S54 Header Design talk

[3staxontheradio]

OEMs are 100 percent of the time constrained by mass manufacturing processes, cost, and serving multiple markets and a broad audience. For example: That PDF has a bunch of marketing copy about their tube bending but in actuality cnc bent tubing like that comes with a lot of limitations on what can and can't be bent. The OEMs are very good at coming up with solutions given the constraints they have but I don't need to deal with the same issues they do and that includes how the rest of the exhaust it built.

[3staxontheradio]

Also on trusting the current pipemax output - I have two things I want to understand better.

1) I'm getting peak power numbers that are optimistic compared to dyno results with a similar torque number. I was starting off trying to match the BMW claimed power and torque for a euro car and if adjust the VE to match torque I get too much peak horsepower. Maybe that's down to street car stuff like an OEM style mufflers nuking the top end but I at least want to make sure it's not calculating way more exhaust gas volume at the top end then reality because that will throw off the primary diameters. There are like 30 different engine / dyno sim options to use and that also impacts the result. Also I think trying to apply chassis dyno results to the sim is a can of worms? I haven't seen other stock or close to stock engines on an engine dyno but if that info is out there it would be helpful. The engine and power calculator is separate from the header calculator but the header calculator uses the info from the engine calculations to determine how much exhaust gas is flowing out of the engine so it matters. Maybe I'm just misunderstanding the workflow.

2) I think it has an accurate idea of when valve events are happening but I'm not sure how to interpret the calculated EGT since EGTs are transient. Is it peak EGT on a WOT pull?. If the EGT it's using is way off the pressure waves will be timed wrong. I can also input a manual EGT if we have a number we think is correct to use.

Pipemax seems really under documented and like users are supposed to be able to just know what it's doing/how to configure it. Digger was super helpful with getting the inputs dialed in better but a lot of it isn't obvious.

The last smaller thing Is getting the cam overlap based settings dialed in a little closer. Though when I tweak a range of settings there the suggested lengths don't change that radically. More overlap seems to just create a bigger window for timing pulses / increase the size range of the suggest lengths

[Obioban]

Yeah I’m mainly stating that you’re unlikely to gain something because you’ll be limited by the muffler

… what? Removing the muffler on the e46 m3 doesn’t gain any power over stock— it is not a restriction

[Bry5on]

Also on trusting the current pipemax output - I have two things I want to understand better.

1) I'm getting peak power numbers that are optimistic compared to dyno results with a similar torque number. I was starting off trying to match the BMW claimed power and torque for a euro car and if adjust the VE to match torque I get too much peak horsepower. Maybe that's down to street car stuff like an OEM style mufflers nuking the top end but I at least want to make sure it's not calculating way more exhaust gas volume at the top end then reality because that will throw off the primary diameters. There are like 30 different engine / dyno sim options to use and that also impacts the result. Also I think trying to apply chassis dyno results to the sim is a can of worms? I haven't seen other stock or close to stock engines on an engine dyno but if that info is out there it would be helpful. The engine and power calculator is separate from the header calculator but the header calculator uses the info from the engine calculations to determine how much exhaust gas is flowing out of the engine so it matters. Maybe I'm just misunderstanding the workflow.

2) I think it has an accurate idea of when valve events are happening but I'm not sure how to interpret the calculated EGT since EGTs are transient. Is it peak EGT on a WOT pull?. If the EGT it's using is way off the pressure waves will be timed wrong. I can also input a manual EGT if we have a number we think is correct to use.

Pipemax seems really under documented and like users are supposed to be able to just know what it's doing/how to configure it. Digger was super helpful with getting the inputs dialed in better but a lot of it isn't obvious.

The last smaller thing Is getting the cam overlap based settings dialed in a little closer. Though when I tweak a range of settings there the suggested lengths don't change that radically. More overlap seems to just create a bigger window for timing pulses / increase the size range of the suggest lengths

In my datalogs, EGT varies pretty wildly from under 300C to above 700C. An estimate around 600-650 for heavy loads seems about right. My datalogs are too big to host online,I’ll look for a good one that has plenty of WOT in it and send via email. heinzboehmer has a bunch too and I’m pretty sure his EGT sensor is in better shape than mine as I occasionally get codes and blips in the logs. So take anything from me with the appropriate grain of salt.

[George Hill]

I have dyno’d a bone stock M3 and it lost power when we pulled the muffler and just ran the sec 2 open.

[heinzboehmer]

In my datalogs, EGT varies pretty wildly from under 300C to above 700C. An estimate around 600-650 for heavy loads seems about right. My datalogs are too big to host online,I’ll look for a good one that has plenty of WOT in it and send via email. heinzboehmer has a bunch too and I’m pretty sure his EGT sensor is in better shape than mine as I occasionally get codes and blips in the logs. So take anything from me with the appropriate grain of salt.

Yep, I have a bunch of street/track datalogs. 3staxontheradio pm me your email if you're interested in them.

[JakeM346]

I have dyno’d a bone stock M3 and it lost power when we pulled the muffler and just ran the sec 2 open.

yes that makes 100% sense. Because as soon as you pull the muffler you loose most of the back pressure.
If you were however to adjust the tune you’d be able to get substantially more power. And that’s why I’m saying….the headers are the smallest problem here

[Tbonem3]

to play the devils advocate, there are diminishing returns with the exhaust primary on a nearly stock engine once you've removed the catted manifold. This should be looked at from the perspective that nice headers look cool and modern manufacturing techniques are fun to play with.
If you go into this with the thought that there are significant power gains to be had over existing products you'll be disappointed

The voice of reason

[3staxontheradio]

Either the output pipemax is coming up with right now is way off or superspint v1s are tuned on a different harmonic. Supersprint says they did extensive development work but I generally don't give those kinds of claims a lot of weight. I would love to see some true a/b dynos with the supersprint v1s and other catless headers. Also euro headers clearly have very short runners and are likely on the 5th+ harmonic.

I guess "significant power" is subjective. I think 5 whp over a big area of the curve is pretty significant. Maybe this is just a semantic discussion because yeah, probably lifting the working area of power curve by 10whp vs a supersprint header is unrealistic. I don't really understand the skepticism around using a stronger pulse/harmonic and then trying to spread that pulse back out with some steps.

Also as I mentioned before, the supersprint headers are damn expensive and I'm not going to spend that much on a mass manufactured product when I can either build a traditionally fabbed header with plausibly better design + a not shit merge for a fraction of that or have fun messing around additive and end up in a similar price point as a supersprint v1 but with something that was way more interesting to make and work on.

There is such a wide variety in runner lengths and diameters between headers and they can't all be good designs. I don't think the s54 is somehow the only performance engine that is agnostic to scavenging and exhaust tuning.

[digger]

Can you elaborate on the output issue? In some respects the older simpler version of pipemax was better

[digger]

I don't think the s54 is somehow the only performance engine that is agnostic to scavenging and exhaust tuning.

it isn't, but there is only so much you can do compared to something that already works well and helping one area usually hurts somewhere else. It is probably going to require a lot of testing which gets expensive to make several versions

[JakeM346]

… what? Removing the muffler on the e46 m3 doesn’t gain any power over stock— it is not a restriction

Well unless you adjust the tune ofc it doesn’t gain power. And also I’m pretty certain that every muffler is a restrictor. Because that’s how they work….if you wanna make it quieter you have to make the exhaust flow bounce back and around. For power however it’s best if the exhaust flow after the headers is as undisturbed as possible…
theres a reason race cars don’t have any mufflers…besides weight saving (otherwise all race series would have made them mandatory)

[3staxontheradio]

The only thing with the output that gives me pause is that ratio of horsepower it's calculating for 7900 rpm to torque at roughly 5k is skewed. The top end is too strong for any given torque/VE setting. I'm not sure that really.matters though? Is it sizing everything.to make peak torque at the target RPM? Or is it sizing the runner diameter thinking the engine is moving that much more air at the top end? Also I'm happy to do some interating, especially with traditional fab versions. I also got a hookup to run some final parts through ansys but I don't want to cache that favor in until I have an actual candidate for a final part.

[digger]

Well unless you adjust the tune ofc it doesn’t gain power. And also I’m pretty certain that every muffler is a restrictor. Because that’s how they work….if you wanna make it quieter you have to make the exhaust flow bounce back and around. For power however it’s best if the exhaust flow after the headers is as undisturbed as possible…
theres a reason race cars don’t have any mufflers…besides weight saving (otherwise all race series would have made them mandatory)

mufflers can gain or lose power, the muffler is not just a restriction device it also changes the effective length and will cause reflected waves to to the change in cross section

[digger]

The only thing with the output that gives me pause is that ratio of horsepower it's calculating for 7900 rpm to torque at roughly 5k is skewed. The top end is too strong for any given torque/VE setting. I'm not sure that really.matters though? Is it sizing everything.to make peak torque at the target RPM? Or is it sizing the runner diameter thinking the engine is moving that much more air at the top end? Also I'm happy to do some interating, especially with traditional fab versions. I also got a hookup to run some final parts through ansys but I don't want to cache that favor in until I have an actual candidate for a final part.

This points to there being some restriction that pipemax hasn't accounted for, since pipemax is an empirical program (not a simulation) developed mostly on pushrod v8 with open collectors i wouldn't be too concerned as the s54 engine has full induction and airbox cleaner and full mufflers.

i would adjust for the hp to be correct as that is what pipemax is mostly using and the pipe diameter is most strongly influenced by peak hp compared to anything else

the primary length are based on peak hp rpm and the assumed associated power band for that.