I took a little trip over to Wisconsin to pick this up from Section 8 on Nasioc. It is a custom long tube header design built by TWE that was used on some rally cars and just a few float around the US at the consumer level. He ran it for about 12k miles before he sold his old RS. It's a 4 into 1 design with cross tubes at half length to aid higher rpms as well. Overall runner length is around 60". The primaries were said to be 1.75", but it seems to be more like 1.625" looking at it. The collector exits into a 2.5" pipe size. My bro has a Car Sound 59956 high flow metal substrate spun cat and a 2.5" in-out Hooker Max Flow muffler that I'll be inhereting as he upgrades from 2.5" to 3" on his FXT. The design is bigger then what should be using for a non-modified engine, but it'll leave room for expansion down the road whenever I step to hotter cams and do any head work. Section 8 did mention in his posts that despite the larger diameter design, the long runners maintained good low end torque, so I'm hoping to retaing good low end performance through the long tube design but a lot more breathability up top. I've got plans for the intake manifold as well. I'm looking at a port and polish plus ceramic coating by Grimmspeed. I have one 8mm spacer on the intake manifold with good results on bottom end grunt. I need to test the setup with the new exhaust once installed and see if I might try a second 8mm spacer depending on the shape of the torque curve. Fun stuff. :biggrin:
That thing is sweet, but I don't get it. I understand the purpose of equal length tubes, I also understand the reason for the cross tubes, but what I don't understand is the reason they kept it all seperate and did 4 into 1. Wouldn't it be so much easier to do 4 into 2 with larger piping one cross tube and then saves some money on fab work and save a little space as well. I mean if this design truely is better, by all means, go for it. I'd just like a little info on why it's better.
road runner knows alot. you should see him spit game over on nasioc. btw nice headers a little weird lookin but hooray for rublez!!
I know the basics, but I don't know specifics. Specifics require things like testing, flow analysis, things I am unable to do. I neither have the hardware nor modeling software. The basics are two things, flow velocity and pressure wave tuning. Diameter defines flow velocity. You pick a certain diameter to pick a certain peak rpm efficiency. The length is the pressure wave aspect and you're simply timing the bouncing of the pressure waves to the opening of the exhaust ports. You can pull a vacuum on the combustion chamber and promote more scavenging. Smaller diameter, lower rpm focus. Longer runner, lower rpm focus. The diameter tends to pick a peak point and the length tends to tilt the torque band to one side or the other, putting emphasis above or below the peak point. It's not that simple, but it's a gist of what happens from the design, at least what I've come across in my own searching and reading. In this sense, this header will have a higher peak point but the long runners will tilt the band towards the low end. The cross bars are an attempt to maintain good high rpm use by creating a shorter pathway for the exhaust pulse pressure waves. Basically, you get more interactions through the entire rev range. All of this is just me guessing. In a basic sense, it functions. In a technical sense, I can't say how well or in what proportions. After the install, I plan to do some road dyno pulls and compare to my current near stock setup. I'm curious to see how the power band changed. One note I will point out, thanks to williaty's work on Nasioc, the cams seem to be the primary limiter will a roll off above 4k rpm. As well, Cobb has done flow testing of the heads and intake, and there are design limitations there as well. Basically, the engine is well engineered but engineered for lower flow levels of 250 cfm and down. This favors designs focused on low and midrange torque and not high rpm use. High rpm functionality and high HP numbers mainly come from more significant design improvements towards high flow use. Apparently, the throttle body, intake runners, and heads need to be ported for increased flow. Cobb's looked into this a bit, and the gist from their old tech articles (can still be found using www.archive.org), one essentially has to do major work to fix/improve the limitations. Because of this, I have qualms using this header on my own car at my current build level. I don't intend to do major work. I intend to upgrade slowly and do some improvements towards flow, but it won't really be optimized without a ton of work. Still, I expect it to be better then stock overall, and the header is a good design and still geared for low, mid rpm use. I expect gains but nothing massive. I don't know what the power band will be though. I intend to tune the power band with the intake runner. I have the option to lengthen or shorten runner length to tilt the power band one way or the other and to port the intake runners to bump up peak torque as well. I have some additional control over just the exhaust. The beauty of a complex system like an engine is that is is the sum of many parts. It means you can change the engine's behavior in more then one way. I just don't know what optimum is unfortunately. One can only guess via various online calculators. It's just a really neat header and something I want to try out.
One of each I assume. There's what I assume is a spot for a temp sensor up near the head flange. Then there's a second down after the collector that would be for the O2.
It is a cool piece, but I am not interested. UEL is the only way for me. I would rather have the sound (and ability to go turbo) than having a huge set of EL headers and 5 more HP. I recall seeing these headers (or ones like them) for sale awhile ago when I was looking at NA parts for my car. They definitely are a unique Subaru part, my guess is not many people out there have them. good find!
Yeah, Section 8 had them for sale for around 3 years. It is odd that no one jumped on them other then the fact that he was selling them (+Thrush glasspack and rest of piping) for $900. It's TWE and well built, but relative to something like an OBX header, it's tough to justify the money spend. He cut the price in half a while ago just to get the parts out of his garage, but luckly for me no one showed interest. I think he said there were only 4 in use not under rally cars and his was the only SS version of the design. I'm just happy get a nice piece out of a garage and under a car where it should be. I'm just curious about the results, but that'll be sometime this weekend...well unless I feel like tossing it on and running with no cats. :laugh: It's designed to bolt directly to any cat back and terminates where the stock midpile would. I could toss it on using my stock remaining parts.
^ they are indeed rare then. I think I have only seen 1 or 2 people post about these on RS25. It is one of those things that when it was new had a very big "awe" factor of people dreaming about putting them on their Subies, but very few people could actually justify the money for such a purpose built header. Correct me if I am wrong but didn't 6mtizzle have a TWE header on his RS before he sold it? I thought he had one that couldnt use a normal catback exhaust though. I am very curious to hear your impression about this header and any gains you think you might get. Especially ones backed with dyno graphs
Similar in idea to the MRT manifold, but that was a 4-2 IIRC. I want to hear these things when they are on. Loud pipes save lives.
Patric Olsen did do dyno runs with his MRT and OBX headers. http://forums.nasioc.com/forums/showthread.php?t=1031584 road dyno on the mrt vs. stock(and cai) http://forums.nasioc.com/forums/showthread.php?t=816371 MRT pic for reference: Obviously a 4 to 2 to 1 stepped design. The cross brace is functional. There are some arguments about the functionality of the 4 to 2 to 1 design in terms of pressure waves. Nascar does use this type of design though. The general gist is that 4 to 1 > 4 to 2 to 1. The MRT header was well liked back in the day, but I think Cobbs 4 to 1 and TWE's 4 to 1 packages simply made more power and MRT sort of fizzled out of production.
Header's on. I'm currently running Section 8's Thrush glass pack center pipe section hooked to my bro's FXT stock back section and muffler. It's the only package that'd work until I get something custom made for it. It's pretty darn quiet but with a little rat-rod kind of tick as you rev up, maybe an exhaust leak somewhere. :laugh: It'll sit like this till my bro pulls off his back section so I can inherit and use his parts and get the full thing setup. Initial impressions of v0.5, at least testing out the header, I can't say differences are vast but there are some notable details. 1) bottom end comes on smoother, a little more pull way down near idle and a slightly rising torque band till a peak at 4k versus stock's flatter torque from 2k with a peak at 3.5k. 2) top end is up around +5hp maybe with gains ramping up from 3.5k up. The profile mimics stock. The engine has a slightly easier time up top, but the exhaust is not the defining factor of the power band. The curve mimics stock exactly on every pull. I assume the cams are the main driving factor at the moment as williaty's testing and thread on Nasioc points to. Cam gains can be considerable when finally done. 3) engine responsiveness is quite a bit better. Engine reaction relative to your right foot is more immediate. This is probably the more significant difference in behavior other then a little smoother power delivery and a slight gain of power up top. Modulating the throttle is more direct. The ECU hasn't been touched or reset and there's only maybe 10 miles on the v0.5 setup. I'd like to let the ECU adjust some before really saying anything. I'm just looking at initial numbers. I have some graphs I'm looking at of 4 road dyno pulls in 3rd gear from under 1.5k to redline. The road section is slighlty hilly, so one way is a touch faster then the other. An average of the two sets yields moderate gains below 1.5k (interestingly) and slight rising gains from 3.5k up. Stock was a hair more from 1.5k to 3.5k but just slightly. Torque and HP numbers are up just a hair. v1.0 will get put together, and I'll give the ECU some time to learn. I'll then make a series of pulls on the same stretch that I made my past pulls (not doing now, different city). Everything is SAE corrected to current weather conditions and plotted with the same car settings. I'll say I was expecting a little bit more, kinda, but the engine info I'm aware of points me to exactly what I'm getting for numbers. I'm quite happy with the low end performance of the header. I was half expecting a lot worse, but it maintained solid bottom end pull basically from near idle up. Top end does breath a tad easier, and the car doesn't feel quite so lacking up to, but there is still definite drop off in power above 5k and with either setup this is the same. I intend to try a port & polish on the intake manifold and maybe or maybe not try a second spacer. The power band shape is still decently flat, but I'm not sure what a port & polish will do to the curve. This is still running the stock intake, and I may do something with that. Although results tend to be mixed on a CAI's usefulness, most do point towards useful top end gains. I basically want to get to the point where the only limiting factor is cams. As well, measured intake temps on my car have generally been quite a bit above ambient. The intake manifold has generally been nearly hot to the touch, even after the phenolic spacers. One main thing I can do for power gains and efficiency is to bring down intake temps, and I need to do some work towards that. Spicy cams will most likely come this summer, and the redline will get bumped up some. That'll take full advantage of the rest of the system.
I'll get some sound clips once I get it properly finished. I may get a clip of the current v0.5 just because it sounds kind of neat. It might be useful if someone wants to recreate it. v1.0 is all I'm really concerned about though.
Version 0.5 = TWE header, Thrush glass pack midsection, stock FXT rear section. It's what fit together with no custom modifications. My brother is upgrading from a 2.5" custom setup using a high flow Car-Sound cat + Hooker Max Flow muffler to a 3" Maddad setup + external waste gate. His car is exhaust limited, so he's stepping up to 3". I'm inheriting his 2.5" parts. Version 1.0 = TWE header, Car Sound cat, and Hooker Max Flow muffler. v1.0 is my end point for the exhaust setup. v0.5 is just what works till my bro takes off his old stuff and I can get my car into a shop to get stuff fabricated.
v0.5 clips, mediocre mic on the cam, a little ricey but an interesting sound, decently mild inside the cabin, very livable. It goes to show that the stuck mufflers do their job well.
Hmm, I feel stupid. So the front temp sensor, um, yeah, that's a hole. :laugh: Because there was something in the hole, I would expect it to be shut. Driving around, I kept getting the inkling something was amiss. There was excess noise from the front end, not a ton, but it was unusual and something I couldn't chalk up to "it's just the header's sound." Upon doing my oil change today, I took a second look at the temp sensor bung, stuck a screwdriver through and all the way in. Oh...that is so not sealed. The second inkling that there was something wrong was the fact that my mileage went down a good chunk. From that, I knew something was amiss as well. I have a check engine light, but that was expected from the lack of a cat and rear O2 sensor. It was probably also telling me on of my cylinders was running lean or rich or something. I quickly stopped off at Ace and grabbed a bolt, cut it short, and filled the hole. Now it's a little quieter and without a motorcycle kind of sound rat tat tat. Power's up a little too. I expect mileage to step back to normal as well. The exhaust is getting worked on Friday, so I should have actual results up this weekend and some new sound clips of something that is slightly less ricey/noisy. The Hooker Max Flow has a nice, deep, truck like tone to it.
Wee, v1.0 is complete. It sounds a bit different then expected. It's amazing how much the exact same exhaust sounds different on a turbo car versus a NA car. It's a little noisy from 3k up but very subtle, stock like quiet from 2.5k down and running anything under 60mph. Something's vibrating and raspy in the higher revs. I may try to find it (probably header) and do a little heat wrap over the piping to dampen the movement. It may be unavoidable too. I do a couple road dyno runs tonight or tomorrow and get some graphs up and a couple sound/video clips. So far mileage seems to have gone back to normal after I plugged the gaping hole.
Power results. This is an accumulation of 3 run sets (both ways on same stretch of road) averaged to a mean level. I plotted my current exhaust against my past dyno runs. Basically the LSE Dyno Dynamics run is my base line. All it has is an I-Speed reflash. The second run set - blue and red. This is my car with geddesk2's throttle body spacer and Grimmspeed's 8mm phenolic spacer. TBS = smoother power delivery, IMS = tilting power band towards low rpm. Notice how much more low end was produced. Peak torque was set to the dyno run. The parts did not make power, just reshaped to the torque cuver. The thrid run set - green and yellow. This is my new exhaust. All else is the same. TWE header, Car Sound 59956 cat, Hooker Max Flow muffler, 2.5" piping everywhere. You should notice that I can remove the 8mm intake manifold spacer and retilt the torque curve back towards the top end.
Got a vid up: In car is with open windows. The camera FOV is goofy unless I zoom in, so you get kind of a fast motion effect from it...I forgot about it. The header does have some buzz from the pipes against each other. I intend to wrap them a little to keep then apart. I don't know if this translates some buzz in the exhaust note or just what I hear in the cabin.
