Equal length turbo header: Yay or Nay?
Moderators: The Dark Side of Will, Series8217
Equal length turbo header: Yay or Nay?
I've got a divided turbine inlet on my GT, so I've been contemplating the effects of equal length or just a regular header on spool up time. I've heard that with the divided housing it's usually beneficial to route the interfering pulses into the opposing turbine inlets. Any downsides? I can deal with the complexities of an equal length setup, and I think I have enough room, but is it worth it in a turbocharged application? Regardless of the configuration it's going to be of Inconel 625 construction to deal with the stress. Thoughts, speculations?
Nate
Nate
Indy DOHC Turbo SD4.....someday.
Oh, and f*ck the envelope. (RFT Insurgent)
Oh, and f*ck the envelope. (RFT Insurgent)
I've actually been reading into this recently as I am getting contracted to build some turbo headers for a Nissan.
If I were you, I would get the exhaust gases to the turbo with as little piping as possible. Turbochargers like hot exhaust, not cold exhaust, and long tube equal length headers tend to cool the exhaust a lot. In addition, turbochargers spool faster and smoother with a pulsing exhaust, not smooth flow. Long tube equal length headers tend to smooth out the pulses, moreso the more cylinders you have.
So if possible, build headers, but use shorty style with a smooth collector, not manifolds, but not long tube equal length.
I hope this helps. What motor will it be for?
If I were you, I would get the exhaust gases to the turbo with as little piping as possible. Turbochargers like hot exhaust, not cold exhaust, and long tube equal length headers tend to cool the exhaust a lot. In addition, turbochargers spool faster and smoother with a pulsing exhaust, not smooth flow. Long tube equal length headers tend to smooth out the pulses, moreso the more cylinders you have.
So if possible, build headers, but use shorty style with a smooth collector, not manifolds, but not long tube equal length.
I hope this helps. What motor will it be for?
88GT 3.4 DOHC Turbo
Gooch wrote:Way to go douche. You are like a one-man, fiero-destroying machine.
It's for a Super Duty 4cyl, 2.6l. I'm pretty sure if I do an equal length header I can keep it to around 18-20" to the header flange from the port. Plus I'm contemplating ceramic coating the exhaust ports. Also, Inconel has a very, very low thermal conductivity, so I think I will be able to retain heat very well if I stick to these.
I figured that the equal length header would enhance the pulses basically until they hit the turbine wheel. Hmmm...
I figured that the equal length header would enhance the pulses basically until they hit the turbine wheel. Hmmm...
The header will smooth the pulses out, which is great for N/A, but not generally for a turbo. I'm not saying your car won't ever be quick if you have an equal length header, I'm just saying the hosrty design might be good for 10-15 more hp, it would be cheaper, and easier to build.Indy wrote:It's for a Super Duty 4cyl, 2.6l. I'm pretty sure if I do an equal length header I can keep it to around 18-20" to the header flange from the port. Plus I'm contemplating ceramic coating the exhaust ports. Also, Inconel has a very, very low thermal conductivity, so I think I will be able to retain heat very well if I stick to these.
I figured that the equal length header would enhance the pulses basically until they hit the turbine wheel. Hmmm...
The length from the port isn't totally critical, but the closer the better.
I disagree.
Aaron wrote: The header will smooth the pulses out, which is great for N/A, but not generally for a turbo. I'm not saying your car won't ever be quick if you have an equal length header, I'm just saying the hosrty design might be good for 10-15 more hp, it would be cheaper, and easier to build.
The length from the port isn't totally critical, but the closer the better.
uhh....ok?
That is the entire point of the equal length header. The exhaust is traveling through its respective primary, and then at the collector, all 3 primaries meet, but not all 3 pulses are the same distance down the pipe, since they didn't start at the same time. Then at the collector they fall one behind the other so to say, therefore it equalizes them to a point.
For some reason turbochargers react better to a heavily pulsing exhaust rather than a smooth flow that an equal length ehader will give. It will not be perfectly smooth, but a lot smoother than a shorty header.
That is the entire point of the equal length header. The exhaust is traveling through its respective primary, and then at the collector, all 3 primaries meet, but not all 3 pulses are the same distance down the pipe, since they didn't start at the same time. Then at the collector they fall one behind the other so to say, therefore it equalizes them to a point.
For some reason turbochargers react better to a heavily pulsing exhaust rather than a smooth flow that an equal length ehader will give. It will not be perfectly smooth, but a lot smoother than a shorty header.
-
- cant get enough of this site!
- Posts: 3289
- Joined: Wed Mar 30, 2005 2:37 pm
lookup the aerotech that fueling did with the magnesium block quad 4 2.2l that thing had some serious headers on it, and they look equal length and fueling was a madman with turbos. I disagree with the fact that you will lose 10 -15 hp. Especially with a header made of iconel and ceramic coated (didnt bryson do his own coatings on pistons) You may be able to do them yourself.
im not a turbo guy, but I say equal length and keep them short as possible. But thats more a guess than a recomendation.
im not a turbo guy, but I say equal length and keep them short as possible. But thats more a guess than a recomendation.
-
- Peer Mediator
- Posts: 15631
- Joined: Wed Nov 24, 2004 11:13 pm
- Location: In the darkness, where fear and knowing are one
- Contact:
You'll get better spool by introducing the exhaust pulses to the turbos at regular intervals--ESPECIALLY with a dual scroll turbine housing... this only happens with equal length manifolds.Aaron wrote:uhh....ok?
That is the entire point of the equal length header. The exhaust is traveling through its respective primary, and then at the collector, all 3 primaries meet, but not all 3 pulses are the same distance down the pipe, since they didn't start at the same time. Then at the collector they fall one behind the other so to say, therefore it equalizes them to a point.
For some reason turbochargers react better to a heavily pulsing exhaust rather than a smooth flow that an equal length ehader will give. It will not be perfectly smooth, but a lot smoother than a shorty header.
I agree that this only happens with equal length headers, manifolds will still puls by the way, but I disagree that this is a good thing.The Dark Side of Will wrote:You'll get better spool by introducing the exhaust pulses to the turbos at regular intervals--ESPECIALLY with a dual scroll turbine housing... this only happens with equal length manifolds.Aaron wrote:uhh....ok?
That is the entire point of the equal length header. The exhaust is traveling through its respective primary, and then at the collector, all 3 primaries meet, but not all 3 pulses are the same distance down the pipe, since they didn't start at the same time. Then at the collector they fall one behind the other so to say, therefore it equalizes them to a point.
For some reason turbochargers react better to a heavily pulsing exhaust rather than a smooth flow that an equal length ehader will give. It will not be perfectly smooth, but a lot smoother than a shorty header.
According to "Forced Incuction Performance Tuning," a book by A. Graham Bell, "From my personal experience with the Cosworth turbo 4 cylinder racing engines, the turbochargs seemed to react better with a pulsing exhaust flow over a steady flow. Therefore we engineered a header that optimized these pulses, using alternate primary lengths for each individual cylinder. After the completion of numerous headers, all slightly different in design, the following header offeed the best overall drivability, coincidently it also offered the highest horsepower number. This header used a single 12", 10", 8", and 6" primary brought together in a 6.5" long 4-1 collector. All 4 primaries had to be in order for the best results, but can be in any cylinder so long as the firing order is 1234, which it was. Such as, for cylinders 1-2-3-4, you can have 12, 10, 8, and 6, or 8, 6, 12, 10, so long as they are in order."
This is the only important aspect, he also gives dyno numbers on the same engine, with 3 different headers, a equal length 4-1, a equilength 4-2-1, and the one described.
-
- Peer Mediator
- Posts: 15631
- Joined: Wed Nov 24, 2004 11:13 pm
- Location: In the darkness, where fear and knowing are one
- Contact:
How long were the runners in the other manifolds? Significantly longer runners in the other manifolds could potentially lose enough more heat to compromise the power of the long manifold relative to the short one.
Also, what was the power and driveability difference? Are talking 10 HP on 500 and 200 RPM out of 10,000 worth of spool?
Also, what was the power and driveability difference? Are talking 10 HP on 500 and 200 RPM out of 10,000 worth of spool?
Last edited by The Dark Side of Will on Mon Aug 15, 2005 10:57 pm, edited 1 time in total.
The 4-1 equil length used all 4 at exactly 33 inches in length, and the 4-2-1 used 21" long runners, which according to Bell was the optimum for each set.
I didn't quote this book becuase it is the one that agrees with my point, I chose it becuz it is one of the few I have, and the only that discusses primary length with respect to turbochargers. I would liek to read more upon this though, but I have yet to find a book that discusses it.
And for everyone that preaches maximum boost, I really found this book MUCH more technical and much more informative.
I didn't quote this book becuase it is the one that agrees with my point, I chose it becuz it is one of the few I have, and the only that discusses primary length with respect to turbochargers. I would liek to read more upon this though, but I have yet to find a book that discusses it.
And for everyone that preaches maximum boost, I really found this book MUCH more technical and much more informative.
Any?
I'd say he has nearly 200, from all different engines, from the Duratec 2.5ls, to Cosworth race engines, small block Chevy's, the GM 3.8 V6, and even a 60* V6. There are dyno charts throughout the entire book. And what I love is how thoroughly he examines everyone. He is such a perfectionist, that he gives a shit about 3whp, calling one setup superior becuz it made 3whp (Which is almost negligable), more.
He is also REALLY big on drivability, since he comes from Rally and road racing, where the all out power isn't as essential as the power curve and how it comes on.
I'd say he has nearly 200, from all different engines, from the Duratec 2.5ls, to Cosworth race engines, small block Chevy's, the GM 3.8 V6, and even a 60* V6. There are dyno charts throughout the entire book. And what I love is how thoroughly he examines everyone. He is such a perfectionist, that he gives a shit about 3whp, calling one setup superior becuz it made 3whp (Which is almost negligable), more.
He is also REALLY big on drivability, since he comes from Rally and road racing, where the all out power isn't as essential as the power curve and how it comes on.
Yah, there is a chart comparing the 3, but I can't scan it in, and I don't have the book anymore.
I leave for college tomorrow morning, and it got packed about 2 hours ago, and is as of right now, BURIED. Sorry.
I do remember from it though that the 4-2-1 offered the highest average torque OFF BOOST (Below 4000rpm, redline 9), but once boost came on the 12-10-8-6 made more power. It wasn't much more power, but enough to make him write about it. THe max power was some 14-15hp higher, and maz torque was about 8 higher IIRC.
I leave for college tomorrow morning, and it got packed about 2 hours ago, and is as of right now, BURIED. Sorry.
I do remember from it though that the 4-2-1 offered the highest average torque OFF BOOST (Below 4000rpm, redline 9), but once boost came on the 12-10-8-6 made more power. It wasn't much more power, but enough to make him write about it. THe max power was some 14-15hp higher, and maz torque was about 8 higher IIRC.
-
- Peer Mediator
- Posts: 15631
- Joined: Wed Nov 24, 2004 11:13 pm
- Location: In the darkness, where fear and knowing are one
- Contact:
How much HP was the engine making?Aaron wrote: I do remember from it though that the 4-2-1 offered the highest average torque OFF BOOST (Below 4000rpm, redline 9), but once boost came on the 12-10-8-6 made more power. It wasn't much more power, but enough to make him write about it. THe max power was some 14-15hp higher, and maz torque was about 8 higher IIRC.
Are we talking 15 HP on 300 (5%) or 15 HP on 600 (2.5%)
Here's the rub. Among 12, 10, 8 and 6, the average is 9. So he's got the same total tubing length as a 4-1 with 9" primaries and 6" collector... 15".Aaron wrote:The 4-1 equil length used all 4 at exactly 33 inches in length, and the 4-2-1 used 21" long runners, which according to Bell was the optimum for each set.
The equal length used 33" primaries PLUS however much collector.
The 4-2-1 had 21" primaries, but probably 12" secondaries (33 total) PLUS collector length again...
So comparing those two manifolds to the unequal length manifold, you're talking about 15" of tubing vs probably close to 40... I think that there would be enough HEAT lost in that extra length of the equal length manifolds to account for the differences in both power and driveability.
IOW, his experimental technique was flawed and he really didn't show anything.
A better experiment would have been to team an unequal length manifold with an average length the same as that of the competing designs... that would have showcased the differences between equal length and unequal length better.
Can't fault him for doing it the way he did, though. He's a racer, not a scientist. His job is to find what works best and use it... not waste time and resources figuring out the exact effect everything has on everything else when most of that data isn't useful for going fast.
That's the thing...I'm pretty sure I can keep a very short primary length. I'm doing custom wiring and insulating the isht out of the firewall, so I have some room to keep the header pretty ideal. I'm thinking I can manage at least a 20" primary length. Secondary length will be fairly negligible if I can get it built right.