Twinscroll turbo sizing
Registered User
Joined: Nov 2008
Posts: 579
Likes: 0
I don't have it backwards, I have seen divided scrolls against the undivided counter parts head to head. I'm aware of the division and what is given up to make the divide possible. It's why I said generally you up size for a divided in my earlier post. Have you ran a divided on a S2K in a mid-frame turbo or smaller? Just curious.....
There's blow by in turbos too. Slow down a little and think about the turbo inlet, consider that the impeller wheel doesn't touch the housing, also the center section is efficient in comparison to the outer portion due to engineering constraints in the metal. If you measure the total area of the intake portion you'll find that the % of effective turbine vs total area changes with the overall size of the turbo. Technology continues to improve these factors but it's still a factor. Also the gains on this small divided setup aren't worth the trouble. I guess not everyone can be shown the light. Do less blowing out and more thorough thinking.
This is why many twin turbo setups have given way to single turbo setups. Single turbo "lag" has become more of a moot point thus a single brings greater gains for some of the reasons I explained above aside from some other more obvious reasons. This is all considered even with the gains had by twin turbos which are some of the exact same advantages a twin scroll brings to the table (separating exhaust pulses).
Numerous books by turbo engineers have taught me that smaller turbos are inefficient by nature due to engineering restrictions. I guess you know a lot more than them though on that topic.
I can see you've made your mind up.
Enjoy!
There's blow by in turbos too. Slow down a little and think about the turbo inlet, consider that the impeller wheel doesn't touch the housing, also the center section is efficient in comparison to the outer portion due to engineering constraints in the metal. If you measure the total area of the intake portion you'll find that the % of effective turbine vs total area changes with the overall size of the turbo. Technology continues to improve these factors but it's still a factor. Also the gains on this small divided setup aren't worth the trouble. I guess not everyone can be shown the light. Do less blowing out and more thorough thinking.
This is why many twin turbo setups have given way to single turbo setups. Single turbo "lag" has become more of a moot point thus a single brings greater gains for some of the reasons I explained above aside from some other more obvious reasons. This is all considered even with the gains had by twin turbos which are some of the exact same advantages a twin scroll brings to the table (separating exhaust pulses).
Numerous books by turbo engineers have taught me that smaller turbos are inefficient by nature due to engineering restrictions. I guess you know a lot more than them though on that topic.
I can see you've made your mind up.
Enjoy!
If you mean stepping up to a larger turbo with for the purpose of a larger turbine to compensate for the divided inlet, that can help to some extent. The divided manifold and housing lend to quicker spool and response, which allows you to move to a larger turbo. Personally, I prefer to use a slightly larger housing if possible, and have the smaller turbine wheel with a lower moment of inertia. The larger housing will have a less constricted inlet and allows for a less radial path to the wheel, increasing efficiency even more than the pulse dividing manifold and housing design already have.
Yes, I have run divided and undivided setups on the S2k in small, mid, and large frame combinations. I have used those combinations on many more than just a F series powerplants.
Onto the turbo itself. If you are using "impeller" as a reference to the turbine wheel, its a helpful tip to let you know the term is incorrect. An impeller provides movement and energy to a fluid, and not the reverse. Impeller can only be used to describe the compressor wheel. The miniscule amount of air that "seeps" past the compressor wheel and housing inlet is under .5% I would personally be more concerned with the overall efficiency of the compressor wheel and housing, as it will have a more pronounced effect on overall performance. Now, because we are talking more about the turbine side of the equation, I should go into that.
Turbine wheels are (generally) cast in 713C inconel, and the housings in cast iron. Both materials have an extremely low rate of thermal expansion, and that is one reason they are chosen. After a turbine wheel is cast it is machined to meet tolerances inside of the housing, and so it can be balanced properly. The gaps between the inducer of the turbine wheel and housing are very small, much like the compressor wheel and housing.
Twin turbo setups are ditched in favor of single turbos in street performance applications due to simplicity and cost. In some applications twins will not fit (especially if not designed as such from the factory), and that being another reason to switch. In all out performance, the twins reign king. In OE ultra high efficiency setups twins are preferred. BMW will soon be releasing a triple turbo engine, and Porsche will do the same next year.
Reading a couple turbo books for beginners does not make up for real world experience in engineering and turbocharging.
Registered User
Joined: Dec 2006
Posts: 24,119
Likes: 4
From: Seattle / Kalifornia
Stepping up in terms of A/R is what I meant, though yes the entire turbo size would be a good move as well since it'll be more efficient on top and equally efficient on bottom due to the divided housing. So long as we're talking about a properly setup system and a large enough setup to begin with for it to be efficient.
You found small frame turbo did better divided? I've yet to see that be anywhere near true with setups posted on here so far. Do you have dyno sheets and such and some specs so we can compare more closely the facts? We'd need to consider charge piping size, what manifold, turbo, inter-cooler size etc. I'd be surprised to see any substantial gain, or hell, at this point any gain at all. So far I've seen worse numbers on small frame stuff divided.
The gaps are small yes, they still exist, I'm aware of the metal used in this application, though as you're probably aware there's two other more expensive choices for the wheels too that are becoming used more often. There's also the core section which serves no purpose in terms of flowing air. Also as you know the first bit of blade is dull and wide for strength, this is the portion I was also mentioning being inefficient.
^ The point here was that if you take a very tiny turbo and compare it to say a GT35R looking at the compressor wheel section you'll note that when measuring total area that the amount of usable compressor wheel vs total area is better on the GT35R. It's simple math. That's not even considering the core section used for rotating the entire mass which shares the same characteristic. If we were to measure these factors you'd find there's a fair bit of difference in usable area vs wasted area. THIS is the factor I'm speaking of.
Twin turbos are often ditched partially for this reason though sometimes for the reasons you listed. It's not always easier to go single. Some cars it's much easier to pipe and setup a twin setup. In OE ultra performance setups twins are generally preferred yes and I support that measure, though they rely on cheaper turbos to accomplish what a better quality single could have done. For them they still look at things from a financial perspective, even ultra performance cars have to make a profit. If you want to talk efficiency let's just move on to variable pitch turbos, they win hands down. Yet they ....don't work effectively in small sizes due to engineering restraints, again, a matter of a technology unable to be accomplished in a small physical space. There will come a day no doubt, it's just not yet.
I believe the writers of these books have enough knowledge to be able to make the argument, it's not a conclusion that I came upon myself it was written as fact.
I've gone from a twin turbo to a single turbo in my previous vehicle, it gained all around. Though we can't use that example as the turbo I applied was of better quality, newer, etc.
You found small frame turbo did better divided? I've yet to see that be anywhere near true with setups posted on here so far. Do you have dyno sheets and such and some specs so we can compare more closely the facts? We'd need to consider charge piping size, what manifold, turbo, inter-cooler size etc. I'd be surprised to see any substantial gain, or hell, at this point any gain at all. So far I've seen worse numbers on small frame stuff divided.
The gaps are small yes, they still exist, I'm aware of the metal used in this application, though as you're probably aware there's two other more expensive choices for the wheels too that are becoming used more often. There's also the core section which serves no purpose in terms of flowing air. Also as you know the first bit of blade is dull and wide for strength, this is the portion I was also mentioning being inefficient.
^ The point here was that if you take a very tiny turbo and compare it to say a GT35R looking at the compressor wheel section you'll note that when measuring total area that the amount of usable compressor wheel vs total area is better on the GT35R. It's simple math. That's not even considering the core section used for rotating the entire mass which shares the same characteristic. If we were to measure these factors you'd find there's a fair bit of difference in usable area vs wasted area. THIS is the factor I'm speaking of.
Twin turbos are often ditched partially for this reason though sometimes for the reasons you listed. It's not always easier to go single. Some cars it's much easier to pipe and setup a twin setup. In OE ultra performance setups twins are generally preferred yes and I support that measure, though they rely on cheaper turbos to accomplish what a better quality single could have done. For them they still look at things from a financial perspective, even ultra performance cars have to make a profit. If you want to talk efficiency let's just move on to variable pitch turbos, they win hands down. Yet they ....don't work effectively in small sizes due to engineering restraints, again, a matter of a technology unable to be accomplished in a small physical space. There will come a day no doubt, it's just not yet.
I believe the writers of these books have enough knowledge to be able to make the argument, it's not a conclusion that I came upon myself it was written as fact.
I've gone from a twin turbo to a single turbo in my previous vehicle, it gained all around. Though we can't use that example as the turbo I applied was of better quality, newer, etc.
Registered User
Joined: Nov 2008
Posts: 579
Likes: 0
Stepping up in terms of A/R is what I meant, though yes the entire turbo size would be a good move as well since it'll be more efficient on top and equally efficient on bottom due to the divided housing. So long as we're talking about a properly setup system and a large enough setup to begin with for it to be efficient.
You found small frame turbo did better divided? I've yet to see that be anywhere near true with setups posted on here so far. Do you have dyno sheets and such and some specs so we can compare more closely the facts? We'd need to consider charge piping size, what manifold, turbo, inter-cooler size etc. I'd be surprised to see any substantial gain, or hell, at this point any gain at all. So far I've seen worse numbers on small frame stuff divided.
The gaps are small yes, they still exist, I'm aware of the metal used in this application, though as you're probably aware there's two other more expensive choices for the wheels too that are becoming used more often. There's also the core section which serves no purpose in terms of flowing air. Also as you know the first bit of blade is dull and wide for strength, this is the portion I was also mentioning being inefficient.
^ The point here was that if you take a very tiny turbo and compare it to say a GT35R looking at the compressor wheel section you'll note that when measuring total area that the amount of usable compressor wheel vs total area is better on the GT35R. It's simple math. That's not even considering the core section used for rotating the entire mass which shares the same characteristic. If we were to measure these factors you'd find there's a fair bit of difference in usable area vs wasted area. THIS is the factor I'm speaking of.
Twin turbos are often ditched partially for this reason though sometimes for the reasons you listed. It's not always easier to go single. Some cars it's much easier to pipe and setup a twin setup. In OE ultra performance setups twins are generally preferred yes and I support that measure, though they rely on cheaper turbos to accomplish what a better quality single could have done. For them they still look at things from a financial perspective, even ultra performance cars have to make a profit. If you want to talk efficiency let's just move on to variable pitch turbos, they win hands down. Yet they ....don't work effectively in small sizes due to engineering restraints, again, a matter of a technology unable to be accomplished in a small physical space. There will come a day no doubt, it's just not yet.
I believe the writers of these books have enough knowledge to be able to make the argument, it's not a conclusion that I came upon myself it was written as fact.
I've gone from a twin turbo to a single turbo in my previous vehicle, it gained all around. Though we can't use that example as the turbo I applied was of better quality, newer, etc.
You found small frame turbo did better divided? I've yet to see that be anywhere near true with setups posted on here so far. Do you have dyno sheets and such and some specs so we can compare more closely the facts? We'd need to consider charge piping size, what manifold, turbo, inter-cooler size etc. I'd be surprised to see any substantial gain, or hell, at this point any gain at all. So far I've seen worse numbers on small frame stuff divided.
The gaps are small yes, they still exist, I'm aware of the metal used in this application, though as you're probably aware there's two other more expensive choices for the wheels too that are becoming used more often. There's also the core section which serves no purpose in terms of flowing air. Also as you know the first bit of blade is dull and wide for strength, this is the portion I was also mentioning being inefficient.
^ The point here was that if you take a very tiny turbo and compare it to say a GT35R looking at the compressor wheel section you'll note that when measuring total area that the amount of usable compressor wheel vs total area is better on the GT35R. It's simple math. That's not even considering the core section used for rotating the entire mass which shares the same characteristic. If we were to measure these factors you'd find there's a fair bit of difference in usable area vs wasted area. THIS is the factor I'm speaking of.
Twin turbos are often ditched partially for this reason though sometimes for the reasons you listed. It's not always easier to go single. Some cars it's much easier to pipe and setup a twin setup. In OE ultra performance setups twins are generally preferred yes and I support that measure, though they rely on cheaper turbos to accomplish what a better quality single could have done. For them they still look at things from a financial perspective, even ultra performance cars have to make a profit. If you want to talk efficiency let's just move on to variable pitch turbos, they win hands down. Yet they ....don't work effectively in small sizes due to engineering restraints, again, a matter of a technology unable to be accomplished in a small physical space. There will come a day no doubt, it's just not yet.
I believe the writers of these books have enough knowledge to be able to make the argument, it's not a conclusion that I came upon myself it was written as fact.
I've gone from a twin turbo to a single turbo in my previous vehicle, it gained all around. Though we can't use that example as the turbo I applied was of better quality, newer, etc.
Turbine wheel blades dont significantly decrease in thickness, proportionate to their size. I have dealt with tiny turbos like the GT15 up to big ones like the GT55, and I can tell you the leading edge does not get much thinner relative to the diameter of the wheel. The center hub of a turbine wheel also grows proportionately to the wheel diameter, and this is for LCF strength. Another thing to know, as the mass of the turbine wheel grows the more it will be affected by speed. Meaning a larger turbo needs to be stronger for a given size than a small wheel, because the larger will be more likely to tear itself apart at speed.
Now, there is one thing to know about the efficiency of turbine wheels. Small turbos are generally used on engines that see a variety of RPMs, and are meant to work happily at all of them. Large turbos are often used on steady state machines. One example is the GT45R, which was originally designed for a diesel generator. This generator stays at a fairly constant RPM and therefore the turbine wheel can be ultra efficient at the operating level it was designed for, but at a variety of revs its not a great turbine wheel.
Yes, a smaller turbo does have a minor disadvantage compared to larger turbos because hub diameter decreases disproportionately to the outer diameter of the inducer. Not only that, but the inlet area increases disproportionately to the diameter as well. But the compressor is not really the subject of the discussion nor the thread, so I wont delve more into that.
Yes, twins are sometimes easier to package, one example being V or W cylinder arrangements. Single turbos on either one can be an arduous job, and generally not a necessary one.
By ultra performance Im not talking about OEM, I mean the aftermarket segment. The quickest and fastest in professional racing run twins, because a single simply wont make the power. The new Turbonetics 122mm turbo (designed for sled pulling) is a beast at 2500hp potential but I know several guys running S500SX turbos that make over 1800 each. The S500SX much smaller, its easier to package two, and they make more power when combined. A single just does not any sense in this specific application, which is how every setup must be tailored.
When it comes to OE systems we can look at the BMW N54 and N55. The N54 is twin turbo, and the N55 a single twin scroll turbo. The N55 is cheaper, easier to package, spools slower, and makes less power. For cost and packaging the N55 is better, but for performance the N54 takes the cake.
Yes, the writers have good knowledge, but those books are designed for beginners. They will help you understand better how a turbo works, but just reading one does not turn someone into a Borg Warner engineer overnight. Ive never read a single book on turbocharging, but when it comes to the subject I know what works and what does not, and its all based on real world experience. My previous knowledge and understanding of physics and engineering helps a little too.
Would you put yourself in the hands of a heart surgeon who has never operated, but instead tries to console you with "but I read a book about it"? Probably not.
Registered User
Joined: Dec 2006
Posts: 24,119
Likes: 4
From: Seattle / Kalifornia
I agree with all the first part, that's why I said it's not worth the gain on a small/midframe rather than saying it's impossible for it to benefit from it. If the guy is going to run a GT42 I'd say no question. If he's gonna run a GT3071.... don't bother. Though many setups I've seen fail to be matched up properly enough for it even to start being worth discussing a dividing housing. Most are better off spending that money on a better manifold and making major gains.
I know what you're explaining to be true however the facts I stated are also still true and they don't produce equivalent efficiency as they decrease in size. That's the bottom line. Keep in mind smaller turbos generally survive faster speeds before ripping apart too, the 2.0 and 2.2 S2000 motors have a similar situation, a longer stroke but slower rpm. (though actually piston speed on the ap2 is faster, that's besides the point)
I've found I can source most turbos to diesel applications. I'm ASE certified for diesel trucks, I've seen a lot of turbos seeing a lot of boost for a lot of years. It's true about what you said to a point, an already spooled turbo doesn't need to vary in rpm as much thus mass matters less.
The disproportionate change in usable surface area such as the blow-by ring on the outside and shaft are the main points I'm pointing to. That was my entire point here really on that part of the topic.
If money is no object and a huge rpm range is needed a twin is better suited, if the cars got enough gears and can operate in a narrow range, go single still. If you have a V or W arrangement, I agree, nearly always twin or quad even is better suited, hell a quad divided housing on a monster engine would be great, every cylinder get's it's own tubing all the way to the wheel in that case.
Some of the books I've read are for beginners, others were really not. I hope you're not assuming you're the only one with real world experience. I've owned 7 turbochargers that I ran in my own vehicles at this point. That's not counting truck applications where I have a bit of drive time there too.
So I'm now a surgeon who has operated on a patient before AND went to med school, I don't see the disadvantage. I think you're making some assumptions that because I've done reading I don't have first hand experience, that's not the case here.
I know what you're explaining to be true however the facts I stated are also still true and they don't produce equivalent efficiency as they decrease in size. That's the bottom line. Keep in mind smaller turbos generally survive faster speeds before ripping apart too, the 2.0 and 2.2 S2000 motors have a similar situation, a longer stroke but slower rpm. (though actually piston speed on the ap2 is faster, that's besides the point)
I've found I can source most turbos to diesel applications. I'm ASE certified for diesel trucks, I've seen a lot of turbos seeing a lot of boost for a lot of years. It's true about what you said to a point, an already spooled turbo doesn't need to vary in rpm as much thus mass matters less.
The disproportionate change in usable surface area such as the blow-by ring on the outside and shaft are the main points I'm pointing to. That was my entire point here really on that part of the topic.
If money is no object and a huge rpm range is needed a twin is better suited, if the cars got enough gears and can operate in a narrow range, go single still. If you have a V or W arrangement, I agree, nearly always twin or quad even is better suited, hell a quad divided housing on a monster engine would be great, every cylinder get's it's own tubing all the way to the wheel in that case.
Some of the books I've read are for beginners, others were really not. I hope you're not assuming you're the only one with real world experience. I've owned 7 turbochargers that I ran in my own vehicles at this point. That's not counting truck applications where I have a bit of drive time there too.
So I'm now a surgeon who has operated on a patient before AND went to med school, I don't see the disadvantage. I think you're making some assumptions that because I've done reading I don't have first hand experience, that's not the case here.
Thread
Thread Starter
Forum
Replies
Last Post
TwistedFabrication
S2000 Forced Induction
5
Sep 3, 2014 08:50 AM