JasonC SBB

Perhaps the smallest trim 2871 compressor (48 trim) is what the Potato ought to be - best for those who want 15-18 psi or 300-320 hp:
http://www.turbobygarrett.com/turbob...R_743347_1.htm

48 trim:


52 trim, like Savington's:


Here's the potato:

Savington

iTrader: (31)

Jason, the turbine maps don't show flow capability. I shouldn't have used the term "peak flow". They show the point at which the turbine begins to choke and build a pressure differential to actually perform work.

The corrected mass flow flatlines on the 2871R .86 at 21lb/min - but a 300whp Miata is moving ~33lb/min of air, and when you correct it (assuming EGTs of ~1600F and TIP of ~7psi) you end up at ~57lb/min - way more than the 21lb/min it can supposedly move.

If Garrett would release the real turbine maps, including the shaft RPMs at various corrected flow rates and pressure ratios, then we can start to get a better picture of how it all works.

Paging jkav.

JasonC SBB

Most of the (mass) flow is wastegated. The question is, what is the TIP (turbine inlet pressure) at the compressor's required shaft power and RPM, at the engine's mass flow rate and boost pressure...

The turbine plots however, do suggest that the GT28, even the 0.64, is sized for more flow than the GT25...

VitaminD

iTrader: (14)

250 is the goal(I cant see needing more in the near future), the car is a dedicated track car that I primarily run at Road Atlanta. 1.8 is getting freshened up and m-tuned rods installed.

The dyno sheets that I have found the 2560 seems to fall off in the upper RPMs, I will be spending a lot of time there.

I'm curious to see what a 2860 .64 a/r turbo dyno looks like, I hope someone post one.

JasonC SBB

On a 99 head the 2560 will do 265 hp. So the turbo can flow it.

Savington

iTrader: (31)

We'd need the full curves and RPM labels printed on the turbine plots to really know that for sure, though. If you know the peak mass flow of the compressor, then you know what shaft RPM and pressure ratio it occurs at. With that, we could compare to the shaft RPM of the turbine and look at the pressure ratio required to maintain that shaft RPM. Given compressor mass flow, EGT and turbine outlet pressure, we can calculate what the turbine inlet pressure will be. If the TIP exceeds the inlet manifold pressure, you've got a problem.

devin mac

iTrader: (2)

I've always wondered why getting our hands on the turbine maps is so damn hard, while compressor maps are a dime a dozen. Discussions like this remind me how frustrating it is to only be able to know half the story of what's going on, and half the information to really properly match up a turbo to one's goals.

JKav

iTrader: (1)

99% of people wouldn't know what to do with a turbine map. Still, it is unfortunate they're not out there. I think Borg's done the best job yet with their 'match-bot' thing, but even that lacks some turbine information to do a full, proper match.

Even with all the information on the map there are enough assumptions that go into a match that you always need to validate with a physical test on-vehicle. VE profile & true turbine inlet temp are big ones. Pressure losses through the comp inlet & CAC. Muffler system losses. On-engine pulse behavior at low expansion ratio has a HUGE effect on effective turbine efficiency that cannot be captured on a steady state turbine map (which all of them are).

Another thing is that the flow and speed on a turbine map is corrected to a higher inlet temp, so 1 lb/min of map turbine flow does not directly equate to 1 lb/min of map compressor flow.

Speaking of speed mismatch, here's something fun. A compressor wheel with a low degree of backward curvature to its blades turns slower for a given PR than a wheel with a high degree of backward curvature. So it behaves like a larger diameter wheel when it comes to matching to a turbine.

JasonC SBB

On a street car, this is typical. The mythical "crossover operation" on race cars, where TIP < MAP, doesn't happen on street cars with streetable turbos, AFAIK.

I measured TIP on my GT2554 and GT2560, at 10 psi boost. In both cases, TIP > MAP. On the 2560 I saw from 15-20 psi gauge pressure in the exhaust manifold. On the 2554, I saw TIP spike to 30 psi above 6000 RPM (suggesting that I'm hitting the efficiency wall (choke) of the compressor).

This is the primary reason why street turbo setups want minimal cam overlap. One way to reduce it, is to advance the exhaust cam. This also improves low-end spoolup. This is also probably why turbo setups like to have the VVT more retarded than n/a setups. (I compared my VVT map to an n/a setup some time ago).

JasonC SBB

The 0.86 a/r GT2860 seems to be more "sensitive" to this than the GT2560. I see much more variation in spoolup in dyno plots, between a log manifold and a manifold like the AbsurdFlow or BEGI S5(?).

Hustler saw a huge gain in spoolup switching over from the BEGI log to the AF. (He claims full boost ~1000 RPM lower, which is a bit hard to believe). I see huge differences in other dyno plots, with the difference seemingly due to log vs. pulsey/tubey manifold.

FM's results for the 2860 vs the 2560 on their log mani were awful:
http://www.flyinmiata.com/tech/dyno_...emy_potato.pdf

JasonC SBB

hmm, this small potato (0.64) made 175 ft-lbs (EDIT: at 3000 RPM) on a dynojet, and 344 hp peak, on a VW 1.8T:

http://forums.vwvortex.com/showthrea...Gt28RS.64-dyno

24 kg-m = 175 ft-lbs at just 3000 RPM:

JasonC SBB

What changed between the above dyno plot, and an earlier one you posted with the pathetic spoolup?: (look at the numbers from 3000-3500 RPM)

edit: oh d'uh, it was the log -> absurdflow.

sixshooter

iTrader: (12)

Everybody knows Greek horsepower is... different.

Savington

iTrader: (31)

It's 345tq, not 175tq. Making 344hp with only 175tq means ~10,300rpm.

Doesn't mean the TIP isn't sky high, too. If they can combat it with high boost pressures (they are) and a good intake (they are) then you can drive more air through the thing. Our problem is restrictive intake manifolds, and once you fix that, we can't get the turbine pressure high enough to drive the shaft speed up to 24psi without inducing detonation.

Also, 24psi is a PR of ~2.6:1, and 345hp is ~38lb/min. Way, way off the end of the compressor map for a 2860RS, and probably way way past the max shaft RPM for the turbo too.

JKav

iTrader: (1)

1.8t's VVT helps here too, as you've probably found (dial out overlap as TIP rises). Also, Miata cams are very mild.

Miatas have knocky engines. Chamber squish might be an area to look at.

JasonC SBB

Ahk, see my edit, I meant:

175 ft-lbs at just 3000 RPM

JasonC SBB

If the VW makes 345 hp with a small (0.64) potato, I don't get why on a miata it wouldn't make 300.

MartinezA92

iTrader: (4)

Because the BP sucks?

Savington

iTrader: (31)

You probably can, if you are willing to run massive boost pressure, super retarded timing, etc.

JasonC SBB

Nah, when turbos are choking, you can't eke power out no matter how much you raise the boost. EGT will rocket, etc.

But I'm not convinced the small tater will choke. Maybe for some reason the small tater chokes with log manifolds and less so with more tubey manifolds.