The effort and hours we racers will put into this hobby are sort of maddening. 5 horsepower can mean the difference between a win and a 2nd, 3rd, or last place sometimes. Recently my brother and I have talked about building our own engine dyno so we could test engines, develop headers, and be crazy people. We might be losing our minds, over a potential 5 horsepower.
If you are reading this, you have likely put a set of headers (or a header) on a car. Or a bunch of cars. You know the basics of why this common aftermarket addition to almost all cars makes a pretty big bump in power usually. Taking a glance at virtually all factory exhaust manifolds reveals the true purposes they were designed for....usually to remove exhaust from the car in a super reliable way, while being as reliable as an anvil (and usually as heavy). Most factory exhaust manifolds are simple, log-style cast iron pieces, which retain heat pretty well and help light off catalytic convertors, are cheap to produce, rarely crack in the lifespan of a car, and are easy to package into todays tight engine bays we so commonly see in production cars. OEM pieces are almost never optimized for performance, and we can understand why.
First off, before I go any further here, lets get one thing straight.... the PERFECT header for a car doesnt exist. You'd be giving up top end potential for midrange, midrange for top end power, peak torque, blah blah blah. The perfect header really should be optimized for the desired purpose of a car. Aftermarket headers are so misunderstood, even by most long-time, tech-savy enthusiasts. Virtually every single "off the shelf" performance header will make power (sometimes a lot of power) when compared to the factory choking, wheezing manifold, but rarely are they optimal for the engine they are bolted to. The wormhole for header-optimization is a deep one, once you start reading on the internet. I've spent the better part of a month reading about this in my free time, as I am building what I hope to be my "ultimate" header for my d15 SOHC powered scca STL-class racecar. I'll be documenting its process and theory here, and hoping to condense some of what I've learned reading into an easier-to-process article.
The engine in the racecar is pretty basic, and follows the class-rules of the Scca STL class, while also slotting in pretty nicely with NASA Honda challenge h2 . It features 11:1 compression, .425'' lift cam, stock port head, with only gasket matching allowed. A good valve job, and a stock intake manifold (either the manifold off the chosen engine or the original Engine from the car is allowed, so I obviously chose the d16a6 89' crx SI engine over the throttle body injection d15 engine manifold). The engine , with the current header I have on it, seems to fall off in power around 7500 RPM, a bit sooner than I think it should after talking with the designer of the cam and several Honda D series experts. The header is a modified off-the-shelf piece I hacked together with the help of my tig-welding-buddy, Kyle.
The current header is the popular "bundle of snakes" primary setup , with a real merge collector and megaphone I fabbed up and kyle expertly welded on. The "choke point" at the bottom of the collector was 2'', which I now feel is too big, going from 1 5/8'' primaries. The off megaphone acts as a draw on the exhaust as it lets the pulse expand to a lower pressure area rapidly, and it is likely where I found some horsepower with this setup. Megaphones can be pretty useful if packaging allows them to work on the car, and we'll get into that in the next couple articles.
So, we had a pretty good setup going here. But, over the past few months, the deeper I read into "perfect" header theory, the further I felt I was from optimized-for-the-engine. The ideal primaries (the initial tubing directly off the head) are 1.5" OD for this head, and this header is larger than that. The current prevailing theory for primaries is to keep them as small as possible without being too restrictive off the head(prevailing=NASCAR development, high end pro road racing , etc) MANY headers are larger than the port, some by a LOT, because the older thinking was simply "make everything as big as possible!" for top end....because it worked, to a point. Some theory on why it sometimes works for top end power to have a bigger primary tube is that the "step" up to the bigger tubing creates an anti-reversion chamber of sorts, keeping the column of air from wanting to reverse into the exhaust port when the valves slam shut and the velocity stops. Exhaust Gas velocity is very important, and big primary tubes are the killers of exhaust velocity . Think of them as a larger volume air compressor, with a small pump....its takes a longer time to fill them than the same small pump would on a tiny little tank. The main goal of mine in building a header for my racecar is to get each exhaust pulse as far away from the port as possible before the valves slam shut(and the column of exhaust effectively stops "wanting" to move)
Strap in. We are gonna geek out. In the next article I'll ramble for a while on how to preserve exhaust velocity, while beginning the fab work on my next header, and showing off a few other designs we will be testing against on the dyno. We'll show pics and discuss a few other road race designed exhaust systems, built by leading professionals like Calvin Elston. And.....we'll get into Anti-Reversion, the voodoo-esque technology that is emerging as the best way to make power in the exhaust system. Its about to get nerdy in here. . AND MEGAPHONES.....I freaking love megaphones. So much.