How critical is header size on a typical street/strip engine? We gathered up a 700hp 540ci big-block Chevy and tested four different Hooker Header sets and one custom dyno set. It did make a difference . . . read on to find out just how much.
There's been a lot written about optimizing header design, but most of us can't afford one-off, custom-built, ideal-for-the-combination headers. Realistically, we use what we can buy over the counter. At least for popular V-8 engines and chassis, major header companies such as Hooker (a division of Holley) often offer a whole passel of tube sizes. This is definitely the case for the venerable big-block Chevy, on which we recently had the opportunity to put five different header sizes to the test on Westech Performance's SuperFlow 902 engine dyno. The headers had primary tubes ranging from 13/4 all the way to 21/4 inches, including one set with adjustable, variable-length primaries.
For these tests on Westech's SuperFlow 902 dyno, the 10.93:1, 540 Dart block-based motor was equipped with Brodix Race-Rite oval-port cylinder heads, a Lunati solid roller cam (0.680/0.680-inch lift, 255/263 degrees at 0.050), an MSD billet distributor, and a Holley 950-cfm carb atop a Wilson 1-inch spacer and Wilson-ported Edelbrock Victor Jr. intake.
From past experience, we knew milder engines aren't that picky about header design; on the other hand, full-race powerhouses almost without exception like really big headers. HOT ROD's pump-gas 540 seemingly fit the bill nicely (see HOT ROD, May, July, and Sept. '08 for complete buildup and test information on this motor's various stages). When detuned to its midrange 695 to 700hp level, the 540 comes in at around 1.26 hp/ci, making it a perfect platform to really wring out a bunch of different header sizes. Westech ran all tests on 91-octane pump gas, pulling from 3,000 to 6,500 rpm. Much as we would have liked to, the big 540 had too much torque to pull any lower on the SuperFlow dyno. Equipped as it was, the engine wanted 36 degrees total advance, and once the fuel curve on the Holley 950-cfm 4150 Ultra Series double-pumper carb was dialed in, the engine ran best with this setup throughout the entire test series. The results we'll be reporting represent the average of the two best runs for each combination, but before we get ahead of ourselves, let's briefly look at what exhaust theory has to say about header sizing.
Unless specifically stated otherwise, all header tests used one of Westech's 18-inch extension sets, which are machined to accept the wide-band oxygen sensors needed to monitor the air/fuel ratios as an aid for correctly dialing in the carburetor. The 31/2-inch-diameter slip-fit version seen here fits the 21/8-inch race headers.
The concept of equal-length tuned headers as scavenging devices is based on the behavior of gases, organ-pipe theory, and wave-action dynamics. I've discussed this in detail before (see HOT ROD, Pit Stop, Sept. '07, "Header Design Theory"), but the short version is that longer header primary-tube lengths usually add low-end torque, while shorter lengths should move the powerband up into a higher-rpm range. In other words, primary-pipe length determines the engine speed at which the system tunes.
ARP head bolts and studs are taller than stockers. Watch for interference with the header flange and grind the flange for clearance as needed. Spark-plug-to-header interference on a big-block usually isn't a problem.
Primary-tube diameter establishes the velocity of the exhaust-gas slug as it travels down the pipe. In simple terms, you want the tubes big enough so they don't form a restriction but not so large that the inertia of the incoming gas becomes insufficient to scavenge the cylinder of all exhaust gases.
Longer header collectors tend to broaden the header's power and torque tuning range below the peak torque rpm point. In the real world, playing with longer collectors is an easy and quick tuning trick that can lower quarter-mile e.t.'s by fattening the torque curve, provided the rear suspension can manage the extra torque without breaking the tires loose.
So much for basic concepts. Let's get that dyno rockin'.
1 3/4-Inch Headers (Hooker 2118)
685.7 LB-FT AT 4,400 RPM
667.1 HP AT 5,900 RPM
We tested the headers starting with the smallest set, then moved progressively up the food chain to the largest set. First up was Hooker's entry-level headers with 13/4-inch primary tubes and 3-inch collectors. At only 30 inches in length, the primaries were quite short, apparently designed more for convenience than for maximum performance. This small-tube set in theory should maximize low-end torque at the expense of top-end power, and in fact the headers' 685.7 lb-ft torque peak did occur at just 4,400 rpm; 667.1 hp was on tap at 5,900. These rpm values turned out to be the lowest peak points of the entire test series. From prior experience, we knew the 640 in its present configuration should exceed 690 lb-ft and 690 hp, so there's no doubt the small-tube headers were holding the engine back.
1 7/8-Inch Headers (Hooker 2250)
696.8 lb-ft AT 4,600 rpm
691.8 hp AT 6,000 rpm
Hooker PN 2250 has 17/8-inch primaries and 31/2-inch collectors. The primary tubes were 5 inches longer than the 13/4-inch-od set-35 inches as opposed to 30 inches. In fact, 34 to 36 inches is typically the sweet spot for most serious performance mass-produced headers. The torque peak increased 200 rpm to 4,600 rpm, while the power peak occurred 100 rpm higher at 6,000 rpm. Power would continue to peak at 6,000 through the 21/8-inch header tests. At the peaks, output was up by 11.1 lb-ft and 24.7 hp, the largest incremental improvement we'd see in this test series. Overall average output also showed a significant increase, with the 17/8 set outperforming the 13/4 tubes everywhere except at 3,000 and 3,400 to 3,600 rpm. Over 3,600, as rpm increased, so did the amount of gain, to the point that the 17/8 headers were worth more than 30 hp at 6,500 rpm.
2-inch headers (hooker 2217)
696.8 lb-ft at 4,600 rpm
702.0 hp at 6,000 rpm
Switching to the Hooker 2217 headers with 2-inch primaries and 31/2-inch collectors breached the 700hp mark for the first time in these tests. Although, torque actually declined a tad by 1.2 lb-ft. In terms of total area under the curve, the engine did gain another 4 lb-ft and 4 hp on average. Westech next removed the 18-inch collector extensions used up to this point and the engine nosed over on the bottom, with a massive power- and torque-curve dip below 4,500 rpm. The engine was also down slightly above that point: Power even dropped by 6.5 hp at the peak. There was also a 2.7 lb-ft loss at the new, 100-rpm-higher torque peak of 4,700 rpm. The dominant influence of collector length would continue throughout these tests.
Hooker PN 2386 adjustable headers fit '67 to '69 Camaros and '68 to '74 Chevy II/Novas. Intended for race cars, they won't clear power steering or air conditioning, although they do accommodate power brakes. We found collector length had a much greater effect on the power curve than varying the primary-tube lengths did.
2 1/8-inch adjustable headers (Hooker 2386)
695.4 LB-FT at 4,700 RPM
707.7 HP at 6,000 RPM
With their adjustable-length primary tubes plus supplied collector extensions, Hooker 2386 race headers are designed to be tuned in the field. Using Westech's own 18-inch oxygen-sensor bung extensions, we tested 30-inch, 34-inch, and 36-inch primary configurations. The 30-inch and 36-inch primaries were also tested without collector extensions.
Overall, the torque numbers were very close to those in the 2-inch tests. Looking at total area under the curve, the 36-inch primary configuration with 18-inch collector extensions was top dog, averaging 581.8 hp at 6,000 rpm and 644.9 lb-ft at 4,700 rpm. These numbers show an average improvement over the best 2-inch header configuration of just 0.7 hp and 0.5 lb-ft. The minimal overall average improvement was generated by trading off a little low end for more midrange and top end. The overall average torque output would be the highest seen with any of the tested header sets.
Generating 695.4 lb-ft at 4,700 rpm, the 36-inch primary configuration also had the highest peak torque number of any of the 21/8-inch header configurations. However, the 540 made the highest peak power number with the 21/8-inch adjustable headers after removing the 18-inch collector extension but retaining the 36-inch primaries; as with the 2-inch headers, the engine now became peaky, losing power downstairs. Running the 21/8 headers with 30-inch primaries and no collector extensions dropped the numbers further. This engine just does not like short primaries.
2 1/4-inch Headers (custom)
687.7 lb-ft at 4,700
716.6 hp at 6,100 rpm
The last set tested was Westech's pet dyno headers with 21/4-inch primaries and 4-inch collectors. With 18-inch collector extensions, the giant headers were worth about 9 hp over the best previous 21/8 combo. Now peaking at 716.6 hp at 6,100 rpm, the 540 generated the best peak power numbers yet seen with the 540's present camshaft and cylinder head combination. However, at 687.7 lb-ft, peak torque was down by 7.7 lb-ft. Although overall output change over the entire test range remained within one number on average, this was achieved by skewing the power curve toward the top end. Below 3,600 rpm, the engine was down up to 23 lb-ft and 13 hp, yet at 6,500 rpm, the headers gave the engine about a 14hp, 14-lb-ft advantage.
Conclusions
At least on this engine combo, the tests show that collector length has a much greater effect than changing the primary length. The multiple combinations evaluated using the 21/8 adjustable headers illustrate this point best. Leaving the 18-inch collectors in place even as primary-tube length was varied, differences in torque and power numbers were minimal. But removing the 18-inch collector had a large impact, making the engine peaky. Without the collector extension, the engine became more sensitive to primary length changes as well.
Overall, the peak torque and power rpm points remained close on all tests, with a maximum variance of no more than 400 rpm. The ported Wilson intake, the Lunati cam, and the Brodix heads primarily determine this engine's general operating range.The peak torque number changed by only about 10 lb-ft overall, as predicted by the theory that an engine's peak torque number is closely tied to its displacement. On the power side, moving up from the baseline 13/4 headers to 21/4 headers was worth 50 hp. Total area under the power and torque curves was similar from the 17/8 headers on up-it's just that the meat of the curve tended to move slightly upward.
If the engine is going in a properly geared race car, both the 21/8 and 21/4 headers perform very close to each other, with the latter getting a slight nod due to superior top-end power output. Consider the 2-inch headers for all-round street use or in a heavy car where it is desired to maximize low-end torque production. Leave the 13/4 and 17/8 sets for milder engines. They just don't cut it on a big roller-cam 540.
DYNO RESULTS SUMMARY HOOKER HEADERS PART NO. CONFIGURATION VALUE PEAK @ RPM AVERAGE OVERALL LOW AVERAGE (3,000-4,700 RPM) HIGH AVERAGE (4,800-6,500 RPM) 2118 13/4" primaries, 3" collectors w/extension Torque (LB-FT) 685.7 @ 4,400 627.9 648.3 607.5 Power (HP) 667.1 @ 5,{{{900}}} 563.9 479.1 648.8 2250 1 7/8" primaries, 31/2" collectors w/extension Torque (LB-FT) 696.8 @ 4,{{{600}}} 640.6 652.2 629.0 Power (HP) 691.8 @ 6,000 577.2 482.3 672.1 2217 2" primaries, 31/2" collectors w/extension Torque (LB-FT) 695.6 @ 4,700 644.5 653.3 635.7 Power (HP) 702.0 @ 6,000 581.1 482.7 679.6 No collector extension Torque (LB-FT) 692.9 @ 4,700 620.9 610.7 631.1 Power (HP) 695.5 @ 6,000 563.9 453.2 674.6 2386 2 1/8" primaries, 31/2" collectors w/extension Torque (LB-FT) 693.1 @ 4,700 644.2 649.8 638.6 Power (HP) 706.4 @ 6,000 581.5 480.1 683.0 As above w/4" primary extensions Torque (LB-FT) 692.2 @ 4,700 644.0 650.3 637.7 Power (HP) 705.5 @ 6,000 580.9 479.7 682.0 As above w/6" primary extensions Torque (LB-FT) 695.4 @ 4,700 644.9 650.0 639.9 Power (HP) 707.7 @ 6,000 581.8 479.4 684.3 6" primary extensions, no collector extension Torque (LB-FT) 695.2 @ 4,600 633.2 627.1 639.4 Power (HP) 709.5 @ 6,000 574.4 465.1 683.8 No primary or collector extensions Torque (LB-FT) 690.0 @ 4,700 628.6 618.6 638.6 Power (HP) 708.4 @ 6,{{{100}}} 570.4 457.6 683.2 Custom 2¼" primaries, 4" collectors w/extension Torque (LB-FT) 687.7 @ 4,700 644.7 646.4 643.1 Power (HP) 716.6 @ 6,100 582.7 477.2 688.3Test Header Specs And Prices
All headers below except the dyno headers are manufactured by Hooker Headers and were priced at Summit Racing on 6/14/08. All linear dimensions are in inches or fractions thereof. Original big-block frame stands and mounts may be needed to ensure correct fit in the listed stock applications.
PARTS AND PRICING DIMENSIONS APPLICATION ACCESSORY CLEARANCE TRANSMISSION CLEARANCE HIGH-TEMP BLACK PAINT METALLIC-CERAMIC COATING MANUAL AUTOMATIC PART NO. PRICE PART NO. PRICE PRIMARY COLLECTOR PORT SHAPE PS PB AC COL. FLOOR COL. FLOOR 2118HRK $449.95 2118-1HKR $685.95 13/4 x 30 3 x 10 Square '70-'72 Camaro Yes Yes Yes No Yes Yes Yes '64-'74 Chevelle Yes Yes Yes Yes Yes Yes* Yes* '71-'74 fullsize Yes Yes Yes No No Yes Yes 2250HKR $499.95 2250-1HKR $729.95 1 7/8 x 35 31/2 x 10 Square '70-'72 Camaro Yes Yes Yes No Yes Yes Yes '68-'74 Chevelle Yes Yes Yes Yes Yes Yes Yes 2217HKR $499.95 2217-1HKR $729.95 2 x 35 31/2 x 10 Square '70-'72 Camaro Yes Yes Yes No Yes Yes Yes '64-'74 Chevelle Yes Yes Yes Yes Yes Yes Yes 2386HKR $829.95 2386-1HKR $1,195.95 2 1/8 x 30-38 31/2 x 12-18 Round '67-'69 Camaro No Yes No No Yes No Yes '68-'74 Chevy II/Nova No Yes No No Yes Yes Yes Westech custom dyno headers 2¼ x 34 4 x 12 Round — — — — — — — —
*Except Powerglide.
