Effect of equal-length exhaust headers for the 991.2
11-09-2017, 01:04 AM
#1
Rennlist Member
Thread Starter
Effect of equal-length exhaust headers for the 991.2
I'm curious to learn what effects equal-length headers would have on the 991.2 cars in terms of the exhaust note. Would one expect a smoother tone at idle?
11-09-2017, 01:45 AM
#2
Rennlist Member
I will never forget a shop with a star welder (and he was a star...his welds were pure art) when I was looking at their latest project, headers for the 2.5-liter Boxster. They were ridiculing the design of the factory headers, which did look pretty awful and very cheap. Each manifold pipe had an oval, "wine-glass" enlargement as it approached the exhaust port. The rest of the factory headers looked like junk. The new headers were gorgeous. All stainless, perfect welds, equal-length, just really pretty. And they clearly looked like they would flow a lot better, too. Nice collectors, based on decades of building race exhausts (that worked). Then I never heard about the system again. So I asked a local dyno operator, who said, "Oh yeah. That system. It lost a lot of power on the dyno, and we couldn't seem to get much of it back."
Mufflers and cats are something else, and I suspect there are gains to be made there in terms of power and better noises. I think the first tuner to do better cats for non-PSE 991.2s with factory S tips in black (or compatibility) and either no mufflers or some sort of X pipe that sounds great will get a lot of business. I skipped PSE because I don't like the looks, and don't love the sound…
Last edited by stout; 11-09-2017 at 02:00 AM.
11-09-2017, 08:31 AM
#3
RL Community Team
Rennlist Member
Rennlist Member
I'll start off by saying I've waded into the shallow end of the kiddie pool of turbo tuning world. Certainly there are some pretty insane turbo builds out there that I just don't follow.
Having said that, you need to be careful with your exhaust between the engine and the turbo, or it will be very easy to introduce lag. The potential to lose exhaust energy (pun intended) is extremely high, and that energy is what spins the turbo. Most manufacturers are moving the turbo as close as possible to the engine, even integrating the manifold into the block, to retain as much heat and mechanical exhaust energy as possible. Back in my Volvo days, though, there was someone who crafted an equal length header for then5 cylinder turbo. Someone here might know more....
The results are not always pleasant sounding. Listen to the farty F80 M3. The double turbo setup results in a fairly unpleasant and distinctively in-straight six sounding exhaust noise. The single turbo B58 sounds loads better. And the S63 in the M5 uses a pretty trick crossover manifold where each turbo is fed from both banks of cylinders to eliminate lag...and the result is an exhaust that sounds more V10 than V8. The Mercedes 4.0 V8 is the current AMG models doesn't use that and sounds worlds better.
The 991.2 sounds very much like a flat 6. Just muted. IMO the gains are turbo back. Listen to the Kline exhaust. It sounds great. Of course it's catless but perhaps their sport cat set will sound as good.
Having said that, you need to be careful with your exhaust between the engine and the turbo, or it will be very easy to introduce lag. The potential to lose exhaust energy (pun intended) is extremely high, and that energy is what spins the turbo. Most manufacturers are moving the turbo as close as possible to the engine, even integrating the manifold into the block, to retain as much heat and mechanical exhaust energy as possible. Back in my Volvo days, though, there was someone who crafted an equal length header for then5 cylinder turbo. Someone here might know more....
The results are not always pleasant sounding. Listen to the farty F80 M3. The double turbo setup results in a fairly unpleasant and distinctively in-straight six sounding exhaust noise. The single turbo B58 sounds loads better. And the S63 in the M5 uses a pretty trick crossover manifold where each turbo is fed from both banks of cylinders to eliminate lag...and the result is an exhaust that sounds more V10 than V8. The Mercedes 4.0 V8 is the current AMG models doesn't use that and sounds worlds better.
The 991.2 sounds very much like a flat 6. Just muted. IMO the gains are turbo back. Listen to the Kline exhaust. It sounds great. Of course it's catless but perhaps their sport cat set will sound as good.
11-09-2017, 02:10 PM
#4
I'll start off by saying I've waded into the shallow end of the kiddie pool of turbo tuning world. Certainly there are some pretty insane turbo builds out there that I just don't follow.
Having said that, you need to be careful with your exhaust between the engine and the turbo, or it will be very easy to introduce lag. The potential to lose exhaust energy (pun intended) is extremely high, and that energy is what spins the turbo. Most manufacturers are moving the turbo as close as possible to the engine, even integrating the manifold into the block, to retain as much heat and mechanical exhaust energy as possible. Back in my Volvo days, though, there was someone who crafted an equal length header for then5 cylinder turbo. Someone here might know more....
The results are not always pleasant sounding. Listen to the farty F80 M3. The double turbo setup results in a fairly unpleasant and distinctively in-straight six sounding exhaust noise. The single turbo B58 sounds loads better. And the S63 in the M5 uses a pretty trick crossover manifold where each turbo is fed from both banks of cylinders to eliminate lag...and the result is an exhaust that sounds more V10 than V8. The Mercedes 4.0 V8 is the current AMG models doesn't use that and sounds worlds better.
The 991.2 sounds very much like a flat 6. Just muted. IMO the gains are turbo back. Listen to the Kline exhaust. It sounds great. Of course it's catless but perhaps their sport cat set will sound as good.
Having said that, you need to be careful with your exhaust between the engine and the turbo, or it will be very easy to introduce lag. The potential to lose exhaust energy (pun intended) is extremely high, and that energy is what spins the turbo. Most manufacturers are moving the turbo as close as possible to the engine, even integrating the manifold into the block, to retain as much heat and mechanical exhaust energy as possible. Back in my Volvo days, though, there was someone who crafted an equal length header for then5 cylinder turbo. Someone here might know more....
The results are not always pleasant sounding. Listen to the farty F80 M3. The double turbo setup results in a fairly unpleasant and distinctively in-straight six sounding exhaust noise. The single turbo B58 sounds loads better. And the S63 in the M5 uses a pretty trick crossover manifold where each turbo is fed from both banks of cylinders to eliminate lag...and the result is an exhaust that sounds more V10 than V8. The Mercedes 4.0 V8 is the current AMG models doesn't use that and sounds worlds better.
The 991.2 sounds very much like a flat 6. Just muted. IMO the gains are turbo back. Listen to the Kline exhaust. It sounds great. Of course it's catless but perhaps their sport cat set will sound as good.
You're not off on your comments. There's many levels of considerations with headers and arguably for the Porsche engineers, it was drive ability. The sound difference with any header swap on our car will be minimal.
The stock monolith design has unequal runners and I believe that's to help low end spool and the elimination of lag. The elimination of any sense of lag with these cars is what I think was prioritized by the engineers on the 9A2 engines. Placing the turbos as close to the block as possible is obviously a way of combating this, the other is unequal length runners. Simply stated, the closer the exhaust pulse is to the turbo, the faster it will spool. The unequal lengths of the runners really isn't an issue with the vehicles at stock power levels. I think from my personal experience and those 991.1 owners who've driven a 991.2 and commented, it feels as if the turbo's don't pull hard to redline the same way a NA motor does... this could be argued, as a result of the header design. Its effectively choking up the combustion cycle of the motor at high rpms.
I personally believe swapping out headers at this stage of the game where there is basically zero information on their impact is premature. The whole next level of this equation is if one wants to tune their vehicle. Often the trade offs for swapping cast/log/monolith style headers for tubular equal length headers while running stock will not be apparent when tuned.
Typically what you see when your increasing runner length and making them equal in length is a shift in the torque curve. You may loose some initial low end spool, but the tradeoff is much more mid range torque and higher overall tq/hp numbers.
The benefits should also be exaggerated once tuned. The volume of exhaust gases moving through the turbo will be higher, thus the low spool issue shouldnt be an issue and the increased and balanced flow should facilitate increases.
11-09-2017, 03:55 PM
#7
Rennlist Member
Am I the only one bothered by catalyst delete on street cars?
Seems like an unfortunate mod given how little power is lost with catalysts on modern engines these days. And then you have the legal issues, the CELs, etc....
Seems like an unfortunate mod given how little power is lost with catalysts on modern engines these days. And then you have the legal issues, the CELs, etc....
Trending Topics
11-09-2017, 11:31 PM
#9
RL Community Team
Rennlist Member
Rennlist Member
11-10-2017, 02:04 AM
#12
The cost savings come from: not have the exhaust manifold, one less gasket, many less fasteners.
There's also less weight.
I do believe with the integrated manifold in the head, there is less volume and therefore should help some with transient response. But it does choke off top-end power.
11-10-2017, 02:36 AM
#13
You're not off on your comments. There's many levels of considerations with headers and arguably for the Porsche engineers, it was drive ability. The sound difference with any header swap on our car will be minimal.
The stock monolith design has unequal runners and I believe that's to help low end spool and the elimination of lag. The elimination of any sense of lag with these cars is what I think was prioritized by the engineers on the 9A2 engines. Placing the turbos as close to the block as possible is obviously a way of combating this, the other is unequal length runners. Simply stated, the closer the exhaust pulse is to the turbo, the faster it will spool. The unequal lengths of the runners really isn't an issue with the vehicles at stock power levels. I think from my personal experience and those 991.1 owners who've driven a 991.2 and commented, it feels as if the turbo's don't pull hard to redline the same way a NA motor does... this could be argued, as a result of the header design. Its effectively choking up the combustion cycle of the motor at high rpms.
I personally believe swapping out headers at this stage of the game where there is basically zero information on their impact is premature. The whole next level of this equation is if one wants to tune their vehicle. Often the trade offs for swapping cast/log/monolith style headers for tubular equal length headers while running stock will not be apparent when tuned.
Typically what you see when your increasing runner length and making them equal in length is a shift in the torque curve. You may loose some initial low end spool, but the tradeoff is much more mid range torque and higher overall tq/hp numbers.
The benefits should also be exaggerated once tuned. The volume of exhaust gases moving through the turbo will be higher, thus the low spool issue shouldnt be an issue and the increased and balanced flow should facilitate increases.
The stock monolith design has unequal runners and I believe that's to help low end spool and the elimination of lag. The elimination of any sense of lag with these cars is what I think was prioritized by the engineers on the 9A2 engines. Placing the turbos as close to the block as possible is obviously a way of combating this, the other is unequal length runners. Simply stated, the closer the exhaust pulse is to the turbo, the faster it will spool. The unequal lengths of the runners really isn't an issue with the vehicles at stock power levels. I think from my personal experience and those 991.1 owners who've driven a 991.2 and commented, it feels as if the turbo's don't pull hard to redline the same way a NA motor does... this could be argued, as a result of the header design. Its effectively choking up the combustion cycle of the motor at high rpms.
I personally believe swapping out headers at this stage of the game where there is basically zero information on their impact is premature. The whole next level of this equation is if one wants to tune their vehicle. Often the trade offs for swapping cast/log/monolith style headers for tubular equal length headers while running stock will not be apparent when tuned.
Typically what you see when your increasing runner length and making them equal in length is a shift in the torque curve. You may loose some initial low end spool, but the tradeoff is much more mid range torque and higher overall tq/hp numbers.
The benefits should also be exaggerated once tuned. The volume of exhaust gases moving through the turbo will be higher, thus the low spool issue shouldnt be an issue and the increased and balanced flow should facilitate increases.
But these 991.2 headers in particular I feel are particularly poor, they are basically a log manifold with a single runner sticking out the side of it to feed exhaust to the turbo. If one were to ask anyone to draw a 3-runner manifold for optimum flow, no one would draw what the factory manifold looks like. The factory manifold does not have any apparent effort put into the design to create a nice merge or even turn the flow nicely into the stub sticking out going to the turbos. I imagine the design requirements were more focused on cost (for obvious reasons) and volume (minimal volume for quicker cat light-off. I imagine the stock header is double walled stainless to better retain heat for cat light-off, and therefore heavier than an aftermarket tubular header.
The tests the OEMs really need to do well on are emissions and fuel economy/CO2 per km/mile. The stock basic log type manifolds meet those requirements. A 3.0L turbo engine making only 370hp-450hp is really quite under-stressed, so for Porsche to hit their power targets is a simple matter of adjusting the boost level. Therefore, an inefficient header is really not an issue at these power levels.
I believe a nice equal length header with a proper merge collector will improve spool and power because the stock design is just that poor. Even if not equal length, just a better merge would be a big improvement. I suspect we will see a good equal length header with good merge come to market.
11-11-2017, 12:07 PM
#14
I will disagree with some of your assessment of these particular 991.2 headers. I do think your assessment is accurate to a large number of log-type manifolds. Even the McLarens have loggish type exhaust manifolds.
But these 991.2 headers in particular I feel are particularly poor, they are basically a log manifold with a single runner sticking out the side of it to feed exhaust to the turbo. If one were to ask anyone to draw a 3-runner manifold for optimum flow, no one would draw what the factory manifold looks like. The factory manifold does not have any apparent effort put into the design to create a nice merge or even turn the flow nicely into the stub sticking out going to the turbos. I imagine the design requirements were more focused on cost (for obvious reasons) and volume (minimal volume for quicker cat light-off. I imagine the stock header is double walled stainless to better retain heat for cat light-off, and therefore heavier than an aftermarket tubular header.
The tests the OEMs really need to do well on are emissions and fuel economy/CO2 per km/mile. The stock basic log type manifolds meet those requirements. A 3.0L turbo engine making only 370hp-450hp is really quite under-stressed, so for Porsche to hit their power targets is a simple matter of adjusting the boost level. Therefore, an inefficient header is really not an issue at these power levels.
I believe a nice equal length header with a proper merge collector will improve spool and power because the stock design is just that poor. Even if not equal length, just a better merge would be a big improvement. I suspect we will see a good equal length header with good merge come to market.
But these 991.2 headers in particular I feel are particularly poor, they are basically a log manifold with a single runner sticking out the side of it to feed exhaust to the turbo. If one were to ask anyone to draw a 3-runner manifold for optimum flow, no one would draw what the factory manifold looks like. The factory manifold does not have any apparent effort put into the design to create a nice merge or even turn the flow nicely into the stub sticking out going to the turbos. I imagine the design requirements were more focused on cost (for obvious reasons) and volume (minimal volume for quicker cat light-off. I imagine the stock header is double walled stainless to better retain heat for cat light-off, and therefore heavier than an aftermarket tubular header.
The tests the OEMs really need to do well on are emissions and fuel economy/CO2 per km/mile. The stock basic log type manifolds meet those requirements. A 3.0L turbo engine making only 370hp-450hp is really quite under-stressed, so for Porsche to hit their power targets is a simple matter of adjusting the boost level. Therefore, an inefficient header is really not an issue at these power levels.
I believe a nice equal length header with a proper merge collector will improve spool and power because the stock design is just that poor. Even if not equal length, just a better merge would be a big improvement. I suspect we will see a good equal length header with good merge come to market.
11-11-2017, 02:05 PM
#15
Burning Brakes
Just a note, OEMs integrate the exhaust manifold into the head of the engine for two primary reasons: emissions and costs. Traditionally, there's a large and heavy exhaust manifold between the head and the turbo. The exhaust manifold acts as a big thermal sink soaking up a lot of heat. This prolongs cold start and therefore increases cold start emissions.
The cost savings come from: not have the exhaust manifold, one less gasket, many less fasteners.
There's also less weight.
I do believe with the integrated manifold in the head, there is less volume and therefore should help some with transient response. But it does choke off top-end power.
The cost savings come from: not have the exhaust manifold, one less gasket, many less fasteners.
There's also less weight.
I do believe with the integrated manifold in the head, there is less volume and therefore should help some with transient response. But it does choke off top-end power.
As usual, there is no free lunch.