Tri-Y (4-2-1) stepped vs. UEL 4-1
 

So I was staring at some of the dynos for some of the many header offerings and it got me thinking. It's widely accepted or assumed that the 4-2-1 tri-y style headers have a tendency to give better low and mid-range power while the 4-1 equal length headers tend to yield less in the low-mid range and more in the top end, relatively speaking.

Now, looking at some of the results from the many UEL 4-1 offerings we have for the FA20 it seems that they also have the same effect of the traditional tri-y header compared to the el 4-1 offerings. Another upside to the UEL 4-1 header is that in terms of manufacturing and design these are the easiest to work with and the try-y (4-2-1) being the most difficult to manufacture and design.

What are your thoughts abut this? Which is the best choice for our little boxer engine and why?

Performance wise, the EL gives the best results, and has the typical 4-banger exhaust note. The UEL is almost specifically related to our Boxer engines and gives you that subaru exhaust note, at a slight performance "nerf" it's almost negligible if you stay N/A without a full build, so the choice is really up to you.

My thought is that the 4-2-1EL header was designed and optimized to this engine and seems to show the best gains in both the torque dip and up top.

The other headers are still useful in that they remove the manifold cat which allows better tuning than with the oem manifold cat.

I wont do either without a good tune and an overpipe.

I wouldn't be surprised if people figure out that the right tune, overpipe and gutted oem manifold are about a good as some headers.:popcorn:

IMHO with the exhaust scavenging on this engine any catless header is a smart mod for all especially superchargers . The last thing you want is an over heated broken down cat shedding into the combustion chamber.

How would pieces of a cat get into the combustion chambers with a supercharger?

Sent from my GT-S7560M using Tapatalk 4

With a supercharger or any tune that is a lot different than stock the possibility of the cat getting too hot and breaking down exists. What I've seen on other cars is little pieces of the cat get trapped in the ex manifold and end up sucked into an exhaust valve, this engine like others is designed to pull (scavange) some exhaust for emissions.

Its just smart insurance to remove manifold cats when making big changes to factory tunes. That's any modern car.

UEL fits this car, but its all an opinion. Its nice to have choices because everyone's build is different.

I'm not up to speed with aftermarket Subaru headers, do uel's typically have theflex-joints like Borla or none like rev-works ?

If anyone wants to see how bad this can be, look up what happened with Sentra SE-R Spec V's when their stock cat broke down on A LOT of cars that were 100% stock. Nissan replaced a lot of engines before removing the cat that was causing havoc.

Those cats are still there, in the same place and still having some issues on the qr25. It was not just the Sentra's but also the Altima's that had that issue. It uses cam phasing to create an egr effect, same theory behind the fa20 cam phasing. As the cat breaks down over time it cases the engine to ingest the material from the cat and causes cylinder damage. I think Nissan has upgraded the cat quality some sense there initial release.

JDL has a braided flex bellow while Borla just has flex joints. Flex joints rattled on the JDL so they swapped them to the braided. Rev Works chooses not to use them.

Yes its common and can be exacerbated by aftermarket tunes, especially a sc tune where there is a lot of unburt fuel in the exhaust. any catless header should be a prerequisite to a supercharger IMHO.

With my "regalonia" :lol: l want a header before I go messing with an NA tune. Saving for the HKS.

Actually almost all OEM headers are UEL, but not to a degree that you would associate it with a "boxer rumble".. Really never understood why Subaru went with such an unequal design. For boosted applications, sure you can make up for power loss, but on N/A (the majority of cars they sell) they never revised it. UEL sure is a preference on sound, but is certainly less efficient.

As far as 4-2-1, I would much rather have this design. Only problem is, not very many of those making headers for us understand the concept of "tuning" the runners. Nameless is doing it with theirs, but I can't think of anyone using a stepped design currently, possibly the JDL. This tuning of the runners and getting the proper length are where power is made, not just by slapping on larger diameter pipes for increased flow capacity. 4-1 requires less fine tuning which I would assume is the reason for most companies to go this route. There is only 1 harmonic frequency, with 4-2-1 there are 2.

Also, like Nameless is finding out, to get the optimal length for their runners, they need more room than our bay currently allows. This is why they are taking so long. Well, that and poor time management lol.

Id like to know where you got the info that "almost all OEM headers are UEL" It's design is inefficient both performance and economy-wise. Why would a manufacturer decide to design it as such?

in theory 4-1 can make more power at a specific rpm but have a narrower power band. but w dual vvt and tuning its not so obvious thats the case anymore. 4-2-1 have 2 peaks so in theory they give up peak power for a wider powerband. but look at hks header... it makes more peak power than a lot of 4-1 headers. the r&d is really more important than configuration

regarding sound 4-2-1 should be quieter than 4-1 can anyone confirm? thats been my experience w other 4-cyl cars

Actually, their initial prototype was very impressive with the TQ peaking where the dip used to be. Even with a cat mated to it, it was much more impressive than the HKS offering. http://www.ft86club.com/forums/showp...59&postcount=3

The first dyno in that link is their 4th? revision with the space (fan shroud?) issue.

Is that right? Still very impressive especially considering this was achieved without a tune, since they tested this before there were any tuning solutions for the ECU available. I can only imagine what it would have looked like with a proper tune on E85 fuel.

I've been considering fabbing my own header since I do have some 1 5/8" and 1 3/4" SS tubing in the form of aftermarket honda headers laying around. I saved them instead of selling them just for an occasion like this.

I've never worked on a boxer engine before this car save for some work done on some WRX bolt-on stuff. Hoping to learn what works on these things before I even consider busting out my chop saw and welder.

Tuned as well.

The ECU? With what?

Ecutek, I think? English Racing was working on sorting out the small left-over dip, plus it seemed to be changing a bit on them, maybe the IAM. It's somewhere in the Nameless thread. I'll link if I can find it.

Edit @solidONE

Starting ~post #1211

http://www.ft86club.com/forums/showp...postcount=1211

This is the guy tuning it.


http://www.ft86club.com/forums/showt...on+dip&page=31

OEMs cut corners. It's far cheaper to produce a "good enough" manifold than going through extreme r&d.

Do not assume because they are a reputable company that they would put the best possible part to accomplish a task that a mediocre part can do to their satisfaction.

Everything being the same (runners, length, diameter, etc.) an Equal length header will always make more power than an un-equal.

The UEL is just something of a gimmick for sound and uniqueness on these platforms.

Thank you.

What are your thoughts on 4-1 vs 4-2-1?

Every platform is different, so I can't say what is ideal for the Fa20 right now. Right now, it looks like longer runners are doing better than the shorter runners, whether it be 4-1 vs 4-2-1. The issue is space and packaging.

I wish I had access to the facilities and resources to do a 100% full test on every conceivable design to make the best header, That would be awesome :brokenheart:

We are looking into some possibilities on our end for something, nothing solid yet though!

As of now we are recommending the HKS headers as a really nice choice.

Aren't the HKS headers just a blatent rip-off of the OE ones but minus the cat?

Sure it pairs the same cylinders and is 4-2-1, but that's about it.

The HKS is closer to true equal length, has much better merges for the pairs, smooths out the 90 degree turns at the start of the secondaries, has much longer secondaries (instead of the cat). I'm sure the primaries and secondaries are larger as well.

HKS:
http://www.garagefrs.com/wp-content/...8-1024x675.jpg

OE:
http://www.garagefrs.com/wp-content/...9-1024x675.jpg

Honestly, I think if you improved the lower merge, and replaced the cat with longer secondaries, the OE wouldn't be all that bad.