Header Ho-Down

A CHP Shootout Between Four Sets of Header Sizes

Scott Crouse Jan 8, 2004 0 Comment(s)
148_0404_header_01_z 1/5

Our 440hp, 440-lb-ft Smeding Performance test mule displaces 383 ci and represents a very popular small-block segment. It retails for $4,895 and comes less carburetor, ignition, rocker covers, and accessories. With more than 60 dyno pulls on its tally sheet, it keeps getting stronger.

Put your hat down and wet your whistle pardner, because this story's about a bunch of hot air. We all know the airflow in and out of an engine is what the horsepower game is all about. Taller intake manifolds, larger cylinder heads, and radical camshafts are all players that generally increase the amount of air an engine will move. However, without the help of a properly flowing exhaust system, the engine's potential power will fall short. Big-tube headers are capable of moving large volumes of air with minimal restriction, but bigger isn't always better. A high-velocity small-tube design can be much better at promoting a strong power curve. To test this, we'll compare four popular small-block headers with different primary pipe sizes on a stout street/strip crate engine.

Our crate mule is a Smeding Performance-built 383ci small-block that has proved itself worthy during several CHP dyno-test sessions. Smeding conservatively rates it at 440 hp and 440 lb-ft of torque right out of the box. We tested this engine last month in our "Size Matters" carburetor shootout and it produced well beyond its advertised power ratings. This header shootout called for intermediate-length small-tube Flowtech headers measuring 1 ½ inches in diameter, as well as three full-length large-tube Hooker headers measuring 1 5/8, 1 ¾, and 1 7/8 inches. We would have used a long-tube header design for the 1 ½-inch tube diameter size, but no off-the-shelf piece was available.

Before we begin, let's review the important aspects of header design and the theories behind them. When mass-producing a product of this nature, it's nearly impossible to ensure an exact manufacturing tolerance from one piece to the next. The fact that header companies are able to pre-manufacture pieces that will fit on a variety of cars is impressive, but if you want more, investigate a custom-built set.

A mass-produced shelf header typically starts at $99; the custom-built header usually begins around $1,000. The builder of a properly tuned header must account for the many engine variables responsible for altering the power curve, e.g., camshaft timing, compression ratio, engine size and speed, intake design, and so on. Engine exhaust airflow tuning correlates to header-tube design, for which the most critical factors are the length and diameter of the tube and the size of the collector. All of these variables will shift the engine's power curve around peak-torque rpm, at which point the exhaust gas moves fastest.

The first dimension to consider is the diameter of the tubing. A large-diameter header tube is capable of moving huge volumes of air to promote peaky torque numbers and horsepower numbers at the upper end. A small-diameter header tube increases the exhaust-gas velocity (air speed) to help scavenge the exhaust port efficiently at low engine speeds. This promotes a strong low- to mid-range torque curve but has a tendency to restrict top-end horsepower potential once the exhaust port begins to move more air than the tube diameter is capable of discharging. The balance is to choose a tube diameter that will provide the best power curve and acceptable peak numbers.

Length plays an equally important role in tuning the power curve, as does the diameter of the tubing. Long-tube designs (typically 30-plus-inches) create a strong exhaust-gas pulse signal throughout the pipe that enhances low- to mid-range torque and horsepower. On the other hand, as the tubes become shorter, the tuning effect has less time to enhance the engine's low-speed benefits. A short header tube will allow the low- to mid-range power to fall off while increasing the upper-end potential of the engine. Many companies offer headers in intermediate and short-tube lengths mainly for clearance reasons, but these same applications can be used to alter the power curve of the engine.

Once tube diameter and length are balanced to deliver the best power curve, collector size becomes the next concern. Collectors join several primary tubes into one mutual area where all the gases are forced to combine and vacate the exhaust system. A long collector tricks the primary tubes into thinking that they are longer than they actually are. This is a way to lower the power curve to produce more bottom-end torque and horsepower without having to sacrifice top-end power. When 1 7/8-inch-or-larger primaries are ordered, it's common for them to come without collectors, because engines using tubes this big are typically intended for upper-rpm operation. So before ordering headers, the engine builder should always consider the engine's operating range and choose tubing dimensions accordingly. What you want is usable power. A dedicated drag-race engine takes little advantage of its low-rpm power, but a road-race engine lives or dies by it. We, however, want a motor that runs strong across the entire power curve, and that's why we found out which header would make the best overall power on our Smeding-built 383.

148_0404_header_02_z 2/5

1 ½-Inch Flowtech Intermediate-Length Headers

Rather than begin at the bottom and put cast-iron exhaust manifolds on the 383ci for comparison, we figured that anyone with a 440hp stroker wouldn't insult the engine with such restriction. We bolted 1 ½-inch Flowtech headers on the Edelbrock heads. Matching gaskets are part of the Flowtech kit, but we used Fel-Pro pieces because of their reputation for durability and seal efficiency. We were surprised to see that the horsepower was down from the advertised peak number by only 11 hp, and the grunt curve showed in excess of 21 lb-ft of additional peak torque above the advertised number. It appeared that the small-diameter tubing provided some impressive low-end and midrange torque numbers, while the intermediate-length design promoted as much upper-end power as possible to create an outstanding street/strip power curve.

RPM; TQ; HP

3,000; 429; 245
3,100; 434; 256
3,200; 438; 267
3,300; 442; 278
3,400; 445; 288
3,500; 449; 299
3,600; 452; 310
3,700; 455; 321
3,800; 457; 331
3,900; 459; 341
4,000; 460; 350
4,100; 461; 360
4,200; 459; 367
4,300; 459; 376
4,400; 456; 382
4,500; 455; 390
4,600; 454; 398
4,700; 452; 405
4,800; 451; 412
4,900; 447; 417
5,000; 441; 420
5,100; 436; 424
5,200; 430; 426
5,300; 425; 429
5,400; 420; 432
5,500; 414; 434
5,600; 408; 435
5,700; 401; 435
5,800; 391; 432
5,900; 382; 429
6,000; 374; 428

Average; 437; 371

1 5/8-Inch Hooker Long-Tube Headers

After such great results from the intermediate-length headers, we were curious to discover what the 1 5/8-inch long-tubes would provide. Typically, a 1 5/8-inch long-tube header is the ideal choice for street/strip engines making upward of 450 hp, but with the previous numbers looking so good, we admit that the Flowtech stuff had us guessing.

But now we changed the testing parameters, sporting a 34 ¼-inch No. 5 primary as compared to the 17-inch No. 5 primary of the 1 ½-inch collector. After a few pulls, the 1 5/8-inch headers showed that the average power numbers gained 1 lb-ft of torque and 1 hp, and did so by giving up a significant amount of power on the bottom end.

148_0404_header_03_z 3/5

From 4,100 rpm on up, the engine pulled hard, posting two new peak power readings of 468 lb-ft of torque and 441 hp. Though the total power gains carried on longer than the losses, the average power increase is insignificant when compared to the amount of lost low-end power a street/strip car could use. With a drop of as much as 29 lb-ft of torque between 3,000 rpm and 4,000 rpm, the 383's in-car acceleration characteristics would suffer dramatically and hardly make up for it on the top end.

It's clear that this particular engine is happy with shorter, smaller primaries than with longer, larger tubes. The less-than-ideal street/strip performance could be attributed to cam timing, intake-runner design, piston size, piston speed, and many other engine-related tuning factors. Rather than changing the parts combinations to pump up the bottom-end power numbers, we further increased the header size on our Smeding 383.

1 ½- vs. 1 5/8-Inch

RPM; TQ; Diff.; HP; Diff.

3,000; 415; -14; 237; -8
3,100; 416; -18; 245; -11
3,200; 417; -20; 254; -13
3,300; 416; -26; 262; -16
3,400; 417; -28; 270; -18
3,500; 420; -29; 280; -19
3,600; 427; -25; 292; -18
3,700; 436; -19; 307; -14
3,800; 443; -14; 320; -11
3,900; 450; -9 ; 334; -7
4,000; 456; -4; 347; -3
4,100; 461; +0; 360; +0
4,200; 466; +7; 373; +6
4,300; 468; +9; 383; +7
4,400; 468; +12; 392; +10
4,500; 466; +11; 400; +10
4,600; 466; +12; 408; +12
4,700; 464; +12; 415; +10
4,800; 461; +10; 421; +9
4,900; 457; +10; 426; +9
5,000; 452; +11; 431; +11
5,100; 448; +12; 435; +12
5,200; 443; +13; 438; +12
5,300; 436; +11; 440; +11
5,400; 429; +9; 441; +10
5,500; 421; +7; 441; +8
5,600; 414; +6; 441; +6
5,700; 406; +5; 440; +5
5,800; 397; +6; 438; +6
5,900; 389; +7; 437; +8
6,000; 380; +5; 434; +6

Average; 436; 372

148_0404_header_04_z 4/5

1 ¾-inch Hooker Headers; Long Tube

At this point, things began to get a little weird. Typically, as tube diameter increases, bottom-end power numbers fall while top-end numbers climb. We swapped out the 1 5/8-inch headers for 1 ¾-inch long-tube headers and found additional low-end power from 3,000 through 3,800 rpm, with increases of roughly 4 to 6 lb-ft of torque and horsepower.

In an effort to uncover the abnormal low-end power increase, we measured all header tubes again and used the same No. 5 primary pipe as the comparison. We found the 1¾-inch long-tube headers were approximately ¾-inch longer than those of the 1 5/8-inch units. The longer tubes and the engine's internal timing probably created a low-end power wave that increased the torque and horsepower slightly down low before giving in to the typical power curve.

Either way, this low-end gain was odd and totally unexpected. Once the small-block reached 4,900 rpm, it began moving some serious air and increased the total power again. Had we been testing a high-performance naturally aspirated 7,000-rpm engine, the 1 ¾-inch headers would have revealed their true top-end potential with a broader power gain high in the power curve.

1 5/8 vs. 1 ¾-Inch

RPM; TQ; Diff.; HP; Diff.

3,000; 422; +7; 241; +4
3,100; 423; +7; 249; +4
3,200; 422; +5; 257; +3
3,300; 422; +6; 265; +3
3,400; 423; +6; 274; +4
3,500; 425; +5; 283; +3
3,600; 425; -2; 295; +3
3,700; 437; +1; 308; +1
3,800; 444; +1; 321; +1
3,900; 448; -2; 333; -1
4,000; 453; -3; 345; -2
4,100; 458; -3; 357; -3
4,200; 462; -4; 369; -4
4,300; 465; -3; 381; -2
4,400; 466; -2; 390; -2
4,500; 467; +1; 400; -0
4,600; 467; +1; 409; -1
4,700; 464; +0; 416; -1
4,800; 462; -1; 422; -1
4,900; 459; +2; 428; +2
5,000; 455; +3; 433; +2
5,100; 450; +2; 437; +2
5,200; 446; +3; 441; +3
5,300; 441; +5; 445; +5
5,400; 434; +5; 446; +5
5,500; 427; +6; 447; +6
5,600; 421; +7; 449; +8
5,700; 411; +5; 446; +6
5,800; 403; +6; 445; +7
5,900; 393; +4; 442; +5
6,000; 385; +5; 440; +4

Average; 438; 375

148_0404_header_05_z 5/5

1 7/8-Inch Long-Tube Hooker Headers With the 1 ¾-inch headers making the Smeding act screwy, we could only imagine what would happen with the big-daddy 1 7/8-inchers. These headers did not come with collector flanges. We put the Hookers to the Edelbrock RPM Performer heads using the favored Fel-Pros and pulled the handle. This time the power numbers danced around like they did with the 1 ¾-inchers and we immediately checked over our previous measurements.

As it turns out, the larger the tube diameter, the more a broad tube bend is required. This in turn lengthens the headers, and in our case resulted in another ¾ inch of total tube length. Again, we felt that this addition combined with the same previous engine-exhaust-wave tuning had something to do with the up and down power curve. If we had ran high-flowing cylinder heads, a huge camshaft, and maybe even a power adder, these big pipes would have shined from 5,500 rpm all the way up.

1 ¾ vs. 1 7/8-Inch

RPM; TQ; Diff.; HP; Diff.

3,000; 430; +8; 246 +5
3,100; 436; +13; 257; +8
3,200; 438; +16; 267; +10
3,300; 436; +14; 274; +9
3,400; 433; +10; 281; +7
3,500; 433; +8; 288; +5
3,600; 435; +5; 298; +3
3,700; 438; +1; 308; +0
3,800; 442; -2; 320; -1
3,900; 446; -2; 331; -2
4,000; 449; -4; 342; -3
4,100; 453; -5; 354; -3
4,200; 456; -6; 365; -4
4,300; 458; -7; 375; -6
4,400; 459; -7; 385; -5
4,500; 459; -8; 393; -7
4,600; 459; -8; 402; -7
4,700; 457; -7; 409; -7
4,800; 455; -7; 416; -6
4,900; 452; -7; 422; -6
5,000; 449; -6; 427; -5
5,100; 445; -5; 432; -5
5,200; 440; -6; 436; -5
5,300; 436; -5; 440; -5
5,400; 432; -2; 445; -1
5,500; 428; +1; 448; +1
5,600; 422; +1; 450; +1
5,700; 415; +4; 451; +5
5,800; 407; +4; 450; +5
5,900; 398; +5; 447; +5
6,000; 390; +5; 445; +5

Average; 438; 374

Conclusion

Our goal was to deliver a straightforward story that would provide readers with a rough idea about what the advantages of real-world header sizing are in terms of power. The odd dyno curves with the larger headers sent us scurrying for answers, and the ones we found pointed us toward internal engine tuning. While this test may have best benefited our Smeding 383ci combination, there is still much to be learned from it.

All of the long-tube designs made more upper-end torque and horsepower than the small-tube pieces, but they also lost power in the midrange relative to where their peak power was made. As for the smallest of the headers tested, we meant to provide a baseline number but instead netted the best overall power curve for a street/strip vehicle. We feel comfortable concluding that the shortest and smallest tube design held true to our initial theories by promoting as much upper-end horsepower as possible with a short tube while building an impressive amount of torque with a small diameter. The four collectors were all pretty much the same, and we didn't test them thoroughly enough to notice any operational differences.

The perfect parts combination is nearly impossible to bolt together the first time. Find an engine combination you like, use the recommendations in this story, and choose a header that's bigger and shorter for the races or smaller and longer for the street. After all, it's not the size of the header that counts--what matters is how you use it.

Parts List

Description; Mfr.; PN

1 ½-inch headers, '73 Camaro; Flowtech; 31108FIT

1 5/8-inch headers, '73 Camaro; Hooker; 2136-1

1 ¾-inch headers, '73 Camaro; Hooker; 2117-1

1 7/8-inch headers, '73 Camaro; Hooker; 2239-1

Stock SBC gaskets, 1.38x1.38-inch; Fel Pro; 1444

Hooker and Stahl adapter plate gaskets for large tube headers, 1.81x1.81-inch; Fel Pro; 1407

Vortec gaskets, 1.50x1.50-inch*; Fel Pro; 1404

LT1 and LT4 gaskets, 1.39x1.41-inch*; Fel Pro; 1470

383ci crate engine; Smeding; Call

* Parts not used for this application

Sources

Hooker Headers
Aberdeen, MS 39730
662-369-6153
http://www.hookerheaders.com
Vrbancic Brothers Racing
909-930-9980
www.customcarbs.com
Smeding Performance & Machine Shop
Rancho Cordova, CA 95742

COMMENTS

subscribe to the magazine

get digital get print
TO TOP