Are headers really needed?

A good set of headers has long been the go-to for the best bang for the buck upgrade. But are they really needed for a high horsepower build? Has a car ever rolled off the production line with a set of long tubes on it ?

Today engines are making 300-500 hp  commonly and 700+ not unheard of and they all do it on manifolds or something like a shorty header.

There are certainly good and bad manifolds, and pros/cons to both.

Depends a lot on the kind of engine, cylinder pairing and the firing order. Also, I am assuming you mean naturally aspirated, as manifold design for turbo applications is quite different.

I always found that the exhaust manifold on my old M44 powered 318ti was pretty well done. 

To pile on to whatever gets said here- what about on a rotary?

What motor? I think the only reason you don't see lts on oe is because the cats have to be as close as possible for quickest light off. Even modern LS motors pick up big torques and hp with proper long tubes,and they have pretty optimized stock manifolds.

5.0 mustang had factory headers.

426 max wedge had a header like cast manifold

Ls manifolds look like cast headers.

Coyote engines pick up a ton of power, both peak and throughout the RPM range, with a good set of tuned long tube headers.  Other engines get minimal gains.  I feel it truly depends on the car/engine/application.  In a majority of the situations though, I'd say yes it's worth it.

On modern performance cars, the ROI is pretty small, IMHO.  A side benefit of emissions and efficiency engineering...

Used to be any engine modifications to increase horsepower was not complete without a good set of headers. Increasing air intake has to have a way to remove exhaust more effectively. I think one of the reasons you don't see tube style headers from OEM's on factory performance cars is simply cost. Tube headers are more expensive to manufacture and are less durable then cast iron manifolds. OEM's measure cost in fractions of a penny so don't want added cost or warranty issues. This doesn't even consider the costs of additional emissions testing, additional noise and heat that has to be dealt with. With the use of turbo chargers and super chargers OEM's can probably increase horsepower to what will sell to the public knowingly leaving horsepower on the table and let the consumer deal with aftermarket exhaust.

Intuitively, my gut says that while some modern performance cars can still get 'big gains' from headers, that while the nominal gains are similar to the header gains of yore...Due to the significantly higher original output, the percentage of power/torque gain is typically going to be considerably less than it used to be...But I'm more than happy to be proven wrong on that.

That being said, if the goal is making power, I think the physics dictates that you'll still generally be giving at least some up by not having long tube headers.

In reply to Gearheadotaku :

It really depends on the motor.  V8's typically need them because of its firing order.  Adjacent cylinders can be 90 to 270 degree apart.
While the Jaguar V12 has a steady 60 degree separation and can be grouped to maximize extraction.  The wonderful part is each ( of 4 ) cast iron manifolds only weigh 4 pounds. while proper race headers require 27 feet of tubing!   Just a little note; 16 pounds of cast iron weighs less than 27 feet of tubing. on a Racing Jaguar after you fit the 27 feet of tubing into the engine compartment your reward if you did everything right is a maximum 5% boost in power. 
V8's if properly done can be many times that.  

 

Be careful what you call headers.  Most "headers"  are not tuned to provide any extraction effect and thus are nothing more than tubular exhaust. 
True headers are designed based on the camshaft to provide maximum extraction at a certain RPM.  Good headers hold the same volume in each tube. Thus all tubes are exactly the same length, size, and degree's  of bends. Clearance dents definitely take the set out of the race header catagory and put them in the tubular manifold range.  
 

I love the sound headers make when they land in the scrap bin.

There are a lot of garbage ones out there and a few good ones. Purchase carefully.

Rather a limited set of data points, but the Engine Masters episode that dyno tested headers that were progressively battered until the tubes were an absolute mess raised some interesting questions.

Headers are really interesting. A properly tuned set will have standing waves with low pressure at the exhaust ports, which sucks the exhaust right out of the cylinder. They're basically a musical instrument, and if it's not overwhelmed by sheer exhaust volume you can actually hear the resonant frequencies.

You can also stagger the tube lengths so that each cylinder hits peak efficiency at a different engine speed - you lose a strong peak but gain a wider spread.

The problem is that this is all tough packaging. On my 2.0 Miata engine, I built a header with 31" primaries to get the frequencies I wanted. Four 31" tubes that are 1.75" in diameter is a tough fit.

Compared to an off-the-shelf performance header, here's the straight before/after dyno chart. No other changes. According to my calculations, that Brand X header was optimized for 9000 rpm. My little toy was fascinating to hear on the dyno, you could hear various resonances and beats as the engine moved through its rev range.

In reply to Keith Tanner :

That's a fairly significant jump for doing some math and welding some tube.

Dusterbd13-michael said:

5.0 mustang had factory headers.

426 max wedge had a header like cast manifold

Ls manifolds look like cast headers.

None of those were longtubes.

Even if true long tube headers are overkill for most engines it's clear that the OEMs are reaping the benefits of more thorough and calculated manifold design. Today's manifolds look more like the aftermarket shorty headers of 20 years ago. For example:

This is the factory manifold from a new Mazda SkyActiv engine.

Here is the manifold from a ~20 year old Protege.

That's a very cool factory setup. Notice how they managed to get the cat after the last collector, where it will do the least damage to the tuning. The ability of modern engines to alter their cam timing, ignition and throttle position in order to heat the cat quickly lets them do that, where a less sophisticated engine would need the cat physically closer to the head in order to achieve light-off. I can tell you that a Skyactive pulls no engine vacuum when it's in cat warming mode - it's got the throttle wide open. That's one of the reasons they make such a weird noise when cold.

Most reports on the factory exhaust manifolds like I have on my '61 Pontiac were they performed just about as well as tube headers.

ShawnG said:

There are a lot of garbage ones out there and a few good ones. Purchase carefully.

There are a lot of really bad exhaust manifolds out there too. In most cases, at least in my experience, a bad header is better than a stock manifold

frenchyd said:

True headers are designed based on the camshaft to provide maximum extraction at a certain RPM.  Good headers hold the same volume in each tube. Thus all tubes are exactly the same length, size, and degree's  of bends.   

Very good answer. Plus the calculations used to design a tuned exhaust system take into consideration the exhaust valve/port dimensions.

All this talk has me thinking there has to be some gains in making a better set of manifolds (to feed the turbo) on the Intrepid

the later sohc 2.3's in the rangers (the OBD2 stuff if i remember right) had what people call the 'factory header'. It's somewhat sought after for budget all-motor guys.

The 'factory header'

the older style manifold

I was always impressed with the JDM integra type r's OEM header. I guess several aftermarket companies were too b/c there are knockoffs. Currently have one on my integra and with the OEM headshield, it has a stealth look to it. Several dyno sheets show the JDM header produced power gains as good as several aftermarket brands. An OEM upgrade for US cars too. 

¯\_(ツ)_/¯ said:

To pile on to whatever gets said here- what about on a rotary?

yes they are needed. 

over stock manifolds, mass savings is one benefit. Depending on the port there are certain primary lengths and tubing diameters that are important, also the use of overflush flange as port matched will have a negative pressure like what is stated above for the miata example by Keith. You want to utilize the free vacuum effect. 

even when you buy headers on summit racing (for a lot of cars) they advise to not port match the exhaust to the cylinder head.