My Sr20 Build (part 2) Twin Scroll

[discopotato03]

Nismoid I'd say the increase at that point is a whole lot of things coming together in unison and all complementing themselves - a system .

I don't really have time for a great long spiel but that never stopped me before ...

Anyway most people are six cylinder minded and they need to remember fours are different like 180 crank degree power strokes where the inline sixes are less at 120 degrees . In a true twin scroll system the exhaust events are further apart so the cylinder blow down event has more time to completely exhaust the cylinder and allow scavenging in the valve overlap phase .

Now because the TS does the high exhaust gas pressure thing when needed and low pressure when that's needed it relies on having very little resistance to flow downstream in the exhaust system . Geoff says its critical what goes on down stream of the divorced waste gates so if you have to plumb them back in there must be absolute most volume and therefore least pressure rise in that direction .

Now I'm not up on SR20 head development and I can't remember if this fella use the early high inlet port non VCT one or the later one with lower less elegant inlet porting and VCT . I do know the ITB GTiR inlet manifold fits the early one and is way to go in a maximum effort engine though a heavier and more expensive .

If you look at the camshafts involved here they I think are ~ 12mm of lift and ~ 260 deg duration , don't remember if this is on the seat timing or at 40 thou/1mm .

Anyway high lift cams with modest duration are great for making an engine breathe whilst having good trapping efficiency and this tends to make engines torquey or grunty if you like from not a lot of revs .

So add this to the twin scrolling and less pressure behind it and the gates and suddenly you have a system that can better charge and evacuate its cylinders more efficiently so the volumetric efficiency is higher - more to burn/less energy wasted trying to pump out the spent stuff .

Now as to turbos I can theorise but its really hot air because its not like I've been there with these engines and dryers . That said my gut feeling is that the 52 trim GT37 compressor is the pick of the bunch (48/52/56) and possibly gives the best all round results . On paper they don't pump the absolute same maximum that the 56 trim one does but we are talking a whole 4 odd pounds of air by mass with the potential to make ~ 44 Hp - at the top at best .

I never could find much about the 48 trim version because they were not that popular or sold in great volume which is a pity because I think they'd have it all over the 71.1mm 56 trim GT compressor wheel and give very close to the same max air capacity . Sadly people look at maximums so the smaller end of the comp wheel trim range at times is largely ignored with aftermarket turbochargers . Actually a while back one sold here 2nd hand but for more money than I could justify .

Also as I've said in the past the 52 trim GT3037/GT3076R seemed to be the one people used in the past to get good all round results on 2L fours like SR20's , the one I used to quote had a lot of development on its long merge collector 4-1 manifold . It had a T25 flanged HKS 0.73 A/R GT30 turbine housing and started to make positive manifold pressure by 26-2700 revs which is not what people expected at the time with a GT3037 turbo .

I also used to mention that the 52s were what HKS specified when the Japs wanted crazy big twins for outright drag Rb26's and as you'd imagine turbos like these twinned are not terribly responsive on a 2.6 litre engine much under I'm guessing 5500 revs . When you're using that much turbo for the capacity ANYTHING you can do to wake em up a bit earlier is a god send so I'm not surprised they didn't use the 56T ones in that application .

Anyway twin scroll wasn't popular until people like Geoff Raicer decided to develop manifold and gate packages to complete the system . Mitsubishi does it OE on the Evos (4-10) but where most of it is not usable in other apps is because their turbos (4-9) are reverse rotation and the turbos mount the wrong way around ie comp housing towards the back of say an RB25 or inline SR20 .

The one I'm a bit surprised no ones picked up on is that the Evo 10's 4B11 engine used a clockwise rotation TS turbo and Garret even make bolt on GT3071Rs and GT3076Rs for them .

Now if you was to take the twin scroll twin integral waste gate turbine housing and make up an exhaust manifold with the Mitsy flange pattern you can have a twin scroll twin gate turbocharger without the cost and complexity of external gates . There's nothing stopping you asking GCG or whoever to supply the 3037/3067R in 52 comp trim form and provided they can get that Mitsy style Garrett turbine housing bobs yer uncle . Non std manifold yes with an all Garrett turbocharger having most of the smarts .

Actually the aftermarket is on the edge of supplying turbos to the Evo world because they've been starved of turbos with balls (BBs) and only recently has HKS released their GT7460R BB turbos for the 6-9 and 10 Evos .

Anyway other things are now becoming available like E85 juice at the pump so other ways can be used to get turbo engines up a bit sooner with high performance turbos . Hopefully people will take advantage of E85s increased detonation resistance and start using high CRs in turbo engines which should make the snappier and more torquey off boost .

Another short one , cheers A .

[juggernaut1]

Disco

Yes this is the surge ported 52 trim 3076. We had slight surging issues with the previous 3071 setup when it was boosting. So I went with the surge ported version of the 3076 just to be sure I didn't have surging issues...probably just as well I did. Not to mention the surge ported version was rated at slightly more hp than the non surge ported version.

Yes I am also using the S15 low port head with has been subject to JHH engineering work...including clover leaf combustion chamber, porting and polishing and larger ex valves only. VCT was attempted to be retained for additional mid range torque and to spool the turbo earlier. The VCT kicks in at 1500rpm and comes off at 7000rpm. Standard I think it comes on at 1000 and switches off at 5000 rpm or thereabouts.

PS I see BMW is now ditching the twin turbo setup (was it a compound setup?) on the 135i inline 6 and going to a single twin scroll turbo.

[discopotato03]

Twins are always going to be an expensive option and it adds so much extra associated componentry because two turbos need twice as much plumbing .

Twin scroll singles are so much cheaper lighter and less cluttered than parallel or god helps us sequential twins .

The complex part of a twin scroll twin integral waste gated turbo is the turbine housing and if they can get that right the rewards are huge .

If you ever get the chance have a look at the turbine housing on a Mitsubishi twin scroll TD05 Evo turbo . Not particularly large but the work quite well on those 2L fours . We really could use a similar styled TD06 Evo turbine housing because everyone's high flowing them with TD06 family turbines and even the 10.5 cm TD05 housing is a tad small for maybe 260+ Kw 4G engines .

The larger of the two Garrett Evo 10 housings is 0.94 A/R and designed around their GT30 UHP turbine .

A .

[phunky_monkey]

Wow Ian, a 50hp gain at 3800rpm is fantastic! You must be over the moon with it. Glad to see the wait was worth it after all.

I wonder what the exhaust housing porting was worth power/response wise??

[juggernaut1]

Boost graph:

Points of interest:

boost graph in KPa!

starts making positive boost at approx 2300rpm

approx 21.6 psi by 3750

approx peak boost of 23 psi at about 5000rpm then starts dropping.

different hp compared to graph posted above - not sure whether ran more boost to achieve above graph.

doesn't make 1.6 bar of boost at 3600rpm as previously advised - but see above comment.

Edited September 24, 2010 by juggernaut1

[badhairdave]

boost graph in KPa!

lol. love it.

I reckon the way it's ramping onto boost it'd make 1.6 at 3600 anyway.

What boost control are you running out of interest and what spring pressure in the gates.

My old rb25/t300s combo liked to slowly drop boost in a similar way which i put down to the 7psi spring in the gate.

[R31Nismoid]

Bah what is with you WA people and being weird?

You do 1/4 runs in km/h - measure power in HP ... and do boost in KPa lol

Interesting though it doesnt make peak boost until 6250rpm?

And the torque/power have both well and truely been holding for 2500-3000rpm!

[juggernaut1]

lol. love it.

I reckon the way it's ramping onto boost it'd make 1.6 at 3600 anyway.

What boost control are you running out of interest and what spring pressure in the gates.

My old rb25/t300s combo liked to slowly drop boost in a similar way which i put down to the 7psi spring in the gate.

Haha...I'm not even going to ask for AFR's .....there probably in litres per 100klm's Lol.

I'm running 1.2 bar springs (the old gates had 1 bar springs)

Running a 3 port Mac valve as the boost controller which is controlled by the Autronic.

However, the way the Mac valve is plumbed is what I believe is "unconventional". Both the top and bottom ports of the gates are plumbed back into the same port on the Mac valve (using a tee). I guess there must be a restrictor in one of the lines somewhere. This to me could be one of the causes for my falling boost...I'm not sure? This plumbing was carried over from the original setup which was done by the original tuner.

I was speaking to Jeff at C-Red about this boost control side of things who suggested we plumb the gates using a conventional method ....i.e. not using the same mac valve port for the top and bottom ports of the gates. Failing this we may need to use a dedicated boost controller becuase, due to the size of the gates when combined, a small movement in both gates makes a big difference to the boost (read sensitive). In his view a dedicated boost controller will be more precise. I have a Blitz ID3 dual solenoid bost controller at home so this may be worth trying if the conventionally plumbed mac valve does not work.

Edited September 24, 2010 by juggernaut1

[juggernaut1]

Interesting though it doesnt make peak boost until 6250rpm?

Ummm..... I think your looking at my power curve.....the other curve is the boost curve.

Edited September 24, 2010 by juggernaut1

[lachlanw]

do you lot realise kpa is just moving the decimal point 2 places to the left compared to bar

nismoid the power is the blue line and green is MAP ? refering to your peak boost at 6000rpm ?

[R31Nismoid]

lol fark ye whops - was on the phone and looking

Guess thats even for WA comments haha.