I was recently asked this question and completely botched the answer- making it seem like they do nothing. So now is a good time to share what we’ve learned so far.
We have both a short answer and a long answer to this question. There are more factors and details involved but this aspect tends to be the largest lever when it comes to power per psi differences, and the circle of interactions that often gets overlooked. From our understanding and testing, this is what we’ve learned:
Short answer:
Engines tend to lose volumetric efficiency (VE) as exhaust pressure goes up with no other changes. We’ve found them to be pretty sensitive- more than we initially thought. What tends to happen is the relationship between boost and drive/backpressure gets less favorable (higher drive pressure than boost pressure) and you get leftover/remaining exhaust taking up space in the following intake stroke, and also less actual intake air charge in the cylinder due to that. That directly affects volumetric efficiency considering VE is how much of the cylinder is filled with fresh intake air.
It has been helpful to think of lower VE as the engine acting smaller. So you could say as drive pressure goes up and up, the engine acts smaller and smaller.
What an effective turbo manifold can do is take that drive pressure needed to generate boost, which is an average that actually consists of many peaks and valleys from the exhaust pulses, and protect the engine from seeing those higher pressure peaks at the wrong time, and see the lower pressure valleys at the right time. That time being near the end of the exhaust stroke- which is also the valve overlap phase where the engine is most sensitive to drive pressure and determines if remaining exhaust stays or leaves. So when the exhaust stroke is near the end, the combustion chamber can be better cleared of exhaust gas and make room for the incoming intake air, which will increase VE/better cylinder filling. The setup can have a more favorable ratio than the average indicates, or even when the average looks good but things aren’t moving in the right direction when it matters most.
It now sounds less absurd to say that the engine will act larger, even at near the same average drive pressure. All of the more well-known characteristics of a larger engine apply at those points where VE is improved, like making more power at the same psi, better turbo response due to more mass flow through the turbo/turbine at that rpm/load point, able to max the turbo out at a lower boost level, and allows you to get away with more aggressive camshafts. So how much of a VE change? During one back-to-back stock replacement dyno test we saw ~10% VE/fueling/airflow change when under boost with a manifold change alone, which was as high as ~60whp more in some spots at the same boost (mid 600whp setup). Some other comparisons seeing ~25whp or 5% change, some seeing 15-20% VE change around spool up; each setup will depend on where it’s starting from and how sensitive it is. Make power more easily, but keep in mind that the peak airflow potential of the turbo doesn’t change; once it’s compressor limited, that’s the limit for airflow- but in the case of an engine that just increased VE from an exhaust manifold change alone, it’s likely due to keeping hot exhaust out of the intake stroke and allowing more intake air in. Turn overlap into a good thing and open up the sweet spot.
The most exciting part to us is how the mechanisms behind things like runner size, length, and open vs. divided seem to fall into place and become less hand-wavy and more intuitive. With this jumping off point we hope to give a better understanding of how to make your setup work together as a system rather than just looking at individual components. We will be looking at the implications on various design aspects with more articles now that this initial piece is out there.
Long Answer:
The thought process that led us there, addressing gaps or allowing you to point out holes in our reasoning:
To start, it would be most obvious to say a turbo manifold allows you to open up new turbo configurations, or gain better placement, etc. A stock replacement manifold does none of those things but offers an opportunity to do a direct back-to-back test with no other changes, and is also where we gained the most insight into the interesting bits and often unanswered questions. In short, an effective turbo manifold makes the engine act larger.
We can explain.
Quick primer- A turbo uses exhaust heat and pressure to create boost. The engine is actually pretty sensitive to that required exhaust pressure, aka backpressure, or more specifically called turbine drive pressure when driving the turbo to make boost. An engine will tend to act smaller as the turbine drive pressure increases, and make less power per psi of boost just like a smaller engine would. Ever wonder why a larger/more efficient turbo can make more power on the same engine even when all other factors are the same- like rpm, boost level, and temperature? It’s not just cooler, denser air from the more efficient compressor- an efficient intercooler can make the air is just as cool and dense in the intake manifold for both. It’s the turbo being able to run a lower drive pressure at the same boost pressure allowing the engine to also work more efficiently. With setups tuned using speed density this can be tracked pretty well with changes in volumetric efficiency (VE) which is how efficiently the cylinders are being filled (by mass) with fresh intake air. A turbo change alone requiring a change in the VE map isn’t uncommon whatsoever. VE tables that are based off 100% VE (corrected for boost and temp, etc.) can be thought of as how large the engine is acting vs. its actual size, and we find it very helpful to think of the VE map like an engine displacement map to better wrap our heads around it. And just like a larger engine tends to make more power with the same boost, it generates more mass flow through the turbine to drive it which can increase turbo/boost response too. The displacement analogy holds really well since it really can ingest more air per revolution. On airflow metering setups (MAF, MAS, etc.), you will see it as more airflow at the same boost and rpm points. All related, just calculated using different variables. It should perhaps be said that you can’t just change the VE values, the VE change must come from a change in fueling requirements. If the engine leans out from more airflow at that VE load cell, a wideband sensor will indicate that change and you’ll dial in the VE map appropriately. Not being dialed in in the first place, or changes in the wideband, etc., can lead you astray.
So how does exhaust pressure ultimately affect how much of the cylinder is filled with fresh intake air/boost? Isn’t VE how much air is trapped in the cylinder and mostly due to camshafts and the intake manifold side of things? That question plagued us for a while since we kept seeing VE/fueling changes from exhaust manifold swaps alone, and in certain tests over 10% increase in VE and airflow across a surprisingly large rpm range. This was unexpected and hard to explain, but importantly backed up with a similar power/torque increases where boost and rpm lined up between the two, making it harder to dismiss as skewed data.
The simplified answer comes from looking at what happens during the short time that the intake valves and exhaust valves are open at the same time. Valve overlap. The amount of drive pressure and the amount of boost pressure when both intake and exhaust valves are open dictates which direction things move. More boost pressure than drive pressure can begin to fill the cylinder with fresh air as exhaust is expelled during overlap, keeping things moving in the intended direction and making more power by flowing more air. More exhaust/drive pressure than boost pressure during overlap can push exhaust back up into the intake if bad enough, definitely hurting power and airflow.
BUT.
Boost pressure mostly stayed the same (the whole point of the comparison test) and average drive pressure actually went up a bit with our manifold (more on that later) which made it even more confusing. So it couldn’t be explained by a more favorable average boost to drive pressure ratio. Saying “average” on purpose because exhaust between the port and turbo isn’t constant flow. Exhaust pulses, a squiggly pressure trace that is only referenced as an average drive pressure value. And the exhaust manifold plays a strong influence on how high and low those squiggles go, and when the engine sees them. The turbine drive pressure mostly shifts that whole exhaust pressure trace up along with it.
So, scavenging?
We’ve known about scavenging increasing airflow, but always heard that it tends to happen over a narrow rpm range- not the wide rpm range we’ve seen. And, to have increases of 10% more airflow, it seemed like the small window of opportunity when both intake and exhaust valves are open (overlap phase) would just pull the intake charge out the exhaust before it got 10% more intake air into the cylinder, explaining the fueling change but not the associated power increase. All didn’t seem to sit right in our heads.
Residual exhaust gas?
Exhaust gas that doesn’t make it out of the cylinder, takes up space that could have otherwise been fresh intake air, directly and negatively impacting VE. Taking more boost to pack the same amount of intake air in alongside the leftover exhaust gas, adding boost but not making more power. Just adding spice. We had heard of residual exhaust gas before, typed it even, but never really considered what it does to VE until trying to figure out where an exhaust only VE change was coming from. That also would mean a turbo manifold would mainly only improve setups based on how much residual exhaust gas there was in the first place, and seemingly there was a lot.
Initially, we probably put too much emphasis on residual exhaust gas for the differences in VE after first considering it. Assuming there was 10% of the total cylinder mass being removed in exhaust gas- not 10% less exhaust gas but meaning at least 10% of what was in the cylinder WAS exhaust gas seems a little much, too. Still plausible, since on a 10:1 compression ratio engine, roughly 10% of the cylinder volume is chamber volume that the piston doesn’t push out, and if there is no charge motion to move it out, it would just stay there and take up space during the following intake stroke (and probably even more if it’s over 1:1 drive pressure to boost pressure ratio but things get complicated quickly since gases are squishy).
It’s probably not all residual exhaust gas though and not all scavenging, just like it’s unreasonable to assume that all exhaust gas makes it out or that none is left. It’s not a perfect process. But residual exhaust gas and scavenging seem like two sides of the same coin. In order to give the intake air charge somewhere to fill, you need to create a void by getting rid of the exhaust gas. So it’s a spectrum, they go hand in hand. Every bit seems to matter and will show up in VE, helping keep things moving, allow you to generate more airflow sooner, make more power with less boost, take advantage of overlap, and give more room to play with for cam timing and overall tune.
More to come looking at runner diameter, runner length, cylinder to cylinder interference, variances, open vs. divided, etc. and how they affect the entire setup as a system- and your experience from the driver’s seat.