S54 Header Design talk

[9kracing]

Oh this thread again?? lol

Horsepower is made before the collector. Period.

[JakeM346]

I think there may be some terminology confusion here. My understanding is that back pressure is NOT affected by scavenging and you’ve got these two effects crossed up. An example visualization of scavenging is like a sine wave of pressure centered around an average pressure. The average pressure would be back pressure. The sine wave and its amplitude would be scavenging. Pressure at the valve would be back pressure +/- pressure from scavenging. This looks like the reason for the miscommunication.

Also, to respond to the “guys, just…” - it does appear that many folks have done just this, one change controlled test at a time with headers and mufflers, and the data does not seem to support your hypothesis. Although generally what you’re saying bay be true, the specifics of this muffler for whatever reason or three appear to tell us otherwise. When the physics doesn’t match measured reality, it usually means your physics is incomplete or incorrect in my experience.

Scavenging creates pressure waves. These have an effect on the pressure differential in the exhaust system. Most often leading to a low pressure zone as soon as the pipes merge into the larger diameter pipe. However if the rear muffler is to restrictive the pressure in the complete exhaust system increases. Which will then (of course with a dampened effect due to the volume of the exhaust piping between the rear muffler and headers) decrease the effectiveness of scavenging.

If you wanna say any different...explain to me why in f1 using na engines the exhausts were built like this Formula One engines - F1technical.net.
Just a manifold going in a sometimes as big as possible diameter to increase scavenging effect while keeping the length of the exhaust as short as possible (to reduce friction-slowing the exhaust flow).

[3staxontheradio]

I think the root of this is a combination of people not agreeing on what counts as substantial power and also viewing exhaust systems through the lens of being a system of bottlenecks and restrictions and then further applying the idea of bottlenecks to think that flow restriction would actually cancel out or just cap the effect of pressure waves/scavenging. That's not how it works at all.
I'm going to lay this out in the simplest terms that most people here probably already understand:

As long as velocity is maintained pressure waves will be generated when there is a significant change in exhaust pressure (usually flow area). A negative pressure wave hitting an exhaust port will increase VE at whatever RPM its timed to. There is no other possibility here. Increasing the pressure differential on either side of the exhaust valve will move more air. All headers do this, even if you ran individual runners with no merge that dumped to atmosphere (See ITB trumpet tuning) There would be a negative pressure wave from where the exhaust gasses hit atmosphere. That wave makes it way back to the valve and no matter the length of the design it would at some RPM coincide with a valve opening and increase VE at that point. This is entirely independent of bulk flow through the system, again assuming the caveat of a minimum velocity threshold is maintained. Exhaust wave tuning is not a route to unlimited power and it's one component in increasing VE in a specific RPM range.

There is a number called a reynolds number. It's a "dimensionless quantity" and it describes the relationship between viscous and inertial forces in a fluid at any given moment. Higher velocity = higher reynolds number = higher inertial forces=stronger pressure waves. There is a minimum velocity threshold where scavenging effects fall off to the point of not being significant. "back pressure" is a term that comes from measuring a positive pressure in the exhaust compared to atmosphere. You can easily have high back pressure and high velocity. Just undersized your runners and tubing. Similarly if you want the worst of both worlds you can have oversized runners, oversized exhaust tubing, and then something like supertrapp muffler with a ton of baffle plates installed. In that case you have poor velocity from the oversized tubes terminating at a small restrictive orifice, velocity at the exit will be huge because small exit but that doesn't do anything for the rest of the system. The restrictive muffler is not really the cause of the poor velocity or lack of scavenging in the system. So again, scavenging effects and back pressure are not a direct relationship.

We don't need to go into a discussion about reversion but that's the only rational I can think of for claiming muffler change necessitates vanos/overlap tuning. Yes, if you switch from a particularly restrictive muffler that was creating high back pressure to a straight through design you could get away with more overlap because less reversion happening. If you are going the other direction you are making bad choices.

[JakeM346]

I think the root of this is a combination of people not agreeing on what counts as substantial power and also viewing exhaust systems through the lens of being a system of bottlenecks and restrictions and then further applying the idea of bottlenecks to think that flow restriction would actually cancel out or just cap the effect of pressure waves/scavenging. That's not how it works at all.
I'm going to lay this out in the simplest terms that most people here probably already understand:

As long as velocity is maintained pressure waves will be generated when there is a significant change in exhaust pressure (usually flow area). A negative pressure wave hitting an exhaust port will increase VE at whatever RPM its timed to. There is no other possibility here. Increasing the pressure differential on either side of the exhaust valve will move more air. All headers do this, even if you ran individual runners with no merge that dumped to atmosphere (See ITB trumpet tuning) There would be a negative pressure wave from where the exhaust gasses hit atmosphere. That wave makes it way back to the valve and no matter the length of the design it would at some RPM coincide with a valve opening and increase VE at that point. This is entirely independent of bulk flow through the system, again assuming the caveat of a minimum velocity threshold is maintained. Exhaust wave tuning is not a route to unlimited power and it's one component in increasing VE in a specific RPM range.

There is a number called a reynolds number. It's a "dimensionless quantity" and it describes the relationship between viscous and inertial forces in a fluid at any given moment. Higher velocity = higher reynolds number = higher inertial forces=stronger pressure waves. There is a minimum velocity threshold where scavenging effects fall off to the point of not being significant. "back pressure" is a term that comes from measuring a positive pressure in the exhaust compared to atmosphere. You can easily have high back pressure and high velocity. Just undersized your runners and tubing. Similarly if you want the worst of both worlds you can have oversized runners, oversized exhaust tubing, and then something like supertrapp muffler with a ton of baffle plates installed. In that case you have poor velocity from the oversized tubes terminating at a small restrictive orifice, velocity at the exit will be huge because small exit but that doesn't do anything for the rest of the system. The restrictive muffler is not really the cause of the poor velocity or lack of scavenging in the system. So again, scavenging effects and back pressure are not a direct relationship.

We don't need to go into a discussion about reversion but that's the only rational I can think of for claiming muffler change necessitates vanos/overlap tuning. Yes, if you switch from a particularly restrictive muffler that was creating high back pressure to a straight through design you could get away with more overlap because less reversion happening. If you are going the other direction you are making bad choices.

This is basically all i wanted to say summed up. The last part is especially what i was trying to get at
So basically i think the rear muffler is the most restrictive part in the system. People are allowed to have different opinions. I was just sharing mine

[Bry5on]

Back to header design and physics now. Let's for the sake of argument continue to ignore the muffler as a source of back pressure, regardless of religious e46 M3 affiliation.

The one concept I've personally always struggled to visualize is a tapered header runner (effectively infinite steps) and what that does to constructive/destructive interference of the resulting wave propagation. Somehow this is the 'ideal' shape but I must be dense as I can't 'see' the effect like simple wave propagation from a discontinuity. Since we're talking 3D printed headers, is it relevant to consider actually nailing this down, and is there a textbook that covers this well? I have 'Scientific Design of Exhaust and Intake Systems' here but that's the only literature I have handy.

[digger]

Oh this thread again?? lol

Horsepower is made before the header. Period.

fixed

[bigjae46]

Bimmerworld race and stock OEM exhaust manifold - 313 peak WHP on a canned tune. Can't tell me an exhaust does nothing.

[digger]

I came across this guy on YouTube, he seems to be quite down into the rabbit hole with exhaust designs:

https://ko-fi.com/maisteer/posts

its amazing how cheap SLM stainless is these days, a nice merge collector close enough to the same as a merge from Burns

[JakeM346]

Back to header design and physics now. Let's for the sake of argument continue to ignore the muffler as a source of back pressure, regardless of religious e46 M3 affiliation.

The one concept I've personally always struggled to visualize is a tapered header runner (effectively infinite steps) and what that does to constructive/destructive interference of the resulting wave propagation. Somehow this is the 'ideal' shape but I must be dense as I can't 'see' the effect like simple wave propagation from a discontinuity. Since we're talking 3D printed headers, is it relevant to consider actually nailing this down, and is there a textbook that covers this well? I have 'Scientific Design of Exhaust and Intake Systems' here but that's the only literature I have handy.

This book basically states that its benificial to enlage header diameter before joining individual primaries into larger combined piped. Therefore is estimate a tapere header would have a positive effect. But maybe not more significantly than just enlarging the header pipe a bit before the joining point

in this f1 header its also possible to see the increase in diameter. Maybe at the beginning of the header it even also tapered / this isa v10 engine and the exhaust ports of one chamber are only merged in the manifold. Thus also reducing the distance/time hot exhaust gasses spend in contact with the cylinder head

design a and b seems to lead to best results
(also is it possible to make this thread only accessible to forum members?)

[3staxontheradio]

My way of understanding stepped headers is that it is splitting the primary wave into a series of smaller waves. So one step placed halfway generates a small wave as the diameter steps up and pressure goes down. It also reduces the strength of the primary wave slightly. Your are just spreading that energy out over a wider area. More steps is more smoothly spread out. Sort of what Bryson is saying is that way of understanding what's happening stops being intuitive when the area tapers instead of being in discreet jumps since you can't pinpoint an origin of wave being generated.

I think the answer to what happens with a tapered runner probably has to do with impedance matching (efficiency of transmitting a wave back) and a natural frequency for the horn shape relative to its length. If you look at a cornet horn vs a trumpet - both terminate with a bell mouth but the body of the horn is different. A cornet has tapered shape and a trumpet is cylindrical. Cornets are described as having a warmer and softer tone (spread of frequencies) vs a trumpet which is more piercing. I 100 percent want to have a less woo woo understanding of what tapered runners would do it was one of the initial thoughts for why additive would be cool.

I see two main ways to use additive: use it for merges and transitions to get geometry that would be hard to make by hand. Maybe sparingly for bend geometries if they are hard to make with cut tubes. Build the rest of the header 90 percent traditionally.

Or alternatively do the whole thing as additive, probably in several pieces to keep part size down, and try out some things like tapered runners or whatever other weird geometry people can come up with

I also do think anti reversion is a place to explore. Especially with dual vanos and so much overlap available to us. I would want to steady state flow test any kind of anti reversion geometry directly compared to smooth tube to see if it costs anything on bulk flow. FDM is good for that kind of thing. We could potentially see a wider torque curve by being able to use more overlap though a bigger range of RPM. I think the flanges to the head would be a good place to start and also somewhere along the collector.

[Bry5on]

One of the things I’ve intuitively struggled with is where to start the taper. Should it start at the first-nth harmonic? As early as possible? Where should it end and how do you size the area ratios? I think having a better understanding of the fundamentals would probably assist this, or even vice versa. Maybe I need to do more reading on two stroke headers, as those folks seem to be the experts on the topic since testing is so easy. Maybe we can all pitch in and buy George Hill and engine dyno and convince him a bunch of testing is a good idea? 👀

[3staxontheradio]

My plan was to take the diameters calculated for a stepped header and use them as points on a taper. Basically just turn a stepped header into a smooth taper rather then discreet steps. I would want to either have a lot more confidence in a final design or test out a traditional header on a dyno before trying to do a fully additive part. I would guess an additive long tube header is going to be in the $2000 ballpark.

I still need to do some chassis and engine bay scans + pull my header to get good scanning clearance. Once I do that I can actually do a first design pass. Hopefully I have time to do that this coming weekend.

[George Hill]

Maybe we can all pitch in and buy George Hill and engine dyno and convince him a bunch of testing is a good idea? 👀

I so wish I had a dyno, we would wear that thing out testing all kinds of hair brained ideas.

[Bry5on]

My plan was to take the diameters calculated for a stepped header and use them as points on a taper. Basically just turn a stepped header into a smooth taper rather then discreet steps. I would want to either have a lot more confidence in a final design or test out a traditional header on a dyno before trying to do a fully additive part. I would guess an additive long tube header is going to be in the $2000 ballpark.

I still need to do some chassis and engine bay scans + pull my header to get good scanning clearance. Once I do that I can actually do a first design pass. Hopefully I have time to do that this coming weekend.

So would you start the taper at the head and expand out to the step diameter, or start the taper at the first step, and expand out from there? Intuitively it seems that a straight section that starts tapering at the first step makes sense to me, but what do I know. I'm assuming you're also targeting equal lengths all the way back for that smooth exhaust note? That and evening out cylinder scavenging engine speeds (from the SSV1s that I dislike) are my two primary goals. Well, that and learning by doing.

If you're on this quick of a timeline, you'll beat me for sure, but we may be able to find a way to share some of the iteration cost risk to ease the pain.

I so wish I had a dyno, we would wear that thing out testing all kinds of hair brained ideas.

I'd be making even more visits to Austin

[digger]

Some quick sims i did on a rough s54 model. For simplicity open primaries (zoomies) 450mm long 35, 38, 41mm primary pipe

- 35mm uniform
- 38mm uniform
- 41mm uniform
- 35-41mm taper
- 35-38-41mm step

power


exhaust port pressure 8000