Turbine AR, trim, cams and airflow- what does it all mean?

[Steve]

Hi,

I have discovered that there is plenty of info on compressor sizing, compressor maps etc, but there seems to be a severe lack of definitive info on turbines, and the effect of different sizes and trims and the implications of a particular choice.

Specifically I was interested if anybody had any info on:

Head and manifold flow characteristics,

Cam sizing and overlap,

wastegate sizing,

trim of the wheel,

turbine AR vs airflow and back pressure,

and, selecting systems that work together.

I realise, as probably alot do, that the smaller the AR ratio the sooner the boost builds, the greater the back pressure - which equals less top end and milder cams needed.

But how specifically can you determine for example if you have a 2.5L engine, what turbine size performs best at what rev band (ie from what rpm to what rpm) is there a compressor map equivelant for turbines? - what cams should be used and how much effect changing from say a 0.61 AR to 0.63 to 0.78 etc? optimum overlap for broad power band, the effect of different ovelap and duration on turbine sizing? what is too big and too small? and then how would you determine the same for 2L or 3L?

There must be some sort of quantifiable interelation between airflow, timing, overlap, turbine AR - or am I dreaming? Is it just a case of hit and miss?

Any ideas would be really appreciated, as this has been bugging me for a while. A fair few questions there, but I am hoping to be able to tie the air going in to the air going out a bit better, and hopefully understand better what compromises must be made to produce a given effect:D

Cheers

Steve

[dojobi]

Don't you hate it when you spend a fair bit of time writing up a post and noone answers it?

Anyway, I'm interested in hearing some responses to this thread, so here's a free bump. Sorry, I can't help you out.

[Steve]

Yeah, a bit frustrating, but I wasnt expecting a landslide, because there seems to be so little info out there on the interelationships between these things.

Thanks for the bump:)

[Roy]

From my experience (limited, and havent researched all that heavily) the info your looking for will not be found in the one text.

Try your local Tafe library or motoring bookshop, and see what books they have to offer. As you have probably found the average book doesnt give you too much detail, only broad brush strokes.

There are a couple of old school books that may help out. I think the book series name is Alex Walordy' guid to etc etc.

One is called something like Turbochargers and Nitrous Oxide Systems, and then there are some generic Small block chevy/ford books that have quite a bit of detail on head flow.

I have a crap load of Racecar Engineering (about 6 years worth) which are a pretty technical mag, but generally ignore turbocharging as only indy cars and WRC really are covered.

Hope that helps.

[Steve]

Thanks Roy, its all good. I have read a bit, as you say its mainly broad stuff, not a lot of detail. Heaps on the science of getting the air in.

[Steve]

anybody ese out there got 0.02 they wouldnt mind throwing in?

[raist60]

K Heres muh bit.

There all very general questions so its hard to know where to begin. SO ill just ramble on for a bit.

Turbine trims are figure that describes the pitch of the fins, high values like 90 are considered low flow high spool. Lower values like 80 are considered high flow low spool. With regards to question as to how to match turbine a/r and trim to power figures. There is formulas for working it out, but majority of the time the turbine side is fairly evenly matched with the expected max flow of the compressor side in most pre-built turbos. If you go and mix and match wheels then youd have to sit down and do the math's but why deviate from something that turbo engineers have obviously tried and tested. One thing that ive noticed is that turbo theory doesnt always become fact in real world scenarios, a turbo i had built should have flowed well according to theory, but when put on the car was lucky to push a rb25det to 190rwkw. Simply chaning the turbine housing got me an extra 60rwkw.

Wastegate sizing, im not sure if there is any formula for working out minimum size, but obviously it is a component that doesnt become detrimental* when it gets too big. The bigger the better IMO.

Head / manifold design, this is where i dont have much knowledge but I do think that turbo engines are much more forgiving of slight design / construction mistakes than a NA car. In a NA car any stuff ups will make the car work alot harder to get the air down into the cylinder using vacuum, where as turbo crams it down whichever way it can. I think that even distribution is important (std plenum), i think that aftermarket front entry plenums cant be brilliant for equal distribution at higher rpm / flows unless of course they are using internal baffles / channels.

Cams i dont know enough about to comment.

As for the whole interelation. Well there definitely is but its still always a compromise, every part has a specific design goal whether it be responsiveness, outright power, torque... you cant have it all.

[Warpspeed]

There is not a lot of information around, and those that really do know will not tell you anyway. Others are scared to have a go because others may later prove them to be wrong.

Anyhow, I am game to give it a try.

The first thing is that the exhaust turbine needs to be speed and energy matched to the compressor so that the available exhaust energy can best be used to create boost.

Speed and energy matching also brings in the question of exhaust back pressure, and at what flow you want the most efficient operating point. This all assumes operation below the wastegate opening point.

Now given a compressor flow map, you can figure out that you need a certain shaft RPM to generate a certain desired boost level. The biggest factor here is compressor wheel diameter. So you come up with a particular shaft RPM that you must reach with as little exhaust back pressure restriction as possible.

Now you also need to generate enough shaft torque to turn that compressor wheel, in fact you need surplus torque to accelerate it quickly as well. The amount of exhaust turbine shaft horsepower required to do this might surprise you.

Now given a certain volume of exhaust gas out of the engine to work with, there are several ways of attacking this problem.

You might for example accelerate the gas through a small a/r housing to very high velocity, and use that to drive a large diameter exhaust wheel at high tip speed. Now a small a/r sounds bad, but remember the wheel is big, and the nozzle area will also be big for a given a/r.

The torque developed by the turbine will be high because the wheel diameter is large, but also wheel weight and inertia will be up as well. So you have more turbine power and more weight, these tend to cancel to a certain extent.

Or you can go the other way. Use a smaller exhaust wheel diameter with lower tip speed. You can then use a larger a/r housing, but the nozzle size will not be as big as you might think because the whole exhaust housing is a smaller diameter.

So you get a whole range of wheel sizes and turbine a/r sizes that will sort of work. With the Garrett T3 and T4 range, people have been playing around with these combinations for over thirty years, and you can usually find something that works tolerably well.

Now Garrett have figured it all out, mainly because they now have computer software that can model exactly what is going on in the turbo. The new GT range of ball bearing turbos have the turbines and compressors very well matched for speed and energy. The trend is to smaller diameter turbines with larger a/r.

Also the turbine wheel shape has dramatically changed. The old turbines are large diameter, and skinny, with a small exit eye. Also the blades had a lot of swirl near the exit.

Newer turbines look more like paddle wheels with the major and minor diameters nearer to being the same. Also the blades are straighter.

Another thing, the old turbos were available with a very wide range of exhaust housing a/r. Newer GT turbos usually have only three housings available in each turbine size. Probably the middle size being the most efficient.

So if you order your new GT Garrett turbo with a compressor to suit your needs, the turbine will be exactly right for that compressor. If it does not work, you need a different sized turbo, not some extreme sized exhaust housing.

As for sizing the old T3 and T4 turbines, that depends on what you want it to do. The most efficient operating point is where you just reach the desired boost level just at the point just before the wastegate opens, this really applies to any turbo.

Once the wastegate opens you are throwing away energy, and the power required to drive the compressor all comes from high back pressure. So having a low boost threshold is going to cost you engine horsepower because of high exhaust back pressure.

Measure your exhaust manifold pressure before the turbo, and realize that every psi is costing you about 1% of engine power.

It is not uncommon to find exhaust pressure to be twice boost pressure, or even more. So measure it, and be prepared for a surprise.

[Warpspeed]

As far as wastegate size goes, it needs to be large enough to bypass all the exhaust the engine can create after the wastegate opens. So there are two variables.

If your red-line is 8,000 RPM, and your wastegate opens at 7,000 RPM it does not need to be large because roughly seven eights of the exhaust is going through the turbine, and one eighth through the wastegate.

On the other hand, if your wastegate opens at 4,000 RPM, at red-line roughly equal turbine and wastegate flows will be required. This is not strictly true, but you get the idea.

The same sized wastegate is going to be able to flow a lot more if the exhaust back pressure is 35 psi than if it is only 12 psi.

In order for the wastegate to properly control boost, the area of the control diaphragm needs to be large with respect to the area of the poppet valve area. Otherwise exhaust back pressure may just blow the wastegate open. This will reduce boost, but it is not really what you want. So do not be tempted to make the wastegate too large.

Lastly, a long soft spring in the wastegate will control boost better than a short stiff spring because it will vary more in length and open and close the valve further for the same variation in boost pressure.

As far as camshafts go, the starting point is knowing how much exhaust back pressure you have, with respect to boost pressure. The higher the exhaust back pressure, the less valve overlap you can get away with.

Do not be fooled into thinking that boost is going to blow all your mixture down the exhaust pipe in a turbo engine. If exhaust back pressure is 30 psi, and you run 12 psi boost, this obviously cannot happen. Fitting a big overlap cam to this same engine is not going to help you either.

On the other hand, if you run a massive top end only turbo, and an open dump pipe, you might find that you have 20 psi boost, and 18psi exhaust back pressure. A nice bit of overlap might gain you quite a bit of extra power. It will be very peaky though.

So the moral is, big turbo, big cam. Small turbo small cam.

The exhaust opening and inlet closing points decide where in the RPM range the cam is going to work best, just as it does with an n/a cam. I find a bit of extra duration on the exhaust lobes also helps top end power without killing the bottom end.

O/K guys thats more than enough bullshit from me for one session, so flame away.......

[Steve]

Some great replies there, thanks heaps.

Hope nobody minds but it raises a some more questions....

with regard to wastegate size;

Is there any formula for working out flow for given valve diameter, travel, pressure? Not when connected to a manifold, but in a sterile world, similar to the testing of turbos? Do manufacturers actually detail wastegate movement range and flow, rather than just valve dimensions? Or is there a general rule of thumb used?

with regards to turbine size/dimensions;

is there a way to determine what sort of response a turbine wheel/AR combo will give at a specific gas flow? Or do we just match the compressor, fit the turbo (utilising the mid sized housing as a start point) and then experiment with different size housing until a desired result is achieved?

with regard to cams;

how do we determine the amount of overlap per psi back pressure that is acceptable before performance is reduced, and how do we determine how much more duration would be appropriate for the exhaust than the inlet?

Sorry if these are a bit on the vague side again, just seems that this side of things seems to be very overlooked when putting a package together.

Really appreciate the time taken for the replies so far, please keep em coming if you will

cheers

Steve

[Warpspeed]

Hmmm.

Wastegate size, could probably be estimated by calculation of flow area, if you know the pressure drop across the wastegate valve. You could actually measure one on a flow-bench, or estimate the flow coefficient from a similar sized exhaust valve in an engine. Or you could just guess.

If you then know that your wastegate flows the same as say a 30mm plain hole, for example, you could then plug this into a formula for calculating the flow/pressure drop across a 30mm orifice plate. I have this formula here somewhere, but need to go and look for it.

I have never tried any of this this myself. Another figure that you might find useful is a flow in the induction system of 1.5 CFM per horsepower, and an exhaust flow of 2.2 CFM per horsepower (at atmospheric pressure).

As far as response goes, you cannot really do much about it, because it depends on the exhaust energy available, and rotating inertia. But boost threshold can be moved around quite a lot. Exhaust manifold design can also help a lot with this.

Valve overlap versus back pressure is also difficult. The relationship above just referred to a general principle without being too precise about it.

My approach is to use one of those engine dyno software simulation packages to come up with a good cam combination. If there is nothing really extreme about the engine, it will not be very sensitive to valve timing anyhow. If it is cam sensitive, there is something very wrong with the combination.

I like to run minimum overlap that I can without losing lots of power. On my own car I have been experimenting for a couple of years and keep coming back to the same valve timing that works very well. It is 1.6 Litre DOHC 4 valve engine running 15Psi boost, so is probably fairly typical.

Inlet cam 245 duration, opens 4BTDC, closes 61ABDC

Exhaust 254 duration, opens 57BBDC, closes 17ATDC

I am not saying this is the perfect cam for everyone, but it does illustrate a few features. First the exhaust cam is fitted to what would be stock n/a lobe centre angle. Standard timing for a 254 cam is 17/57 57/17, no surprise here.

The inlet cam has nine degrees less duration and is also retarded a fair bit. Both of these reduce the valve overlap to only 21 degrees. A pair of stock 254 cams would have 34 degrees of overlap which is far too much. The overlap can be reduced by late inlet opening, which does not seem to hurt things at all.

Another interesting feature here about overlap is that most of it is on the exhaust side of TDC. This works best where exhaust back pressure is higher than induction pressure on any engine.

Most stock n/a cams are like that as well, with typically about five degrees more overlap on the exhaust side. Stock engines usually have quiet and restrictive exhaust systems.

Typical n/a sports racing cams that run extractors and a low back pressure exhaust system seem to like split overlap.

Mechanically supercharged engines always have far higher induction pressure than exhaust back pressure, and run best with most of the overlap on the induction side of TDC.

The clue here with your turbo engine is to keep most of the overlap on the exhaust side, unless the turbo is truly huge, then run split overlap if you can get the exhaust back pressure down that far.

It is a commonly known trick with the GTR to retard the exhaust cam a few degrees, even though the exhaust duration is only 236 degrees. This makes exhaust opening very late, but it still goes a bit better at the extreme top end.

[Steve]

Thanks again warpspeed, certainly plenty to think about there. Do you by chance know what the standard duration and overlap is for RB25?

Thanks again

Steve

[EnricoPalazzo]

Steve: Once ive finished Croky Bells Maximum boost ill lend it to u if u want, it might interest u!

[Steve]

Thanks Enrico, very kind offer.

I have a copy of supercharged by the same guy, good read. I was going to buy maximum boost too, but decided it was a bit much ($100) when a large majority of the info is covered in supercharged.

Another good one (borrowed off a mate for a while) is 21st century performance - covers some good stuff. Might have to buy a copy of that too one day

[riggaP]

This is a great thread.

Enrico & Steve: can you guys post the full details of those books? Cheers.

[EnricoPalazzo]

Steve: the offers there if u want it mate!

riggaP: Title: Maximum Boost

Author: Corkey Bell

Can purchase online from autospeed i think, however i got mine at Borders, they ordered it in for me. Around 80 bucks it turned out to be, cheaper then the net.

[Steve]

Thanks Enrico, give us a yell when you are finnished, I am sure I'll find something interesting in it:)

riggaP: 21st Century Performance, buggered if I can remeber the author, but someone else here might know his name with luck?

[BOOSTD]

21st Century- Julian Eger

[riggaP]

Enrico, Steve, BOOSTD:

I was discussing a few of these issues on the weekend with a friend. I still dont know enough to comment so keep the replies coming.

[Warpspeed]

Steve, The RB25DET has 240 duration cams, and the early non VVT RB25DE have 248 degree cams. I have no idea what the VVT cams are, I would love to know.

R33 RB25DET inlet cam 240 duration 7.8mm lift, lobe centre angle 120 degrees, inlet opens TDC (zero degrees) closes 60 ABDC.

Exhaust cam 240 duration 7.8mm lift, lobe centre angle 117 degrees, exhaust opens 57 BBDC, closes 3 ATDC. Overlap 3 degrees.

R34 RB25DET inlet cam 240 duration 7.8mm lift, lobe centre angle 100 degrees, inlet opens 20 BTDC, closes 40 ABDC. Exhaust cam 240 duration 7.8mm lift, lobe centre angle 117 degrees, exhaust opens 57 BBDC, closes 3 ATDC. Overlap 23 degrees.

So they have advanced the inlet cam a fair way on the R34.