Anti-reversion and twin scroll
Discussion
I just read a year old thread on "perfect' turbo exhaust manifold design. A lot of it was interesting, but it seemed to miss a couple obvious questions.
I'm in the process of designing an exhaust manifold which I think will be, not unique, but at least mildly innovative. I went online and it looks like there are people who can 3D print objects that can be used as the base for investment casting. I live in Loveland CO in the US where there is a lot of art casting done, and one of the companies can cast stainless steel.
So, my plan is to create a manifold design in Solidworks then have it 3D printed and investment cast. I think it will actually end up easier for me and give me more design options.
This as all for an SCCA D Modified autocross engine based on a GM 1.4 turbo motor. The engines have to breath through a 33mm restrictor, so chocked flow will be around 27 lb/min and max power will be around 270 hp. The turbo I picked is the twin-scroll Borg Warner EFR 6258.
I expect to have more, but my first question is on anti-reversion. For a 4 cylinder motor I believe it's tough, or impossible, to not have one cylinder valve still open at the end of the exhaust cycle when the next pressure pulse comes out of a port where the exhaust valve has just opened. That's the whole reason for twin-scroll. Three cylinders are supposed to be about perfect, so in the case of a 4 cylinder twin scroll, there should be a dead zone when no exhaust valve is open at all.
In this situation, why would we care about anti-reversion at all? I can see where a very long primary might let you use a low pressure return wave for extraction, although I think it would be very long and I'm much more interested in short primaries for reduced lag.
So, anti-reversion, or not?
I'm in the process of designing an exhaust manifold which I think will be, not unique, but at least mildly innovative. I went online and it looks like there are people who can 3D print objects that can be used as the base for investment casting. I live in Loveland CO in the US where there is a lot of art casting done, and one of the companies can cast stainless steel.
So, my plan is to create a manifold design in Solidworks then have it 3D printed and investment cast. I think it will actually end up easier for me and give me more design options.
This as all for an SCCA D Modified autocross engine based on a GM 1.4 turbo motor. The engines have to breath through a 33mm restrictor, so chocked flow will be around 27 lb/min and max power will be around 270 hp. The turbo I picked is the twin-scroll Borg Warner EFR 6258.
I expect to have more, but my first question is on anti-reversion. For a 4 cylinder motor I believe it's tough, or impossible, to not have one cylinder valve still open at the end of the exhaust cycle when the next pressure pulse comes out of a port where the exhaust valve has just opened. That's the whole reason for twin-scroll. Three cylinders are supposed to be about perfect, so in the case of a 4 cylinder twin scroll, there should be a dead zone when no exhaust valve is open at all.
In this situation, why would we care about anti-reversion at all? I can see where a very long primary might let you use a low pressure return wave for extraction, although I think it would be very long and I'm much more interested in short primaries for reduced lag.
So, anti-reversion, or not?
Isn't anti reversion where the inside diameter of the primary collectors is greater than that of the exhaust port. Thus it gives a step..... Makes not a huge difference to the out flowing exhaust pulse, but provides resistance to any reflected pressure waves
For a 4-1 collector the benefits will be grater, but still even with a twin scroll there is likely to be higher than 1:1 pressure ratio between the inlet manifold and the exhaust manifold.
I can't see any drawbacks to be honest,put in the steps!
For a 4-1 collector the benefits will be grater, but still even with a twin scroll there is likely to be higher than 1:1 pressure ratio between the inlet manifold and the exhaust manifold.
I can't see any drawbacks to be honest,put in the steps!
Edited by McVities on Thursday 21st June 23:39
Anything that can help flow = good. Anything that can help prevent reversion = good.
No matter what the end goal, basic principals to help efficiency will always be good if you can implement them.
But even the twin scroll aspect, how much use you could make of such things or how good/bad they might be would also depend on valve timing too and as above mentioned what sort of pre-turbine pressures you'd see.
Although I would say that if it is a good setup, you would be hoping to see less than 1:1 PR, not above it. Old restrictive inefficient setups might see high pre-turbine pressures, but in this day and age there's little excuse for it unless there is something with the build that necessitates it.
Even my crappy basic twin turbo V8, most of the time I'm better than 1:1 and it's simple cast manifolds.
As for the process, it all sounds good. I think some of the GTR manifolds in the aftermarket are made the same way. As you say, the 3d printing process can really open up options that would have been very complex before and you could probably get a cast piece with well finished flow paths and not need to have to manually port them etc
No matter what the end goal, basic principals to help efficiency will always be good if you can implement them.
But even the twin scroll aspect, how much use you could make of such things or how good/bad they might be would also depend on valve timing too and as above mentioned what sort of pre-turbine pressures you'd see.
Although I would say that if it is a good setup, you would be hoping to see less than 1:1 PR, not above it. Old restrictive inefficient setups might see high pre-turbine pressures, but in this day and age there's little excuse for it unless there is something with the build that necessitates it.
Even my crappy basic twin turbo V8, most of the time I'm better than 1:1 and it's simple cast manifolds.
As for the process, it all sounds good. I think some of the GTR manifolds in the aftermarket are made the same way. As you say, the 3d printing process can really open up options that would have been very complex before and you could probably get a cast piece with well finished flow paths and not need to have to manually port them etc
I'm looking for corrections and amplifications of how I understand this, because I assume I'm missing stuff.
Going back to basics, the two cylinder attached to one scroll would align cycles for instance like:
cylinder 2 _______ cylinder 3
intake ___________ power
compression ___ exhaust
power ___________ intake
exhaust _________ compression
The exhaust cycle from one cylinder is so separated from the other that each operates more or less as a single cylinder engine. Overlap and duration aren't enough to cause pressure wave interaction between the two cylinders, especially since I'm planning on short runners. It looks like it will be about 4" between the head and the twin scroll turbo inlet flange. The runners from 1 and 4 look like they will be around 7" long, with 2 and 3 somewhat shorter.
I'm trying to think about what the pressure at an exhaust valve will look like through the time the exhaust valve is open. Starting when the exhaust opens a burst of high pressure gas enters the runner. Velocity should get to sonic very quickly at the valve, so pressure in the runner should start rising quickly. Note, I'm kind of assuming pressure waves in such short runners get bounced and confused fast enough that they can be ignored, since by the time the valve starts to close, which is when you care most, the reflections have all died out.
Through a good part of the exhaust cycle the flow stays sonic, but the pressure in the cylinder is dropping. Sonic flow is directly proportional to the inlet pressure, so flow and the rate of pressure rise in the runner would start to fade, but at the same time the valve opening has been getting bigger. Towards the end of the exhaust cycle the pressure drop shouldn't be enough to maintain sonic flow and the valve is now closing, but there will still be some flow out.
All of this though ignores the rate the gas flows out of the runner and into the turbo. I'm guessing that flow out of the runner is proportional to the rpm of the turbo and the pressure and density of gas in the runner. The speed of the turbo will be effectively constant through a single exhaust cycle, but the gas pressure and density wont's be.
It does seem like the turbo won't be able to fully absorb all the gas in the runner. As the exhaust valve closes there will be some residual pressure that will continue to drop for the next engine cycle. This is the cycle were the cylinder we are looking at is pulling in the next charge, but there is no cylinder exhausting into the runner we're looking at. On the cycle following what you might call this exhaust dead cycle, the other cylinder attached to the scroll enters it's exhaust cycle and the pressure in the runner goes back up.
The question is, how low is the pressure in the cylinder and the exhaust runner when an exhaust valve closes. I think that the turbo will not be able to absorb all the exhaust gas by this point, because if it did the turbo would continue pulling gas out of the runner during the dead cycle and the pressure would go towards vacuum.
As the exhaust valve closes, will there be reversion? It doesn't seem like there would be based on pressure waves with such a short runner. There could also be scavenging based on the velocity and momentum of the gas. Even if there is though, and I've read this effect is not big, if the exhaust valve closes at the right time there would still be no reverse flow.
This ignores rpm though. All of the timing of flow and pressure implies a perfect time to close the exhaust valve, but all that gets shifted as rpm goes up. Also, to empty the cylinder maybe you want to delay closing to not limit flow, but then close the valve a little late and use anti-reversion to limit back flow.
I suppose with all the different rpm for engine and turbo, there will be some cases where there will be reversion, so plan for it.
I suppose also what you really want are high speed pressure measurements on either side of the exhaust valve. And, a V6 to more or less avoid the dead cycle.
Going back to basics, the two cylinder attached to one scroll would align cycles for instance like:
cylinder 2 _______ cylinder 3
intake ___________ power
compression ___ exhaust
power ___________ intake
exhaust _________ compression
The exhaust cycle from one cylinder is so separated from the other that each operates more or less as a single cylinder engine. Overlap and duration aren't enough to cause pressure wave interaction between the two cylinders, especially since I'm planning on short runners. It looks like it will be about 4" between the head and the twin scroll turbo inlet flange. The runners from 1 and 4 look like they will be around 7" long, with 2 and 3 somewhat shorter.
I'm trying to think about what the pressure at an exhaust valve will look like through the time the exhaust valve is open. Starting when the exhaust opens a burst of high pressure gas enters the runner. Velocity should get to sonic very quickly at the valve, so pressure in the runner should start rising quickly. Note, I'm kind of assuming pressure waves in such short runners get bounced and confused fast enough that they can be ignored, since by the time the valve starts to close, which is when you care most, the reflections have all died out.
Through a good part of the exhaust cycle the flow stays sonic, but the pressure in the cylinder is dropping. Sonic flow is directly proportional to the inlet pressure, so flow and the rate of pressure rise in the runner would start to fade, but at the same time the valve opening has been getting bigger. Towards the end of the exhaust cycle the pressure drop shouldn't be enough to maintain sonic flow and the valve is now closing, but there will still be some flow out.
All of this though ignores the rate the gas flows out of the runner and into the turbo. I'm guessing that flow out of the runner is proportional to the rpm of the turbo and the pressure and density of gas in the runner. The speed of the turbo will be effectively constant through a single exhaust cycle, but the gas pressure and density wont's be.
It does seem like the turbo won't be able to fully absorb all the gas in the runner. As the exhaust valve closes there will be some residual pressure that will continue to drop for the next engine cycle. This is the cycle were the cylinder we are looking at is pulling in the next charge, but there is no cylinder exhausting into the runner we're looking at. On the cycle following what you might call this exhaust dead cycle, the other cylinder attached to the scroll enters it's exhaust cycle and the pressure in the runner goes back up.
The question is, how low is the pressure in the cylinder and the exhaust runner when an exhaust valve closes. I think that the turbo will not be able to absorb all the exhaust gas by this point, because if it did the turbo would continue pulling gas out of the runner during the dead cycle and the pressure would go towards vacuum.
As the exhaust valve closes, will there be reversion? It doesn't seem like there would be based on pressure waves with such a short runner. There could also be scavenging based on the velocity and momentum of the gas. Even if there is though, and I've read this effect is not big, if the exhaust valve closes at the right time there would still be no reverse flow.
This ignores rpm though. All of the timing of flow and pressure implies a perfect time to close the exhaust valve, but all that gets shifted as rpm goes up. Also, to empty the cylinder maybe you want to delay closing to not limit flow, but then close the valve a little late and use anti-reversion to limit back flow.
I suppose with all the different rpm for engine and turbo, there will be some cases where there will be reversion, so plan for it.
I suppose also what you really want are high speed pressure measurements on either side of the exhaust valve. And, a V6 to more or less avoid the dead cycle.
Gassing Station | Engines & Drivetrain | Top of Page | What's New | My Stuff