Why is a Twin scroll/Dual WG setup a benefit?

[Strm Trpr]

Here's a little info on twin scroll technology from a mazda forum.
Mazda scientifically tests single scroll vs twin scroll turbos - RX7Club.com

 

[dsm-onster]

^^^But because it's not in a DSM engine bay, it will be mysteriously refuted as immaterial to the arguement . Interestingly that is with a small turbo at just 9psi and the setup outputting 190ft-lbs. Goes to show that this design doesn't "only benefit large turbos" as has been the consensus elsewhere.

I think you are a little quick to claim that the time differences between an evo3 16g dsm and an evo 8 16g evo are because of the turbo. Evos have a decade of engineering and technology in their favor (mostly in the cylinder head).

The only question I will ask is this. Is a twin scroll design a band aid for a poorly designed oem exhaust manifold due to extremely tight space constraints found in todays engine bays? It seems that people using properly designed manifolds (gradual taper at the collector, equal length, ect.) are seeing really good spool characteristics in the evo world.

Edit*
I realized that "poorly" was probably the wrong term to use above. I was mearly stating that comprimises must be made by engineers to make anything they design work.

Not time difference; the output based on mph and weight. . . The same kinds of head port design is found in the 2g head. The same cam profiles can be applied to both. but yes, the test isn't ever very fair from an et perspective.

And, if the twinscroll design makes an inferior performing manifold (due to compromises) work, then what would it do with a great performing manifold? Great performing anything else? Again, this goes back to the better turbine wheel and compressor wheel design out now adays. You could say that they make up for the poor design of the 1g head ports and open manifold setups prevailant around here. But that's not relavent to the fact that they are more efficient to whatever they are attached.

 

[untouchablecc]

So if im correct on this. A non divided manifold or header on a twin scroll can use a wastegate mounted to the manifold or header but you will not see the gains of having a twin scroll housing. Whereas a divided header you will need two wastegates in the header itself or one in the O2 housing.

 

[dsm-onster]

You cannot have the wastegate mounted at the o2 housing and vent both scrolls at the same time unless the interior scroll has a port around the exterior scroll to the gate, like the evo turbine. And further, any single gate setup with a twinscroll manifold and twinscroll turbine needs to have the divider all the way to the gate valve and seal properly. Or the full effect of twinscroll technology will not be seen. Many tests between open and twin designs don't even utilize a full twinscroll (all the way to the gate) setup.

 

[ViciousLord]

So if im correct on this. A non divided manifold or header on a twin scroll can use a wastegate mounted to the manifold or header but you will not see the gains of having a twin scroll housing. Whereas a divided header you will need two wastegates in the header itself or one in the O2 housing.

please re-read the info posted this thread. Exhaust pulses must be kept completely separate until the wastegate opens.

 

[Black95TSIawd]

Didn't AMS do a dyno test a few years back on a larger frame turbo in both a twin scroll and single scroll housing? If memory serves me correct, it was their GT37r turbo aka T67. The twin scroll only gained miniscule power in the low end/mid range, but lost 80whp on the top end. I'll try and search for it.

 

[dsm-onster]

Yea, that is the one I was discussing that had an improperly sized twinscroll housing. And they also never even bothered with the cam timing. We keep saying over and over again that valve events don't have as much an effect on each other, so why be resitricted to timing events pre-programmed into turbo cams.

I am sorry; but in opinion, SAE papers trump AMS testing. For one, as mentioned the testers need to fully understand the concepts of both systems to make such test objective. Imphatically and clearly applicable in our case: the SAE test was done with a "smaller" exhaust pulse flow platform at meager boost.

What's comical is that this isn't "new tech". This has been employed for well over 20 years. Along with Map Width Enhancement (surge porting), billet compressor wheel strength, compressor hub and blade reshaping, superback technology, extended exducer blades, and even radiusing of the volute exit.

 

[wishihadatalon]

You cannot have the wastegate mounted at the o2 housing and vent both scrolls at the same time unless the interior scroll has a port around the exterior scroll to the gate, like the evo turbine. And further, any single gate setup with a twinscroll manifold and twinscroll turbine needs to have the divider all the way to the gate valve and seal properly. Or the full effect of twinscroll technology will not be seen. Many tests between open and twin designs don't even utilize a full twinscroll (all the way to the gate) setup.

You are correct, but if you know a good welder, and know the base material of the turbine housing, you can do something like this.
Browse Album :: wastegate

 

[dsm-onster]

^^^^ +10000 for awesome and simple. Beautiful! But not sure how a typical turbo mount will allow everything to clear the radiator.

 

[ViciousLord]

You are correct, but if you know a good welder, and know the base material of the turbine housing, you can do something like this.
Browse Album :: wastegate

Well look at that, a TS using a single wastegate.

 

[dsm-onster]

Yes. It's not about the number of gates. It's about the isolation and pressure differentials. There's lower pressure going out the gate than in the gate making a u-turn round the isolation **at the valve** into the next exit at the gate. If the flow is isolated between the blowdown and overlap events, then everything is fine. The flow out the gate doesn't change except that you can get more intake charge since the backpressure doesn't affect the intake, and there is faster spool. But otherwise a 44mm gate would work fine for a typical open scroll setup that would work with a twinscroll setup; no creep.

The only reason one would need two gates is because a well chosen twinscroll setup will spool faster than a typically well chosen open scroll setup. A well chosen twinscroll setup will spool faster while flowing the same total flow up top. After all it's about power under the curve, right? See Kiggly's twinscroll system success for the full benefit of a twinscroll system; whilst utilizing overlap with the mivec head (again cam timing greatly affects bottem line results with a twinscroll system, and never fully applied in some "tests"). Thus the total flow out the exhaust won't require so much more open gating considering a properly chosen twinscroll setup is about faster spool with the same topend and not about the same spool with more topend.

Althought the converse of the later part of the above last sentence is completely possible with an a/r twinscroll setup that spools only as fast as a single scroll setup. Considering the nulling of the effect of backpressure, if the spool is the same and there's little to zero effect on backpressure, what do you think happens up top? This is what's going on with the +18cm^2 twinscroll hx40s around here vs. the bolton hx40s (about 1/2 the twinscroll a/r, but still have the same teardrop entrance to the turbine blades). They are spooling about the same, but at any given boost they are soaring up top, achieving more flow at far lower boost.

 

[dsm-onster]

Considering all the "tests" between twinscroll vs. non-twinscroll, another supposition pops up. Similar to 1999 when the V-late world was saying: "Vortec lsWat? I can make more power with my 350." And it was true. Because the speaker had 25 years working with inferior technology. Oh well, look at today.

 

[TSi Kid]

Then is a twin scroll setup worth the money?
I keep trying to find a good comparison (I liked that rx7 one) but is the steep difference in price worth the money over a single scroll setup?

 

[polarmoment]

Well look at that, a TS using a single wastegate.

Welcome to 1983. F1 engineers didn't seem to have an issue with using a single gate.

BMW:

You must be logged in to view this image or video.

You must be logged in to view this image or video.

zakspeed:

You must be logged in to view this image or video.

503e (toyota Group C/IMSA GTP engine):

You must be logged in to view this image or video.

mercedes M117 (sauber C11 Group C):

You must be logged in to view this image or video.

In fact, not only did they use twin scroll setups with singular wastegates, but that was actually the preferred configuration, as evidenced by the fact that NONE of the race engines used a gate-per-volute setup that everyone on forums and magazines seems to think is absolutely necessary.

It weighs less, it costs less, it introduces fewer points of failure, there's less stress on the pipes having to support a second wastegate, it's easier to plumb, etc. It still surprises me that even with 20+ years of technology on its side, the "import tuner" community still hasn't caught up with 80's race tech.

 

[ViciousLord]

Welcome to 1983. F1 engineers didn't seem to have an issue with using a single gate.

In fact, not only did they use twin scroll setups with singular wastegates, but that was actually the preferred configuration, as evidenced by the fact that NONE of the race engines used a gate-per-volute setup that everyone on forums and magazines seems to think is absolutely necessary.

It weighs less, it costs less, it introduces fewer points of failure, there's less stress on the pipes having to support a second wastegate, it's easier to plumb, etc. It still surprises me that even with 20+ years of technology on its side, the "import tuner" community still hasn't caught up with 80's race tech.

Thank you sir!

Even more proof that you CAN in fact have a properly functioning Twin-Scroll setup using a single wastegate.

Same thing I've been saying from day one!

People automatically think you need twin wastegates to get the full benefits of twin-scroll... which is a myth!

 

[Strm Trpr]

Every single one of those ^^^ TS single gate configurations has two separate pipes leading up to a collector that is probably divided right up to the gate valve probably just like the stock Evo set-up.

And if wastegate runners are not divided up to the gate valve face, then the extra length of tubing between the divided manifold collector and the wastegate is more than enough to quell disturbing the precious exhaust gas pulses and not allowing exhaust gas reversion to take place.

I can't wait to get my set-up finished.
I will degree my GSC S2 cams with the engine on the stand.
These cams have 11.2/11 of net valve lift, intake/exhaust and both have durations of 230*.
The TS set-up will allow these cams with a lot of overlap to really shine and hopefully make decent power.
When this set-up hits the dyno, I plan to tweak the cam timing to see what the affects are.

 

[BoostFrenzy]

While all of that 80's race tech is completely doable, try fitting it inside of a DSM engine bay and see where that gets you.

You would need at a MINIMUM a 44MM on a moderate size turbo to have proper sized wastegate runners. The hard part is making clearance for the 2 gate runners to form one AND mount the gate.

Best idea is like the manifold Twicks69 uses but with TRUE divided runners all the way to the gate. It works because it routes one runner behind the manifold where there is more room then like the DNP style where it routes it in front and limits spacing between manifold and radiator (in stock location).

 

[Shearer]

Marketing bs doesn't come from the originator of the twinscroll design.

Just because YOU havn't found any proof doesn't mean that the rest of the modern automotive world hasn't.

Really, You should keep the personal issues with Full Race in the evom forums.

Where did the twin scroll design originate from?

I have done my testing and I am selling the same stuff I have built for years because it works! Trust me I'd much rather sell divided setups to everyone, they are more labor intensive to make which equals more money in my pocket.

It's not personal, it's business. The twin scroll hype has been one of the most talked about subjects in a long time. The idea of having better response while maintaining equivalent or better high RPM power is a sweet deal. It has been endorsed with promising marketing claims from day 1 and ABSOLUTELY NO ONE has been able to provide evidence that it works as promised. Instead they provide a seat of the pants feeling description or a bunch of SAE documents that are irrelevant to the application at hand.

Twinscroll setups are not a mandate for Emmisions Regulation, nor do they affect emmisions. I wonder why the evo has a twinscroll 16g turbo. . . . ????

They are not a mandate for emission regulations by any means, but it is a factor. Emissions regulations are getting far more stringent than they have ever been. And on the flip side of the coin the consumer doesn't want an engine that makes less power than previous years. Keeping paired cylinders isolated from one another in a small turbo(read, very high backpressure) application surely plays into engine performance while maintaining an acceptable emissions output.

Whereas a divided header you will need two wastegates in the header itself or one in the O2 housing.

The only reason why we started doing twin wastegates on our divided manifolds was cost and potential warranty issues. The extra time it took in labor for a single gate a person could essentially buy a second 44mm wastegate. And some additional stresses were introduced and I did not want to deal with warranty issues.

I am sorry; but in opinion, SAE papers trump AMS testing. For one, as mentioned the testers need to fully understand the concepts of both systems to make such test objective. Imphatically and clearly applicable in our case: the SAE test was done with a "smaller" exhaust pulse flow platform at meager boost.

Out of curiousity. Have you done any testing yourself or is your knowledge of the application comes from SAE papers and what you've read on the internets? You have a lot to say about the subject just curious how much of what you say is actual first hand knowledge.

 

[We're on Boost]

Really, You should keep the personal issues with Full Race in the evom forums.

Wait a minute - we need to be nice to this guy! Ron makes 2 different fully divided tubular manifolds for DSM.
Full-Race makes exactly 0.

Plus, Shearer makes 2 different (and really nice) single scroll manifolds for DSM, where you can take your pick between large diameter runners and small diameter runners. Good grief!

 

[dsm-onster]

NO offense intended. Going over to the evom forums every once in a while gives one a headache with all the misinformation and lack of reasoning.

Last pictures I have of my setup before my "internets" capable computer went down this year. On the kids wii to surf the web: click.

My experience. . .Single scroll bolton housing big h1c, fp2x cams, eagle/wiseco 8.3CR 2.0L, adj cam gears, stock intake manifold. No IC. E85 w/ meth injection, FIC1000cc injectors. Did 125mph in the fwd at 30psi. 20psi by 3500ish rpms.

Swapped to awd and my 70-90 went from 1.8 to 2.1sec (weight and gearing). Swapped to tubular t3 header and the 12cm^2 hx35 turbine housing w/ radiused divider and just downpipe; gained all that 70-90 back at 25psi boost. Spool went to 20psi by 4100-200rpms. Link estimated power/torque definately shows a shift to the right. Playing with the cam gears never did a thing either I gained power low and lost it up top, or the other way around (which the setup was already peaky for my taste), or lost power all around. Gotta love FP cams: perfect balance was with them straight up. Ebayed the cam gears. Good flowing hotside. More than a couple have seen over 130mph and a few 10s out of hx35 12cm^2 hotside.

Suspected I tightened my rod bolts too tight (realized the discrepancy on the eagle install paper). Pulled motor, replaced only rod bearings and bolts. swapped on my ugly home diy twinscroll manifold and the orginal h1c 18cm^2 twinscroll housing and finished exhaust. Spool speed came back to close to the bolton level: 20psi by 3700ish. But power didn't seam to change all that much. Still kind of soggy in the middle with a little more off throttle response. But I was running lean with the maf clamp now. VE changed significantly, apparently. Figured backpressure was altering cylinderfilling so I found a set of adjustable cam gears and retuned everything. Spool speed didn't change but dialing in more over lap (retarding the exhaust cam 2 degrees and advancing the intake cam) put me at a quicker 70-90 than with the open 12cm: 1.7sec. Power flattened out on the left earlier per Link and fell up top just slightly after advancing the intake cam. But I like the feeling with this setup, probably inpart because of the 1g intake manifold.

This is what I mean about testing the proper twinscroll housing. The flow is still there up top with the 18cm^2 housing which is larger than two single scroll bolton housings and significantly larger than the 12cm^2 with open manifold. Considering the h1c compressor has a 54mm inducer and much like a 20g/50trim in total flow, I felt the spool of the open 12cm setup was just to laggy for the potential. Much larger 18cm housing still brought spool down to a great level but not so much that is still choked the top. Fortunately, that was the original housing the h1c was with. Also, it took retuning to actually log a difference in power. The car feels very 'n/a-like'; it responds to overlap like a v8. Yes, it seams the effects of backpressure are nulled. However considering the open 12cm^2 hx35 hotside has seen 133-6mph traps with close to full weight T/E/Ls, one can't readily assume my 25psi boost and only 1.9sec 70-90 time is developing "extreme" backpressure. Apparently, backpressure being anything over the boost is going to alter scavenging. After all, it's all about differentials: even 1psi backpressure over boost can be likened to running an n/a motor with no header upgrade out of it's stock header rpm range; zero boost in the intake manifold and begining to develop measurable backpressure in the exhaust.

Yes, it is very interesting how many refute sae papers with absolute results with a low torque setup. The testers are certainly trustworthy, knowledgeable, and unbiased for financial gain (if tech doesn't work, it's scrapped). The turbo in their experiments is certainly NOT out of it's range of efficient production at all; no "extreme backpressure". The exhaust flow is similarly pulsed as an I4, and backfilling is similar to an I4. Further, again the Evo 8 platform is NOT built from the factory anywhere close to the limit of what an evo3 16g turbo can output (same compressor a/r same compressor wheel same bearing section same turbine wheel; reverse scroll). 220whp with better head/intake flow with 8.8cr is NOT pushing an open scroll 16g. So where does the emissions come into play? I DO see Viciouslord's data logs of spool speed. I also DO see mph with respect to weight output results from the stock evo8 turbo when pushed to hell. The AMS testing showed a tremendous gain in the low end with a drop in power up top. Again, if the product was understood enough so that a larger a/r twinscroll housing would have been used along with cam timing adjustments, we're still resigned to speculation for that test: there would have been no loss up top.

. . .Look at Matt Morrison's twinscroll 12cm^2 hx35 as just one example. Tiny housing and low boost still put him in the mph range of 10second 12cm^2 open manifold hx35 setups. The individual volutes of the 12cm^2 housing truely are only the size of a 6cm^2 14b turbine housing. He can give up 500rpms in spool speed and STILL spool MUCH faster than the open manifold 12cm setups. Imagine the topend flow even with the bit restrictive 264/272 setup. Is this speculation? Just as much as the improperly chosen a/r for the ams test and the guess that the sae papers don't apply somehow.

 

[ViciousLord]

Bringing back this from the dead. Why? To inform everyone that a proper twin scroll single wastegate configuration does actually work. If it works for the fastest Time Attack Evo in the world, it'll work for us too.

Below are recent pictures of the CyberEvo setup for this Time Attack season. Twin-Scroll single wastegate

Enjoy.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

 

[fosforito95]

that looks nice too bad that doesnt look like it would fit very good in a 1g..

 

[We're on Boost]

Below are recent pictures of the CyberEvo setup for this Time Attack season. Twin-Scroll single wastegate

That looks very cool ..... Cyber Evo

Have you seen the Supra forum testing that is posted here:

Hks/full-race t4 exhaust manifold test results

It doesn't necessarily prove anything one way or the other, but they did have divided w 2 gates vs divided w 1 gate vs single scroll. Very interesting.

 

[Strm Trpr]

Is it just me, or are these runners incorrectly paired for a TS set-up on a 4G63?

You must be logged in to view this image or video.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

 

[ViciousLord]

Weird.

No, you're right. I have no idea why they would do this. There has to be a reason as to why they decided to build it like this.

More pictures of the Cyber Evo set up...

You must be logged in to view this image or video.

You must be logged in to view this image or video.