dan83590

Phewy... Now that's an answer.

DeanF

iTrader: (1)

If you can cool this Baby

Your onto a WInner for sure

http://www.max-boost.co.uk/max-boost.../big_turbo.jpg

Dean

dan83590

Do you think that one would cause lag on a 2.0 wrx?

DeanF

iTrader: (1)

A good Mapper would be able to sort that out
More of a problem will be trying to see over it !!

Dean