Twin-scroll skeptics...
Mar 24, 2010, 11:13 PM
#166
CART/IRL, Lemans, WRC, TurboF1 all seem to have used single scroll setups at their pinnacle. Many WRC teams were using twin scroll setups until Garrett produced the TR30, which is a single scroll turbo similar in size to a GT30. The TR30 was also used in IRL and Lemans. I believe it has been replaced with a next generation of turbo though, but it's still a single scroll arrangement.
Keep in mind that a circuit racing application is different than a street car app. There are not too many circuits with extremely slow corners; the cars are generally able to stay up in the RPM range where the benefits of twin-scroll design is limited.
In WRC, they use anti-lag, so lag is not an issue
In Lemans, Peugeot uses a V12 and Audi a V10. So they have 6 and 5 cylinders feeding each turbo respectively. With the V12, you could do a pair of twin-scroll turbos, not so much on the V10. But with racing, everything comes down to weight. Those TR30Rs only weigh like 8 lbs or something ridiculous. With their transmissions, they're able to stay up in the rev range quite easily negating a lot of the benefits of twin-scroll.
Also, in racing, the simplier the better. And weight, rather the lack of, is key. So the benefits of twin-scroll don't make much sense on a modern race car.
For a street car? Twin-scroll makes all the difference in the world. The list of twin-scroll turbo cars is only getting bigger (new BMW N55 3.0L engine, 4.4L V8 in X5M/X6M, next gen M5, Evos, JDM STis, GM Ecotec engines, Mini Cooper S, etc.)
Another example of something that makes a huge difference on a street car, but not used so much on race cars: variable valve timing. Why? Because the purpose of it is to make a fatter powerband. But a race car should always be in the upper revs. So variable valve timing just adds extra complexity (something to break) and more weight.
Mar 25, 2010, 12:39 AM
#167
Evolving Member
iTrader: (5)
Join Date: Jun 2008
Location: harvard il
Posts: 123
Likes: 0
Received 0 Likes
on
0 Posts
A little nitpick but IRL never used turbos. Cart, which became Champ Car, used the Cosworth 2.65L with Garrett turbos. And the current Indycars use NA engines.
Keep in mind that a circuit racing application is different than a street car app. There are not too many circuits with extremely slow corners; the cars are generally able to stay up in the RPM range where the benefits of twin-scroll design is limited.
In WRC, they use anti-lag, so lag is not an issue
In Lemans, Peugeot uses a V12 and Audi a V10. So they have 6 and 5 cylinders feeding each turbo respectively. With the V12, you could do a pair of twin-scroll turbos, not so much on the V10. But with racing, everything comes down to weight. Those TR30Rs only weigh like 8 lbs or something ridiculous. With their transmissions, they're able to stay up in the rev range quite easily negating a lot of the benefits of twin-scroll.
Also, in racing, the simplier the better. And weight, rather the lack of, is key. So the benefits of twin-scroll don't make much sense on a modern race car.
For a street car? Twin-scroll makes all the difference in the world. The list of twin-scroll turbo cars is only getting bigger (new BMW N55 3.0L engine, 4.4L V8 in X5M/X6M, next gen M5, Evos, JDM STis, GM Ecotec engines, Mini Cooper S, etc.)
Another example of something that makes a huge difference on a street car, but not used so much on race cars: variable valve timing. Why? Because the purpose of it is to make a fatter powerband. But a race car should always be in the upper revs. So variable valve timing just adds extra complexity (something to break) and more weight.
Keep in mind that a circuit racing application is different than a street car app. There are not too many circuits with extremely slow corners; the cars are generally able to stay up in the RPM range where the benefits of twin-scroll design is limited.
In WRC, they use anti-lag, so lag is not an issue
In Lemans, Peugeot uses a V12 and Audi a V10. So they have 6 and 5 cylinders feeding each turbo respectively. With the V12, you could do a pair of twin-scroll turbos, not so much on the V10. But with racing, everything comes down to weight. Those TR30Rs only weigh like 8 lbs or something ridiculous. With their transmissions, they're able to stay up in the rev range quite easily negating a lot of the benefits of twin-scroll.
Also, in racing, the simplier the better. And weight, rather the lack of, is key. So the benefits of twin-scroll don't make much sense on a modern race car.
For a street car? Twin-scroll makes all the difference in the world. The list of twin-scroll turbo cars is only getting bigger (new BMW N55 3.0L engine, 4.4L V8 in X5M/X6M, next gen M5, Evos, JDM STis, GM Ecotec engines, Mini Cooper S, etc.)
Another example of something that makes a huge difference on a street car, but not used so much on race cars: variable valve timing. Why? Because the purpose of it is to make a fatter powerband. But a race car should always be in the upper revs. So variable valve timing just adds extra complexity (something to break) and more weight.
I like what you said about the variable valve timing. Personally im not a fan of it i like keeping it simple. You are also very right on the wrc comment anti lag i was thinking about trying it out but just seems so hard on the turbo. I have a friend building a hta3582 pretty sure its twin scroll. i want to see how it goes. very interesting thread though and those f1 motors are something of beauty
Mar 25, 2010, 08:33 AM
#168
Ok, I was mistaken on some fronts.
That's pretty much what I was getting at. Street cars there is no doubt a benefit for powerband.
Drag cars, I'd still agree there is a benefit. A GT42 for example barely spools by 7000 RPM in an undivided housing on a 2.0L. Toss it in a good TS setup and it's at full sing by 5500 RPM. A lot of the street class cars have weight limits that are easily reached so you can make up for the weight of the TS setup by pulling it out of other areas.
In my particular setup though, I'm more interested in a limited HP range that still allows a fairly small turbo. I'd prefer to have a bare minimum of weight though and a simple engine setup. The rev range for power that I care about is 5K+ so the difference in response is going to be much less.
I think people are taking my argument as an argument against TS in typical applications. It's not and I'm only arguing for a paticular setup and not a general case.
Keep in mind that a circuit racing application is different than a street car app. There are not too many circuits with extremely slow corners; the cars are generally able to stay up in the RPM range where the benefits of twin-scroll design is limited.
Also, in racing, the simplier the better. And weight, rather the lack of, is key. So the benefits of twin-scroll don't make much sense on a modern race car.
For a street car? Twin-scroll makes all the difference in the world. The list of twin-scroll turbo cars is only getting bigger (new BMW N55 3.0L engine, 4.4L V8 in X5M/X6M, next gen M5, Evos, JDM STis, GM Ecotec engines, Mini Cooper S, etc.)
Another example of something that makes a huge difference on a street car, but not used so much on race cars: variable valve timing. Why? Because the purpose of it is to make a fatter powerband. But a race car should always be in the upper revs. So variable valve timing just adds extra complexity (something to break) and more weight.
Also, in racing, the simplier the better. And weight, rather the lack of, is key. So the benefits of twin-scroll don't make much sense on a modern race car.
For a street car? Twin-scroll makes all the difference in the world. The list of twin-scroll turbo cars is only getting bigger (new BMW N55 3.0L engine, 4.4L V8 in X5M/X6M, next gen M5, Evos, JDM STis, GM Ecotec engines, Mini Cooper S, etc.)
Another example of something that makes a huge difference on a street car, but not used so much on race cars: variable valve timing. Why? Because the purpose of it is to make a fatter powerband. But a race car should always be in the upper revs. So variable valve timing just adds extra complexity (something to break) and more weight.
Drag cars, I'd still agree there is a benefit. A GT42 for example barely spools by 7000 RPM in an undivided housing on a 2.0L. Toss it in a good TS setup and it's at full sing by 5500 RPM. A lot of the street class cars have weight limits that are easily reached so you can make up for the weight of the TS setup by pulling it out of other areas.
In my particular setup though, I'm more interested in a limited HP range that still allows a fairly small turbo. I'd prefer to have a bare minimum of weight though and a simple engine setup. The rev range for power that I care about is 5K+ so the difference in response is going to be much less.
I think people are taking my argument as an argument against TS in typical applications. It's not and I'm only arguing for a paticular setup and not a general case.
Mar 25, 2010, 08:47 AM
#169
There's nothing like learning something new.
Yep. There is a benefit in any application where a 4-cyl engine is fitted with a relatively large turbo, especially if rpm drops below that required to deliver full boost pressure anywhere on the course (or street).
Mar 25, 2010, 10:10 PM
#171
The debate is just that I could get a GT30 based setup with tial housing for 1/2 the weight of a stock frame turbo. But reality, it's only 10 pounds difference, before the weight of the wastegate on the GT30 setup. The GT30 may be slightly less expensive as well.
Mar 25, 2010, 10:30 PM
#172
With a .82 can you deal with losing that much tq on the top end for all your down low responsiveness? The 30R is a good turbo but it's easily over worked on a 2.3 with boost,on a 2.0 as well but it's not as drastic because its not as high at the beginning. You lose 200+ ftlbs of tq, I'llprefer a car that makes peak tq from 4k-8k, not 3k-6k.
Thats a 2.3 30Rcar with mild cams at 30 psi on C16...
To me, you go with the stock frame TWIN SCROLL FP RED or Green, call it a day for simplicities sake and accept it or go bigger with a T4 frame manifold and setupso you can do your 450 whp and make tq to redline on less boost, and make 1100whp so you choose later if thats what you want to do.
If you are really attached to the Tial hotside do it, the v-bands are nice.
Scorke
Mar 26, 2010, 01:47 PM
#173
Evolved Member
iTrader: (62)
Join Date: Feb 2004
Location: Redmond - Lake Tapps ,WA
Posts: 3,336
Likes: 0
Received 6 Likes
on
6 Posts
The precision .63 AR housing is the one to use for a GT30R in T3 open scroll, it's the only one worth looking at. These new T3 twin scroll are interesting though, and the T4 twin scroll on a 35R ish is pretty proven now. Geoff seems to know this stuff better than anyone else and has invested the time to try lots of things.
For reference to that plot posted above, I have mine from last week posted below. Very similar but with a Precision .63AR turbine housing. 2.3L, GT3076R, 28 psi, E85. I make less peak torque, which is expected at 2 less psi of boost, but mine doesn't fall off as much, I drop 100 ft-lbs; ignore the cliff at the end that is the dyno reading after lifting at 7300 RPM. I peak at 450 ft-lbs, and make 350 ft-lbs out the top. And this is with a smaller AR. AMS tested all the turbine housings back in the day and picked the Precision .63 AR because it spooled faster AND made more top end than the garret .63, or garret .82.
V-band is really nice though. It makes life easy, and you don't worry about all your bolts falling out, which is a classic issue for turbine to manifold, and turbine to O2 housing. Twin Scroll V-band is the next new hawtness.
For reference to that plot posted above, I have mine from last week posted below. Very similar but with a Precision .63AR turbine housing. 2.3L, GT3076R, 28 psi, E85. I make less peak torque, which is expected at 2 less psi of boost, but mine doesn't fall off as much, I drop 100 ft-lbs; ignore the cliff at the end that is the dyno reading after lifting at 7300 RPM. I peak at 450 ft-lbs, and make 350 ft-lbs out the top. And this is with a smaller AR. AMS tested all the turbine housings back in the day and picked the Precision .63 AR because it spooled faster AND made more top end than the garret .63, or garret .82.
V-band is really nice though. It makes life easy, and you don't worry about all your bolts falling out, which is a classic issue for turbine to manifold, and turbine to O2 housing. Twin Scroll V-band is the next new hawtness.
Mar 26, 2010, 02:52 PM
#174
Sorry JID but yours falls off just as much as mine does. I would expect if you ran it out to 8500 and kept it wide open the hp and tq graphs would still look the same, the 30R just runs out of breath at this power.It would be more apparent that yours was declining as much as mine was if it was running as much boost.
Peak boost=Peak TQ in this situation, the rest of the variables on my motor were very stock. You can clearly see that my peak comes earlier than yours, and the pull was started later, indicating that it was probably spooling quicker in my scenario.
Regardless of this I am making more tq than you before peak boost would come into play, and at similar RPM indicating that the .82 seems to work better, considering I gave up very little which would be expected, especially with a 2.3.
The Precision .63 is the housing that AMS pretty much exclusively uses, all the other vendors seem to stick to the .82, precision, or the new tial stuff now.I'm just lost as to how you can say the .63 is the one to use with a 30R given our two graphs. Your curve seems to peak later, but with the housing it should happen sooner, no?
Scorke
Peak boost=Peak TQ in this situation, the rest of the variables on my motor were very stock. You can clearly see that my peak comes earlier than yours, and the pull was started later, indicating that it was probably spooling quicker in my scenario.
Regardless of this I am making more tq than you before peak boost would come into play, and at similar RPM indicating that the .82 seems to work better, considering I gave up very little which would be expected, especially with a 2.3.
The Precision .63 is the housing that AMS pretty much exclusively uses, all the other vendors seem to stick to the .82, precision, or the new tial stuff now.I'm just lost as to how you can say the .63 is the one to use with a 30R given our two graphs. Your curve seems to peak later, but with the housing it should happen sooner, no?
Scorke
Mar 26, 2010, 03:45 PM
#175
Evolved Member
iTrader: (62)
Join Date: Feb 2004
Location: Redmond - Lake Tapps ,WA
Posts: 3,336
Likes: 0
Received 6 Likes
on
6 Posts
I looked closer, you are right about the torque out the top, at 7000 RPM they are about the same, and I stop the pull earlier so it's hard to tell after that.
What gear is your pull in? Mine is 3rd.
I think the shift in peak torque is just the timing in the tune. E85 is a little picky at peak torque, and I didn't want to push to the limit, so I run low timing in the area where you are at peak - most likely. My car hits full boost at 4200 rpm in 3rd.
They are pretty close then, it's lame to compare dyno sheets anyway, I should refrain from the urge...
But this is way off topic. Twin Scroll V-band For The Future Win, FTFW.
What gear is your pull in? Mine is 3rd.
I think the shift in peak torque is just the timing in the tune. E85 is a little picky at peak torque, and I didn't want to push to the limit, so I run low timing in the area where you are at peak - most likely. My car hits full boost at 4200 rpm in 3rd.
They are pretty close then, it's lame to compare dyno sheets anyway, I should refrain from the urge...
But this is way off topic. Twin Scroll V-band For The Future Win, FTFW.
Mar 26, 2010, 06:38 PM
#176
My run's were in third. Fullboost at that RPMS sounds about right for the gear if we are talking race gas boost levels.
Hahah agreed on lame about comparing dyno sheets, but it's nice to see how close they are, 3k miles, and 3 years apart! To think of it now I've been through 2 sets of cams, 2 intercoolers, two turbo setups, and 2 rotating assemblies since then
Scorke
Hahah agreed on lame about comparing dyno sheets, but it's nice to see how close they are, 3k miles, and 3 years apart! To think of it now I've been through 2 sets of cams, 2 intercoolers, two turbo setups, and 2 rotating assemblies since then
Scorke
Mar 26, 2010, 08:27 PM
#178
Evolved Member
iTrader: (62)
Join Date: Feb 2004
Location: Redmond - Lake Tapps ,WA
Posts: 3,336
Likes: 0
Received 6 Likes
on
6 Posts
^^^ I don't think it exists, but when it does it will pwn all, and be declared king 
03whitegsr and I have crazy garage fantasies about building one from scratch though 
03whitegsr and I have crazy garage fantasies about building one from scratch though
