Muscle Mustangs & Fast FordsHow To Engine
Ford Aluminum 5.4L Header Swap - All-American Header Test
We Offer A 100hp How-To
Before calling it quits, we decided to install the 1¾-inch headers from American Racing. All I can say is wow. After dialing in the fuel curve, the peak numbers jumped up to 492 hp and 442 lb-ft of torque. We were so stunned by the results that I immediately ordered the headers removed and the stock exhaust manifolds reinstalled and retested. Having never witnessed power gains like this from a simple header swap, we were convinced something else changed when we installed the headers. This allowed us to perform an A-B-A test that ensured the accuracy of our results.
After installing the exhaust manifolds, we were rewarded (or punished) with the very same 366 hp and 364 lb-ft of torque. Though the scavenging effect offered by long-tube headers is significant, this went well beyond a simple header swap. The reason for the huge change in power was the cam timing. The log-style factory exhaust manifolds allowed crossover between the cylinders during the overlap periods. The common log allowed significant dilution of the intake charge, which caused the very low power readings. Had we run even shorty headers in place of the stock exhaust manifolds, chances are the power would have been up significantly (not to match the long-tubes, but certainly better than the stock manifolds). You won't see this type of gain by adding headers to your normally aspirated 4.6L Cobra or 5.4L swap unless you combine the Cobra R intake (reflected waves from the induction system must work in conjunction with reflected waves from exhaust) and the type of wild cam timing used on the GT1000 motor. The significance of the intake manifold will come into play in Part 2, where we run the header test again after adding a supercharger.
Amazed by the results, we still had to run the remaining two sets of headers from American Racing. Equipped with the 1¾-inch headers, the 5.4L produced 492 hp at 6,500 rpm and 442 lb-ft of torque at 5,200 rpm. This compares to 366 hp at just 5,900 rpm and 364 lb-ft at 5,100 rpm for the stock manifolds.
Installation of the larger 1 7/8-inch headers on the normally aspirated 5.4L upped the peak power numbers to an even 500 hp at the same 6,500 rpm, while the torque peak was up to 445 lb-ft at 5,200 rpm. The theory was proven correct, as the larger-diameter tubing increased the power peak but traded some power below 5,000 rpm in the process. Compared to the 1 7/8-inch headers, the 1¾-inch versions offered an additional 10 lb-ft at 4,700 rpm.
The final test on this GT1000 motor was a set of custom 2.0-inch headers. Don't ask for a set for your GT500 as the 2.0-inch headers were done strictly for this test. Equipped with the 2.0-inch headers, the 5.4L produced the same 500 hp at 6,500 rpm, but just 442 lb-ft at 5,300 rpm. These 2.0-inch headers weren't ideal for this combination, as they didn't improve the peak power (compared to the 1 7/8-inch versions), they dropped the peak torque output by 3 ft-lb and lost significantly below 5,000 rpm (as much as 20 lb-ft compared to the 1¾-inch headers).
I wonder what's in store for us when we run the headers on the supercharged combination?
|Power Horsepower Numbers|
|1¾-inch vs. 1 7⁄8-inch vs. 2.0-inch (Normally aspirated 5.4L)|
As is evident from the power numbers, the 1 7/8-inch and 2.0-inch headers increased peak power slightly over the 1¾-inch version, but there was a trade off in power elsewhere in the curve. Compared to the 1¾-inch headers, the 1 7/8-inch versions offered less power below 5,000 rpm but more power from 5,000-6,800 rpm. The highest peak power was shared by the 1 7/8-inch and 2.0-inch headers, but the same 2.0-inch headers lost over 20 ft-lb down at 4,500. For most street applications, the smaller 1¾-inch headers would be best, but for those looking to maximize power past 5,000 rpm, the 1 7/8-inch headers would be the hot ticket. The 2.0-inch headers are simply too large for this normally aspirated 5.4L (both from a fitment and performance standpoint).