What Are Headers Worth?

Bobby Loudpedal takes his copy of Street Rat and scours the pages to find all the horsepower-heaven ads. You know, the ones that claim, “Just bolt this part on and its worth 30 hp.” After adding up all the claims, Bobby exclaims, “Hey, if I crank up all these parts, my motor’ll make 500 hp—eeeeeaazy...” The only problem is, power claims can be very elusive. When it comes to horsepower, too much depends on the engine and its capabilities. One of the classic horsepower adders for a hot street car is a set of headers. But what is a set of headers worth? The answer to that question is an unqualified “it depends.”

In order to get specific, a true horsepower soothsayer needs to know more about the components that are bolted to the engine. To make this concept a little less vague, we decided to do a couple of dyno tests that might make some sense out of this confusing subject. But before we get into the tests, let’s review basic engine theory. As you’ve probably heard before, an engine is really nothing more than an air pump. The more air you can pump into (and out of) the engine, the more power you can make. When designing an engine, the builder needs to determine how the engine will be used. This is important since automotive engines tend to operate over a very wide rpm band.

For example, if we want to build a drag-race engine, the engine will operate at fairly high engine speeds and probably will not be required to operate at speeds below 3,000 rpm. Conversely, if you are building a street engine that will be used to drive back and forth to school or work, this engine will spend a majority of its time between idle and 3,000 rpm. These are two completely different engines that will require two completely different configurations. Let’s assume that both are 355ci small-block engines. Let’s say that the race engine is a drag-race 355 that will operate at speeds from 5,000 to 9,000 rpm.

This engine will require a high compression ratio, cylinder heads with big ports and valves, a long duration mechanical-roller camshaft, a single-plane intake manifold, and a giant carburetor. All of this points to flowing tons of air through the engine, which means this engine would really benefit from a large-diameter set of headers to prevent limiting the airflow exiting the engine.

On the other end of the spectrum, let’s say our grocery-getter small-block 355 has stock iron heads, a very mild flat-tappet hydraulic camshaft, a basic aluminum dual-plane intake manifold, and a Q-jet carburetor. Because this engine operates at low engine speeds (relative to the race engine), this engine will pump far less air through the engine even though the displacement is exactly the same as the race engine.

Now, let’s say that we take each of these engines and dyno-test them first with cast-iron exhaust manifolds and then with a set of headers. Would you expect that each would generate the same horsepower differential? The answer to that question is a resounding no.

It should be obvious that as the engine’s potential for power increases with big heads, a big cam, and lots of compression, any type of restriction in the exhaust system (like small, cast-iron exhaust manifolds) would severely limit power since these manifolds represent a serious “cork” in the system. However, adding headers to a relatively mild small-block is only one of several restrictions to airflow. Upstream, you would still have small intake ports, a short-duration camshaft, and a relatively restrictive intake tract that would not allow the engine to make more power.

This is why you continually see engine buildup stories where a lot of emphasis is placed on combinations of parts. A big cylinder head will not necessarily guarantee a ton of power if the engine is still saddled with a restrictive intake or exhaust system. This “systems” approach to building a hot street engine is complicated by the sheer number of parts available for an engine like the small-block Chevy. There are a couple of dozen cylinder-head manufacturers that all claim to have great cylinder heads. But not all of these heads (or cams, or intake manifolds) are best in all situations. This brings us back to headers.

Test Tubes

All this information is vaguely interesting but rather dull without some real-world testing. So we put together some tests to show what a set of headers will do for both a very mild small-block Chevy and a streetable 435hp 355 that we strangled with a set of exhaust manifolds to evaluate the power loss. First, let’s take a look at the mild street small-block.

The small-block Chevy is a very efficient motor right out of the box. However, its weak spot is the exhaust side of the engine. Generally speaking, if you want to improve power on a stock small-block Chevy, the best place to start is with the exhaust system. If the car is equipped with a single exhaust, the best plan is to convert to a dual system. Mufflers also play a big part in this overall scheme, but don’t equate power with noise. There are several companies like Walker DynoMax, Flowmaster, Borla, and others that offer relatively quiet mufflers that flow a ton of air.

The 350 we baselined for this test was a basic 350 small-block with 8.5:1 compression and a painfully short-duration stock camshaft (194/202 degrees of duration at 0.050-inch tappet lift and 0.383/0.420-inch valve lift). The induction side of the engine utilized a set of stock cast-iron heads with 1.94/1.50-inch valves, a stock aluminum intake manifold, and a Q-jet carburetor with an HEI ignition. The baseline numbers were generated using a set of production, 2-inch–diameter, Chevelle cast-iron exhaust manifolds that exited into a 2-½-inch exhaust system.

The comparison test pitted the stock cast-iron exhaust manifolds against a set of 1-5/8-inch Hooker headers that were plumbed into the same 2-½-inch exhaust system. As you can see from the results of Test 1 versus Test 2, the headers were an unqualified success. If all you looked at was peak horsepower, the cast-iron engine made a best 239 hp at 4,200, while the headers pumped the power up to 255 hp at a 200-rpm–higher 4,400 rpm. This alone is a significant 16 hp, but that’s not all of the story. The headers not only reduce exhaust restrictions, which lowers exhaust backpressure, but the length and diameter of the header pipes also have a dramatic effect on the torque curve as well. This is perhaps a bigger story than just the horsepower increase.

Look closely at the two torque curves and you will see a spike of 53 lb-ft of torque at 3,400 rpm while the 3,200-rpm line reveals a torque increase of 26 lb-ft of torque. These are amazing numbers that average more than 26 lb-ft of torque between 3,000 and 3,800 rpm. But even these numbers don’t tell the entire story. We also loaded both curves into the Racing Systems Analysis dragstrip simulator to see what this power was worth. This is perhaps the most useful information of all since this is a typical change that many hot rodders’ cars appreciate.

For a real-world evaluation, we simulated a 3,500-pound Chevelle with a 3.55:1 rear gear and a Turbo 350 automatic trans with a mild 2,500-stall converter with the car running at sea level with 72-degree air. Shifting at 5,300 rpm, the RSA Quarter program simulated the Chevelle running 14.55/94.40 mph. Then all we did was change the power curve to reflect the power increase from the headers and ran the simulation again. This time, our theoretical Chevelle ran a slightly quicker and faster 14.34/96.40-mph pass. This gave us a dragstrip improvement of 0.21 second and exactly 2 mph. For a painfully mild 350, these are good numbers and the two-tenths and 2-mph improvement is right in the ballpark. Some might consider two tenths to be somewhat less than what they would expect, but this is a real-world number especially considering the engine was basically stock.

Testing The Big Dog

Next, we started with a 436hp 355 equipped with a set of Air Flow Research aluminum heads and a Comp Cams 274 Xtreme Energy camshaft with 230/236 degrees of duration at 0.050-inch tappet lift with 0.552/0.555-inch valve lift. On the induction side, we added an Edelbrock Performer RPM and a 750-cfm Holley carburetor. This engine obviously needed a very efficient exhaust system if it was going to perform well. But for Test 3, we corked it up with a pair of exhaust manifolds similar to the ones used on the mild engine test and tied them in with a pair of 2¼-inch exhaust pipes.

In this case, it’s easy to guess that the stout small-block fizzled with this much restriction on the exhaust side. Test 4 revealed how much headers are worth on a well-prepared engine by pulling out an impressive 68 lb-ft of torque improvement at 5,200 rpm (338 versus 406) that is equivalent to a mild nitrous hit in terms of pure torque. Peak-horsepower improvements were actually even more impressive with a 70hp gain at 6,200 rpm. We did the math and came up with a staggering average increase of 46 lb-ft of torque. That means that throughout the entire 4,000-rpm band from 2,600 to 6,600 rpm, the engine made 46 lb-ft of torque more with the headers at every single data point.

We plugged these numbers into the Quarter program to see what these numbers would mean to a hot street car. Obviously, a lighter car with a deeper gear and all the other goodies would respond to this kind power increase. Our cyber mule for this test case was a 3,200-pound ’68 Camaro with a 3.73 gear, sticky tires, and an automatic with a 3,600-stall converter and a 6,800-rpm shift point. With the iron manifolds, our dream-book Camaro still runs a respectable 12.58/108.7-mph run. But add the headers and this rascal produced an awesome 11.97/116.0 run that was worth 0.61 second and a stout 7.3 mph. See what headers can do?

What Did We Learn?

It should be clear by now that headers do work. Even on a bread-and-butter small-block, headers are always a good idea. Granted, headers don’t help as much when the engine is close to stock as they do on hard-core engines making more than 400 hp. We’ve also learned that even significant increases in torque don’t always mean your car is going to run a half-second quicker just by bolting on a set of headers. The mild combination improved by a couple of tenths and 2 mph. This is very typical of a mild-performance car at the dragstrip. The stronger motor and car combination picked up more, but that’s because the engine has much more potential to make power that was restricted by the exhaust manifold.

So, we come back to our original point that despite tons of testing, what are a set of headers worth? The answer is still imprecise, but you now have a much better grasp on the role that headers play in a performance engine. Isn’t science cool?