I don't like T3 twin scroll. I also hate T4 open. I do believe it is a huge choke point right at the merge for a t3 TS. (I also hate making merge collectors for them). I have made decent power at moderate power levels with them but anything in the 9 sec power range seems to suffer from the small T3 TS. The T4 open is just pointless in my eyes. It's a hard test to go back to back with.

There is no doubt dividing the pulses all the way to the turbine housing has more advantages(with a properly designing exhaust manifold) than just spool times.

 

Got some more data and this stuff is AWESOME...if only I had backpressure probes wired into the dyno already (next week) we could have had even more data.

This is test subject #1, an ER LR2.4 combo which is so close to my car that I literally used the same map. Same head, intake, cams, injectors, etc. The differences are subtle but the results are amazingly similar at least as far as power. This car has a Full Race 8374 T4 TS kit on it with the 1.05 housing. Results-



Before I get the "why did you stop at 32psi" let me say its because the 1.05 is way too small. I have a 25% differential in WGDC to get it to hold as much as it does out the top vs down low (74% up top/48% at peak torque vs the 40% straight across I ran in mine). Good news is there will be a 1.45 housing released later this year (if not already) that would sacrifice very little spool in this turbo and combo, yet deliver much more power out the top.

Here is the same graph overlaid against my 3586 with a 0.82 single T3. On the street I logged 21psi at 4200 so I dont feel it was unresponsive but in comparison it looks like a 2.0L. Since the power is identical at the same boost level on virtually the same build I really feel this was a fair test of the power of twin scroll versus single.



Now for the downside, which some may say isnt a downside, of the 1.05 Twin scroll in this combo. Past 7500 my car starts to take over in a hurry. I didnt show the full revved out pull as it was obvious without even datalogging that I had started to get on the wrongside of backpressure rather quickly with the small 1.05 housing. Since the timing, boost, and AFR were all the same and torque started dropping off so fast its pretty clear this combo needs a bigger TS housing.

Going with the "paper" the 1.05 should have been plenty big as 83.3% of 1.05 is still equivalent to a 0.87 single T4. In practice it differs, much as we found with the 1.15 to 1.28 housings on 42Rs years ago. It certainly spools like it is 60% of the size making a whopping 170 ft lbs more at 4800 rpm than mine. I think this is the main reason TS stuff hasnt really caught on as much in most applications because a little housing change can really make a large difference in how the engine reacts. The 1.15 to 1.28 actually spooled 200 rpm faster and picked up 70whp. Back then we didnt datalog engine functions like we do now on our racecars but you can bet your sweet bippy it was reduced backpressure and the resulting LACK of reversion.

 

Major props for continuing to research things and to post this data. I'm thinking that a LR2.4 or SLR2.2 with a 8374 in the 1.45 housing is going to be a great combo.

 

Did it just a couple months ago. Had a client go from a T3 .82 to a TS T4 1.06. Same car, same engine, etc, etc. Ill try to dig out the logs/charts.

IIRC the spool difference was negligible.

 

Is this on a gt30 car? If so 2L ppl typically complain how the .82 wasn't worth the lag and prefer the .63

 

An important metric I always consider in looking at a dyno chart has nothing to do with boost, but where the 200whp mark is crossed on the rpm axis. That single statistic tells me more about streetability than anything else. In your apples-to-apples comparison, the difference is 600-700rpm, which is huge. A 700+whp setup that can generate 200whp before 4k rpm will be 'lights out' on the street.

People tend to lose track of three important considerations when looking at dyno charts:

1) A dyno chart doesn't reveal transient response - a critical street metric.

2) Whatever lag there is in 3rd gear is worse in 2nd and even worse in 1st.

3) Peak power sells parts, average power wins races.

 

Ted, very true. I think a better comparison of this particular car is it against a 2.0L BBK also on E85. It gives up 30 ft lbs for a few hundred rpm and then completely dominates. The fact that this car makes 300 ft lbs (your 200whp) almost like stock is the big thing to me as well.

 

It was a GTX3576. And the car liked the .82 more then it liked the .63

 

Would be great if the 1.28A/r housing was being made too.
Will the due out soon 1.45 fit the 7064?

 

1.45 isn't necessary on the 7064. Its a 56lb/min turbo vs the 8374 which is a 79lb/min turbo. You won't be pushing the power/exhaust gases to need the 1.45 on the 7064. The 1.05 should be more than enough on the 7064.

 

It still suffers the same fate at over 7.5k. That's why I want the 1.28

 

I haven't even seen a 1.05 7064 dyno on an evo. Mind posting it up?

 

I posted that as an assumption my bad. Until convinced I think no TS turbos are matched.
And by that I mean all results I've seen don't do it for me over 7.5k and believe that when the bigger housings are out then they will do what I need.
And surely this is why they are being developed?

 

My understanding was that the 1.45 was being designed for the 8374 & 9180. Not the 7670/7064

 

Ah I've heard wrong then. And I always thought its too big of a jump up for those turbos and hence my wish of 1.28ish for the 64&76