wantboost

iTrader: (2)

I was on ATPs site and saw this on the front page



http://www.atpturbo.com/mm5/merchant...egory_Code=NGH

They offer them for many turbine wheel sizes... and from the pictures it looks like they might actually contain two completely separated volutes like a true twin scroll housing.

You can see the full length divider clearly here in the GT30/GTX30 housing




They also offer divided inlet v-band twin scroll housings as well.

TheShodan

iTrader: (14)

Oh yeah... You're just now seeing this?

From what insiders are telling me right now, yes, they're using an actual completely divded volute.

There are a couple of "hiccups" with this, but once they're solved its definitely something I'm willing to swap out my housing for to see.

1) Current availability is a bit difficult, despite ATP having it on the website. I believe they have the first official "batch" of housings, (so in certain sizes, i'm quite sure you can get one) but influx of new ones on a regular basis has been difficult due to their (Garrett's) contract disputes with the foundries that produce them.

2) Cost. On average the housings alone are a bit over $500. This is because of the material they're using (to compete more with the EFRs since they are no longer using the investment cast that we've come to love).

3) Limited sizing. Since Garrett really wants to take advantage of how this phenomenon works properly and not just some divided housing nonsense, they're sticking with the GT(X)-R series of turbos only right now from GT30 sizes and up. No cross over to the journal bearing market. Now, that doesn't mean that ATP or other companies themselves won't try for their own versions that are not Ni-Resist material or Investment cast steel, but that's going to be based upon initial test results first. Look to the Evolution , Subaru and Toyota crowds to start this testing first. unfortunately, knowing them, they'll only dyno the car to see if there's any difference, and not use real-world applications.

I'm sure either you, me, Tony the Tiger (who did the initial .78A/R divided housing test and was disappointed).

Now if only Myles would step to a Garrett.... Then we can do some real testing . I think I may talk to him about that later this year. Muhahaha

wantboost

iTrader: (2)

The price is worth it to me if they work, they offer quite a few different aspect ratios for each turbine wheel size which is nice. since the b series ubermotor I plan on building later this year will be nearly 2.3 liters I'm sure I could take full advantage of the true twin scroll housing and have a nice, broad, fat, flat torque curve. I think the GTX3076R or something like it (hint hint ) would be insanely responsive and produce gobs of torque on my motor around 20-25psi on pump

I might even be able to get away with excellent response on Tue GTX3582R or similar turbo with my large displacement, high compression, ported head, big cam plans. add in that everything will be ceramic coated and possibly wrapped/blanketed and I have even better response although I think the 3076 would be the ultimate response torque monster.

ps we need to talk about a turbo for this motor. I've pretty much got my rod length down which will definitely help torque production on top of the 98mm stroke..

it is nice to see Garrett making a real twin scroll housing. maybe now people will stop calling a divided housing a twin scroll housing lol.

still waiting for someone to post "i thought divided is twin scroll" lol

TheShodan

iTrader: (14)

Just wait for it, give it time.. I already get flack on it on EvolutionM about that. Its almost as though I were and extremist starting a military coup when I showed the data regarding that one, man.. But I knew the pot would be stirred a bit. I'm pretty used to it by now.

*Edit 4/30/2014* spoke to a few special people. Basically, availability is just nill right now, and the biggest problem is the v-band turbine inlet that's divided. The users of those twin scroll v-band turbine inlets is that the divided part of the inlet only allows the turbo to face one direction, (just like with the regular divided flanges of the standard T4 divided turbine inlets).. So really, the advantage is just space savings.

Honestly, for the size turbochargers that I use, for the money, I'll just convert to TiAL stainless.

wantboost

iTrader: (2)

What I'm wondering is why are both volutes the same size/ar, at least from the outward appearance? If you look at the Borg Warner EFR TS housings, as well as the EVO3-10 TS housings each volute is a different size.

wantboost

iTrader: (2)

Here's the T3 7163 housing, in the cutaway you can clearly see the aspect ratio difference between each volute



EVO3-9 10.5cm housing



I have a 34mm inlet restrictor EVO9 rally turbo downstairs with the 10.5 single flapper housing, both volutes are definitely different sizes. as confirmed by someone who measured each volute then cut the housing between volutes. also confirmed by people who have logged backpressure on cyl1-4 and 2-3.

I'm wondering if it's because of the runner difference between the paired cylinder groups. most of the time, in most 4cyl applications the 2-3 runners are much shorter than 1-4. especially in regards to the oem turbine housing. Even most aftermarket divided manifolds will have shorter 2-3 runners regardless of design unless someone purpose builds a true equal length divided manifold (not easy, I have one)

Could it be to properly time the cylinder pulses? On the EVO3-10 the longer runners of 1-4 are paired to the smaller volute and 2-3 paired to the larger volute. could this be to allow 2-3 to make up for the shorter runners? I could see if both volutes were the same size that with very different length runner pairs that the timing of pulses reaching the turbine wheel could be horribly wrong

wantboost

iTrader: (2)

Here's the T3 7163 housing, in the cutaway you can clearly see the aspect ratio difference between each volute



EVO3-9 10.5cm housing






these various angles give you a true sense of just how different the volutes are.

I have a 34mm inlet restrictor EVO9 rally turbo downstairs with the 10.5 single flapper housing, both volutes are definitely different sizes. as confirmed by someone who measured each volute then cut the housing between volutes. also confirmed by people who have logged backpressure on cyl1-4 and 2-3.

I'll try to take a pic showing the size difference better

I'm wondering if it's because of the runner difference between the paired cylinder groups. most of the time, in most 4cyl applications the 2-3 runners are much shorter than 1-4. especially in regards to the oem turbine housing. Even most aftermarket divided manifolds will have shorter 2-3 runners regardless of design unless someone purpose builds a true equal length divided manifold (not easy, I have one)

Could it be to properly time the cylinder pulses? On the EVO3-10 the longer runners of 1-4 are paired to the smaller volute and 2-3 paired to the larger volute. could this be to allow 2-3 to make up for the shorter runners? I could see if both volutes were the same size that with very different length runner pairs that the timing of pulses reaching the turbine wheel could be horribly wrong

NotARaCist

iTrader: (1)

Correct me if I'm wrong...isn't the only advantage to twin scroll a faster spool?

wantboost

iTrader: (2)

not really.

yes initial response is better but it offers quicker transient response and a much broader, flatter torque curve.

also pairing the cylinders by pulses increases engine efficiency and reduces back pressure since all of those pulses aren't fighting each other on their way to the turbine wheel.

TheShodan

iTrader: (14)

You're getting better at this, young Jedi..

But with the one Garrett I had, the volutes were of different size to coincide with the angle exhaust pulse.

wantboost

iTrader: (2)

Yea i figured the volutes had to be sized differently, odd they kept the turbine housing an equal thickness in regards to the outer diameter as opposed to the evo3-9 and borg warner housings.

I think they could have done it like those two did to save some weight, unless it's for burst containment although I've never known Garretts specialty non OEM housings to have as much burst containment as the OEM units