When it comes to performance there is but one steadfast rule: All the good air that goes into making horsepower must eventually find its way out. This means that the addition of everything from camshafts to intake manifolds, even superchargers to help improve airflow into the engine, will be useless (OK, maybe not useless but certainly much less effective) if the engine is not able to rid itself of its exhaust. Cork up a serious performance engine and watch it struggle and gag on its own fumes. Adding the right header can add some much needed exhaust flow as well as additional power by means of the scavenging process because a header is much more than a simple set of tubes welded together in hopes of directing exhaust flow. A true header provides not just a path for the exhaust but can also help draw spent gases out of the combustion chamber. The effectiveness of this so-called scavenging process is determined by a number of design criteria, some of which we'll examine and test.
When it comes to long-tube Honda headers-or headers in general for that matter-there are basically two primary designs: the tri-Y and the 4-into-1. As the name implies, the tri-Y refers to a group of three Y sections created by joining the primary tube of runner one with runner four and runner two with runner three. Once the four runners converge into two tubes these two eventually merge to form the final Y section to complete the tri-Y design. In the case of many tri-Y headers, like our test piece from Airmass, the final Y section merges to form a short collector. By contrast, 4-into-1 headers feature no such Y sections, instead merging its four runners into one short common collector. We've also seen short versions of the 4-into-1 design but the Airmass version features long primary runners to help enhance low-speed and midrange power production. The commonly accepted theory is that the 4-into-1 header offers more top end power, while the tri-Y header is designed to bolster midrange torque. Though there is a great deal more to the performance of a header than simply its overall design (4-into-1 versus tri-Y), our testing indicates that the basic layout makes for a strong indicator as to what to expect performance-wise-no matter what brand you might be considering.
Before we get to our test engine and comparison results we should take a closer look at additional header design criteria that might affect power output. A number of variables can be changed within both designs, most of which affects performance. With either the tri-Y or the 4-into-1, it's possible to change both primary and collector tubing diameters and with the tri-Y, add to that the secondary tubing's diameter, which can also be altered. In addition to tubing diameters, it's also possible to change the primary pipes' length, prior to their merging, to form the secondaries (or collector in the 4-into-1). The same is true of the secondary pipes' length and even the collector for that matter. Speaking of collectors, it isn't just lengths and diameters that can be tailored but also shapes. Collectors can be tapered, converging or even diverging at the exit. The exit diameter can be altered as well. By now the many great possibilities when it comes to header design should be obvious. Throw in the near infinite number of engine combinations, even within just one engine family-like the B-series, for instance-and it isn't hard to imagine how difficult (if not impossible) it is to build a header that works best for all applications. It is for this reason that many companies offer both tri-Y and 4-into-1 designs since both have their strengths and weaknesses, although a custom header designed specifically for any given engine combination (and operating rpm) will always provide the best performance.
What better way to demonstrate each design's tuning effect than to compare them on the same engine? To properly test not just the tri-Y and 4-into-1 header designs but also the stock exhaust manifold, we assembled a B16A test mule. The 1.6 liter was chosen for its availability because the B16A is obviously much more prevalent than any B18C mill. A healthy engine was in order. Something both powerful yet representative of what might be run on a typical street car. The short-block's modifications are minimal and consist of stock replacement, forged Probe Racing pistons and matching forged connecting rods. The rod and piston upgrade allows us plenty of latitude in terms of engine speed potential and even gives us the ability to add nitrous at a later date. Though the stock short-block will withstand plenty of abuse, we wanted the security of the forged internals during seemingly endless hours of dyno testing.
The beefed up block was topped off with a hand-ported and slightly milled cylinder head, along with a set of Crane Stage 1 camshafts. The Stage 1s offer 242 degrees of intake duration and 230 degrees of exhaust duration. The lift values check in at .457 of an inch on the intake and .425 of an inch on the exhaust. These represent a healthy jump over the stock B16A pieces. The Crane cams are optimized using a set of the company's adjustable camshaft sprockets. A Holley 68mm throttle body is used to supply airflow to the B16A intake manifold while the fuel system is composed of a modified AEM fuel rail housing a set of 30 lb/hr injectors. An Aeromotive adjustable fuel pressure regulator is relied on to control the fuel pressure on the engine dyno and a Hondata programmable ECU is employed to dial in the air/fuel and timing curves. The dyno exhaust consists of a 2.5-inch section of tubing connected to a 3-inch, 90-degree elbow and finally out to a 6-inch evacuation tube. For testing purposes, each header, including the OEM exhaust manifold, were run through this same exhaust system.
The first order of business was to run the B16A test engine equipped with the stock exhaust manifold. The factory B16A exhaust manifold consists of a cast-iron upper section bolted to a tubular lower section. Though cast iron, the stock exhaust manifold closely resembles a typical tri-Y header. Like many tri-Y headers, the stock manifold pairs runners one and four and runners two and three. The paired runners then merge into a single collector to form the final Y section. After tuning camshaft timing, the air/fuel mixture and ignition timing curves, the B16A produced 195 hp at 7,700 rpm and 139 lb-ft of torque at 7,200 rpm. Thanks in part to the torque-producing nature of the tri-Y design, the little B16A produced over 130 lb-ft of torque from 5,200 rpm (our VTEC engagement point) to 7,900 rpm. Our mildly modified 1.6L B16A was pumping out as much power as a 1.8L ITR engine, and with the stock exhaust no less. We could hardly wait to get the Airmass piece on because we just knew we would break the 200hp barrier.
First up was Airmass' tri-Y design. The first thing we noticed about the Airmass header was its significantly lighter weight when compared to the stock manifold. Our handy dandy shipping scale told us that we just saved 15 pounds by replacing the bulky cast-iron piece with the lighter Airmass tri-Y header. After comparing the stock manifold to the tri-Y we began to wonder how much power the header had to offer, after all, the two designs were so similar. Our fears were soon put to rest after the first pull showed power readings exceeding 200 hp. Equipped with the tri-Y the B16A produced 201 hp at 7,700 rpm and 141 lb-ft of torque at 7,200 rpm. Note that both the stock manifold and the tri-Y header produced horsepower and torque peaks at equal engine speeds, 7,700 rpm and 7,200 rpm respectively. Though camshaft timing and intake design play a major part in shaping the overall power curve, identical peak power engine speeds demonstrate, at the very least, consistency in exhaust design. The similarity in design didn't stop Airmass' tri-Y from bettering Honda's setup though, especially once engine speeds reached 6,000 rpm. From 7,500 rpm to 8,100 rpm, the Airmass one-upped Honda by a good 5 to 7 hp. It's also worth noting that nowhere in the rev range did the tri-Y lose out to the stock exhaust manifold.
After seeing the gains offered by the tri-Y, we were anxious to see how well the 4-into-1 header would perform. Performing a header swap is no big deal, even installing a single piece 4-into-1 header is not terribly difficult. A skilled enthusiast with only two thumbs should be able to knock it out in roughly an hour. On the engine dyno, a header swap takes all of five minutes or so, which is why we chose to conduct testing on such a dyno in the first place. The 4-into-1 header was well worth the time it took to install because it posted the highest peak power figures of the afternoon. Our 4-into-1-fitted B16A produced 204 hp at 7,700 rpm and 143 lb-ft of torque at 7,200 rpm. The 4-into-1 was good for 10 hp compared to the stock exhaust manifold but this additional peak power came at a price, a slight penalty near our 5,200rpm VTEC engagement point. The 4-into-1 design was down roughly 3 to 4 hp across a 600rpm spread (between 5,300-5,900 rpm) when compared to the stock exhaust manifold but, by 6,000 rpm, things quickly turned around and, by 7,000 rpm, the 4-into-1 pulled away with a vengeance.
Frankly, the fact that the 4-into-1 got the best of the tri-Y up top didn't surprise us but how close the tri-Y followed in the 4-into-1's footsteps-even at the top end-and bettering it by 1 or 2 hp down low, as well as supplying more peak torque, did. This alone should be enough to knock the textbooks out of at least one or two bench racer's hands.
|Dyno Data: Stock vs. Tri-Y|