Theoretical header selection mathematics
 

I’m working through some decisions for header selection. It looks like most headers have a primary tube diameter of 42.5mm or 45mm (interestingly the OEM exhaust manifold primary tubes are similar: 42.7 for ’13-’16 and 45mm for ‘17+). After some online reading, it seems that primary tube diameter has a strong influence on what RPM is targeted for torque improvement. The formula is: Peak torque rpm = Primary pipe area x 88,200 / displacement of one cylinder. I assumed a 16 ga wall thickness of 1/16” (1.5875mm). Using that formula, the RPMs where torque is improved appears to be 2300 for 42.5 and 2466 for 45mm. The torque dip starts at around 3,000. Perhaps I’m missing something here, but the primary sizing seems ill-matched to our application.

Imho header runners length does more to optimization/best scavenging to certain rpm ranges then inner diameter of runners. Diameter .. i'd probably care (a very little) only in high power forced induction builds.

An increase in diameter plays an important role. Check the difference in torque of the OEM exhaust manifold before and after the revision:

Blue color: before the revision
Red color: after the revision

http://i65.tinypic.com/2gvqvbp.png

The difference in power is 7 PS (JDM cars had before the revision 200 PS and after the revision 207 PS). I never understood why the NA cars got 5 PS.

The increase in torque has been measured also just by switching the exhaust manifold and get a tune. You don't need the red intake manifold or the other (internal) engine revisions.

Question 1, what is the source of that formula? It seems overly simple to provide an rpm at which the headers are most efficient. A whole lot of assumptions went into that 88200 constant which may not be appropriate for the dimensions, etc of the twins.

I'm going to hazard a guess as to what that constant includes and say that RPM is, assuming all prior assumptions are correct, a primary frequency of the header. Meaning that the header scavenging is most efficient at 2300RPM and every 2300 RPM after that.

Also keep in mind that the FA20 has vvt on the intake and exhaust cams, so the tune plays a factor as well.

Edit: On second thought, my guess was incorrect. I should actually stop trying to think this late...

first, 207 PS = about 205 HP . japan cars didnt get more power it s just a different unit
second you can t use that bad made/low resolution dyno graph to estabilish that those gains are because of the new header. It s header + manifold + tune, and i bet you could get even bigger numbers withouth the mods of the revised models, but just a well done tune

you desperately want to justify the enormous amount of money you spent for that heavy and useless header you bought from japan, but the reality is the gain you had are just because of the cosworth tune and not from the header that still has the huge cat that blocks the gas flow and stops the timings at the same level as the older one

I might be wrong, but IIRC gains in MY2017 came more from better stock tune. As in there was less on plate to gain with aftermarket tune on later generation, vs tuning ecu on older cars. I wouldn't freely single out slight header changes as THE reason of gains and would be extra careful when there were several things changed at same time, eg. less resistance airfilter, slightly different intake manifold, slightly different exhaust header, different ecu tune. If anything, imho most MT kouki gains come from tune & airfilter (as AT kouki-s with old airfilterbox are still 200hp).
But if we are look only when one component changes, eg. between different aftermarket headers, then main differences in curve looked not from ID of runners (most aftermarket ones had larger ID then stock), but rather from runner lengths (eg. short runner headers, UEL headers, long tube headers like ACE's or Nameless, or different overpipe part lengths (150/250/350) for misc. ACE's headers)

Setting aside the dubious validity of the formula I used, there had to be a functional reason Subaru went from a primary tube diameter of 42.7 to 45mm. I agree that it's far too simplistic to attribute that single change to the power increase for the 2017's. But also, I am disinclined to merit the bulk of power increase to calibration changes. They changed the airbox, intake manifold, valves, exhaust, etc...To be honest, the number of things that they changed/tweaked is out of proportion to the actual power increase. Regardless, it seems that the increase in the primary tubing had a reason, and I'd like to learn about it before I buy a header (not that there are that many options. JDL is the only one with the larger diameter, 44.5 mm, the rest appear to be 42.7mm). At this point, I suppose it's just a case of academic curiosity.

If you look after "which aftermarket header provides most gains", then i'd look for diminishing ordered list by features by their importance like: 1) if header is catless or catted, 2) how good ecu tune is optimized for it, 3) header generic construction (UEL/EL) & runners length, 4) ID / runner merge angle / radius of bends / shape of collector / quality of welds. I moved ID to last as playing (relatively) little role in performance. First three bits are much more important.

Correct header design for NA is the most "not thought through" modification out there, one size does not fit all applications.

ACE headers are on the ball with there designs, albeit a quite expensive venture.

I am currenty trying to find a header to suit my daily driver duties with non competitive motorkhana thrown in.

I'm thining long tube EL 4:1, maybe, I'm still trying to figure it out and find a seller that doesn't want my kidney.

I also don't want 304, it needs to be either 321 or mild steel.

I've never had luck eith 304 headers, they always ended up cracking.

Same for the '12-'16 EU cars. It is not a different unit. It is same unit. They were spec'd at 200 PS and not HP. Check below or in the attached file at page 4.

http://i67.tinypic.com/2coo3n8.jpg

The intake manifold helps to not drop the power much in very high rpm's (after 6.5k). That's all. There was a post of celek about this. The tune had to stay within factory standards ...

I am not trying to justify anything. You know that the exhaust manifold with the bigger diameter didn't come in Europe and it wasn't about any change in the cat converter. If there were zero gains and zero change in the exhaust flow why did it not get approval?


And if you want some history of the revised exhaust manifold, a catless version was tested in the Nurburging 24h races back in 2014. The first car that got it afterwards was the GRMN 86. The difference is that it has perfectly equal length (the first one was close, but not exactly equal) and the diameter is increased further. The catless version gives for sure better results, but the cat converter does not cancel the better design. The cost was high to justify a new exhaust manifold for zero gains. I've seen the parts and apart from the manifold they had to create also an entirely new heat shield.

Usually 304 headers fail because of bad welds, improper tunes or high boost overheating the metal, to much twist/play under acceleration, or so old they loose the noncorrosive properties.

Some other members have had headers crack and break because they use stock engine mounts on boosted applications.

The power increase was bigger, but SOA wanted to keep the car "low" profile and not to get much in the WRX territory. They even said to the journalists that the original car was 200 HP, which was never true. The original car was 197 HP = 200 PS. At least for the latter there is evidence and I attached above the original catalog specs. Additionally, Cosworth that tuned the original engine and works with many other OEM manufacturers said multiple times that even the original 197 HP catalog number was too optimistic.

EU kouki specifics are mostly caused by stricter emission standards. Toyobaru for EUDM decided that being Euro 6 compliant was more important then +5hp. IIRC they couldn't get new intake/exhaust/tune bits and still be Euro 6 compliant. As depending on compliancy may depend taxes/dotations/certifications on certain market according to governing legislation/competitiveness among other carmaker offerings in said market countries, both for manufacturer and for car buyer/owner, toyobaru marketing departments evalued/weighted how much said perf. gain vs ecostandard may hurt or promote sales and did decision based on that. I perfectly understand their reasoning. There is reason why most well known traditionally NA ICE manufacturers switched lately to smaller displacement but forced induction engines. It's VERY hard engineering challenge to stay NA with high performance AND be compliant with latest eco emission standards. In that sense FA20 is VERY efficient and well made engine getting close to most of what's possible and "last of breed", that shouldn't be compared with eg. NA engines of past that weren't strangled with as strict emission requirements, and that probably in next iteration cars of this class also will be replaced with smaller turboed engines like for other carmakers.

Don't over think this. There are plenty of reviews here to base your decision on. It's a cheap sports car, not the space shuttle.