Twin Scroll Vs Single

[bigmikespec]

I see ATPTurbo now have a 1.06a/r T4 twin scroll turbine housing for the GT3076R.... this has intrigued me quite a bit! I can't help but think that could be very interesting to try.

So what are you thinking Lith, decent twin scroll manifold (i.e. Full-Race or similar) twin wastegates and GT3076R with 1.06a/a T4 divided on a stock stroke RB26 with its six throttle bodies?

[Lithium]

So what are you thinking Lith, decent twin scroll manifold (i.e. Full-Race or similar) twin wastegates and GT3076R with 1.06a/a T4 divided on a stock stroke RB26 with its six throttle bodies?

Depends on what you want really - I'd be more inclined to go the above using a GT3582R but for an RB25 a GT3076R with twin wastegates etc on a FullRace type manifold would be pretty fantastic. Hell I am rapt with my .82a/r T3 internally gated GT3076R on my RB25 and I'm not even tapping its flow potential yet. I couldn't imagine the same thing with better spool, less restriction and 22psi pumping through it!

[Rolls]

How does everyone think a twin scroll GT3071R would go on an rb25det with the appropriate power, would it come on much sooner than it already does? and if so what about peak power, would it be basically the same?

Would external gate vs internal make much difference? I'm trying to figure out what the ultimate combo for response and around 280kw would be.

Edited June 29, 2011 by Rolls

[Psymon]

Good question! I'm wondering the same thing at the moment. I have a HKS exhaust manifold I could use for a TS setup on an RB25, but i'm wondering if the results would be worth it. I was considering a GT3076R a while back but from memory couldn't find an appropriate housing.

[Rolls]

How much of a spool increase is usual with a twin scroll, 200rpm? more less?

[rob82]

Good question! I'm wondering the same thing at the moment. I have a HKS exhaust manifold I could use for a TS setup on an RB25, but i'm wondering if the results would be worth it. I was considering a GT3076R a while back but from memory couldn't find an appropriate housing.

Why dont you just get a Garrett EVO9 replacement twin scroll GT3071 or 3076 with internal gate.

You will find that the 0.76AR TS GT3071 will spool about the same rate as the 0.63 GT3071 but the power will be more - ie closer to the 0.82. Twin scrolls are about broadening the torque curve.

My pick for 280rwkw with the rb25 would be the TS GTX3071 with the 0.73A/R. I would expect that power at 15psi with full boost around the 3000-3200rpm range.

[black bnr32]

i suggest looking at the popular turbo 4 cylinder cars for answers, such as the evo and any sr20-based car. also, there is some discussion on bimmerforums regarding inline 6s and twinscroll usage.

[Lithium]

A twin scroll GT3071R on an RB25 would be pretty mental - I'd personally make the most of the extra transient response and torque boost the twin scroll gives and use a GT3076R, forgot all the ideas you have about the different turbos with the single entry housings and don't get too hung up on boost curves only having a couple hundred rpm difference in spool... the big draw to twin scroll to me is the increase in VE, remember we run more boost to boost torque but thats not the only way to do it.

Here is an overlay of an SR20 going from a .64a/r GT2871R to a twin scroll GT3071R from Full-Race:

I really think a twin scroll GT3076R would be the ideal on an RB25DET, we already know GT3076Rs are barely noticeably laggier than a 71R and with the extra boost in transient response and inherant torque gained with twin scroll - basically a twin scroll GT3076R should out grunt any open housing turbo bigger than a GTRS... at least once it has over 3000rpm on it.

Food for thought?

[dorifticon]

Super interesting read, I'm currently looking to go from twin HKS 2530's to a much larger single with tubular manifold.

This info has been hard to find thus far, I vote to sticky.

[Dose Pipe Sutututu]

A twin scroll GT3071R on an RB25 would be pretty mental - I'd personally make the most of the extra transient response and torque boost the twin scroll gives and use a GT3076R, forgot all the ideas you have about the different turbos with the single entry housings and don't get too hung up on boost curves only having a couple hundred rpm difference in spool... the big draw to twin scroll to me is the increase in VE, remember we run more boost to boost torque but thats not the only way to do it.

Here is an overlay of an SR20 going from a .64a/r GT2871R to a twin scroll GT3071R from Full-Race:

I really think a twin scroll GT3076R would be the ideal on an RB25DET, we already know GT3076Rs are barely noticeably laggier than a 71R and with the extra boost in transient response and inherant torque gained with twin scroll - basically a twin scroll GT3076R should out grunt any open housing turbo bigger than a GTRS... at least once it has over 3000rpm on it.

Food for thought?

Jebabs!

Look at that torque curve, it's more of a straight line than a curve now!

[Rolls]

I really think a twin scroll GT3076R would be the ideal on an RB25DET, we already know GT3076Rs are barely noticeably laggier than a 71R and with the extra boost in transient response and inherant torque gained with twin scroll - basically a twin scroll GT3076R should out grunt any open housing turbo bigger than a GTRS... at least once it has over 3000rpm on it.

Food for thought?

I guess my thinking was anything much over 280kw on the street isn't really usable so by going the 71 you gain a bit of extra response/midrange and the car will probably end up being faster.

[Lithium]

I guess my thinking was anything much over 280kw on the street isn't really usable so by going the 71 you gain a bit of extra response/midrange and the car will probably end up being faster.

WTF? Ok, I know you and some others have seem some grudge against GT3076Rs but seriously - that post is just being silly. You don't have to like them, but don't muddle things up by making comments which are basically saying "A car running an ever so slightly larger turbo is going to be so laggy and have so little midrange torque and so much more power that it will probably be slower". Try and stick to facts for people who don't know better.

My thoughts:

1) The throttle response is going to be so tiny between those two turbos its really not going to make any difference unless you are having rolling races from 2500rpm-3500rpm, and its going to be hard to call the start accurately enough to determine if its a reaction time or a spool thing separating the cars. Once you are actually in the "fun" part of the rev range, the throttle response is going to be so small I'd venture to say it'd be impossible to quantify the difference.

2) Midrange isn't going to be higher on a GT3071R unless you are running more boost in the midrange, which is a tune thing. Both turbos will be at their controlled boost level well before "midrange"

3) Midrange punch is the most likely part of the rev range where traction will be broken - make up your mind, you want to over come the tyres or you don't?

4) Faster = sum of the parts, if a car can handle everything a GT3076R can throw at it (see Tomei STI, one of the fastest time attack cars on the globe) then its going to be faster than a GT3071R equivalent... albeit not much. If its going to be slower because of the power a GT3076R makes over a GT3071R, like the response - there isn't a night and day difference between the GT3071R and 76R power potential, you're probably already having the same issues with the GT3071R.

This has given me an idea....

[wolverine]

I'm with Lith, there never seems to be much difference between the 71 and 76mm compressors when it comes to response in Garrett housings.

You have to lean on the 71mm compressor to make the numbers where you can sit the 76mm comp in pretty fat area and still have some headroom if you want.

[Rolls]

saying "A car running an ever so slightly larger turbo is going to be so laggy and have so little midrange torque and so much more power that it will probably be slower". Try and stick to facts for people who don't know better.

Uhh I never said that? Don't put words in my mouth.

I said if 280kw is about the limit for the street then what is the point in going the 76? You are just going to make more peak power/torque for a slight loss in spool. Why not keep that increase in spool?

I didn't realise there was that little difference though, what kind of spool differences would be expected between the two? If it really is nothing then I see your point.

Edited June 30, 2011 by Rolls

[Lithium]

Uhh I never said that? Don't put words in my mouth.

Rolls - you said the following: anything much over 280kw on the street isn't really usable so by going the 71 you gain a bit of extra response/midrange and the car will probably end up being faster.

I didn't realise there was that little difference though, what kind of spool differences would be expected between the two? If it really is nothing then I see your point.

Its not nothing obviously, but in the ranges you're likely to be utilising the power the difference in spool will be negligable... especially with a twin scroll housing, which is what this is about.

By setting your power ceiling to the most a GT3071R you do artificially give a GT3071R the advantage, though it'll be such a small one it'd be hard to quantify and that kind of conjecture is totally missing the point of optimising the turbo match to an engine - it makes it seem like you're just making up rules to push your view point. Keep it separate, a car running a stock turbo would be faster than a car runing a GT3071R if they are both running 9psi - but that defeats the purpose if you have a car built to run a GT3071R to its potential.

[Rolls]

but that defeats the purpose if you have a car built to run a GT3071R to its potential.

Ok point taken, so the gains from going 71 vs 76 are so negligible its not worth it, and a 76 provides a much higher ceiling for similar response. I'm still curious though, what is the likely change in response/spool time?

[Lithium]

I started this thread to see if people were interested in trying to nail that down instead of speculate too far.

The higher the engine speed, the smaller the difference - realistically. The main things that cause lag are the overall rotational inertia (the majority of the inertia comes from the turbine wheel which is made of heavier materials) and the wheel design. Both GT3076R and GT3071R have the same turbine which is the source of most of the inertia, and they are both 56-trim GT-series Garrett compressor wheels... on the same turbine shaft etc.

For the sake of picturing the effects of the differences between the two turbos in an overly simplified way - lets pretend both the compressor and turbine wheels on a GT3071R contribute evenly to inertia and try and work out how much more work is required to pump a certain amount of air.

Even without counting the fact the turbine and bearing resistance contribute way more to the equation that I am giving them credit for, the increase in the moment of inertia going from a 56trim 71mm comp wheel to a 56trim 76mm comp wheel on a GT30 core could result in at tops a 7% gain for 13% more flow potential.

Take into account the GT3076R pumps more air for the same rpm - ie needs ~104000rpm to push 25lb/min at 1.5bar of boost whereas a GT3071R needs ~112000rpm (9% more rpm required than something with 7% more inertia!?!) it stands to reason that in terms of transitional throttle a GT3076R might not be as bad as you'd expect.

That is very overly simplified (the maths and such forth get long winded and I should be working), but thought I'd just give some crackpot theory to the comparison

[Harey]

How about a twin scroll T67 on an RB25!!!!!! I think I have a 3/4 chub

That SR20 dyno sheet is unreal. From that example (and others) when choosing a twin scroll setup you should choose a slightly larger turbo then what you would choose if it wasnt twin scroll.

[Lithium]

How about a twin scroll T67 on an RB25!!!!!! I think I have a 3/4 chub

Thats why god made Holset HX35s That was my preferred choice when I got my GT3076R, but I was concerned at the effort to make it work. If I could I'd go back to 2006 and tell myself to go a twin scroll HX35 on a suitable manifold/wastegate setup.

[juggernaut1]

Probably seen this before, but my effort - sr20 .78 twin scroll 3076 52 trim, full-race manifold with twin gates.

Edited June 30, 2011 by juggernaut1