Bullitt 5094's 9-Second Ride - Power Struggle Part 3

Just when the Bullitt's rate of fire was enough to take out the competition, it would seize in the barrel and spit out head gaskets. That's what Bullitt 5094 owner Bob Watson got for spending a lot of money, time, and research on his 5.0L Two-Valve Mustang. Having this sort of thing happen once is something you can recover from, but having it happen several times could convince anyone to bite the bullet.

We couldn't let that happen, so to ensure that Bullitt 5094 would penetrate the 9-second quarter-mile zone reliably, we pulled the spent Two-Valve and swapped it out for a smaller, more potent 4.6L DOHC powerplant. Over the last few issues, we've documented this swap, which was far more involved than a normal Two-Valve to Four-Valve changeover due to the car's unique Bullitt components.

This month we explain some of the driveline changes we made to the car, and get the project payoff both on the chassis dynamometer and at the dragstrip.

With the Four-Valve Bullitt up and running, we strapped it to the nearest Dynojet rollers and turned it loose. Said Dynojet belongs to HP Performance in Orange Park, Florida, which is where the 5.0L Two-Valve engine experienced its baseline dyno testing for our "Safety First" tech article (Dec. '07). While utilizing the Snow Performance water/methanol injection system, the car produced 602 rwhp and 531 rwtq with 26 degrees of total timing and 18 psi of boost. We opted not to use the water/meth injection on the 4.6 until we had it running just the way we wanted.

With the new 4.6L Four-Valve engine between the Bullitt's fenders, the Mustang made 693 rwhp and 551 rwtq. This was with only 16 degrees of timing and 15 psi of boost. The Four-Valve can probably handle more timing, but we didn't push it given the sizable gains we had already achieved. The drop in boost pressure shows how much the airflow through the engine has increased over the previous Two-Valve, as we were using the same pulley size on the Vortech T-Trim that we employed in our baseline tests. Also, where the mass airflow meter that we used on the Two-Valve was working in its operating range, the extra flow through the Four-Valve made it tap out in a hurry.

With our newly acquired 90 extra horsepower, surely 9-second e.t.'s were only a couple of passes away, but that wasn't the case. The first track sessions only offered a handful of mid-to-low 10s-not too much of an improvement over the Two-Valve mill's effort. From the timeslips, we deduced that the car's 60-foot times were 0.3 slower than previous runs, and slicing that 0.3 off again would knock over a half-second off our elapsed times.

Though we were making more horsepower, we were lacking torque down low in the rpm range. The short/straight runner intake and smaller displacement needed a looser converter to let the engine rev higher at the hit than it did with the tight 2,500-rpm-stall speed converter we were using. With that in mind, we contacted Darrin Burch with BC Automotive. Burch had provided the 4R70W transmission that was in the car, and after filling him in on our recent track outing, we decided to replace the old converter with a new design from Circle D Specialties in Houston, Texas.

We called Circle D owner, Chris Sehorn, and told him we needed a lock-up converter for a 4R70W. We also told him that the car was making 700 at the wheels. These are pretty high numbers for a 4R70W installation, and Sehorn's first question was, "You're leaving lock-up open at the track, right?" Nope-we need wide-open throttle (WOT) lock-up capabilities. Surprisingly Sehorn answered, "No problem."

He took information from the dyno sheets and notes on how the car was performing at the track, and sent us a prototype converter to try out. All we had to do was promise to return it for inspection/rebuild after we abused it sufficiently. To date, we haven't felt inclined to remove it, as it has worked flawlessly. (Bob promises to send it back sometime, Chris.)

Armed with a better-suited converter, we were ready to get back to the track-except for one small issue. Our subject Bullitt has been using a stock Mach 1 automatic ECM to handle all of the systems on the car, including automatic-transmission shift points during passes. At this performance level, we're pushing the stock EEC-V computer capabilities, and it's taken some research and testing to get it to work as well as it has. As much as we tried to prevent a WOT shift to Overdrive during the car's 130-mph passes, the EEC was overriding our best lock-out efforts and calling for a 3-4 shift just before the traps. While we do lock up the converter going down the track, we don't let the car go into Overdrive.

A WOT Overdrive shift is a very bad thing for a 4R70W passing 700 hp, and the computer snafu necessitated a trip from Florida to the BC Automotive digs in Indianapolis, Indiana, for a repair/rebuild/upgrade. Since the initial Bullitt 4R70W upgrade, Burch has come up with more innovative tricks to strengthen the internal workings of the transmission, including the intermediate clutch assembly. According to Burch, the intermediate clutch pack is the Achilles' heel of the 4R-series transmissions. With the gearbox apart, Burch confirmed that the OD drum was toast and replaced that as well. We also managed to come up with a solution to prevent a 3-4 WOT shift at the track.

With the driveline back together, it was time for some testing and tuning. With the old converter, Watson got the best 60-foot times by flashing the converter from a dead idle off the line. The Circle D converter, with its 3,000-rpm-stall speed, worked best by powerbraking the engine to 2,000-2,500 rpm and then punching it. With this technique, the car was pulling mid-to-low 1.4-second 60-foot launches that were dead straight, thanks to a recent fortification of the 8.8 axle. The rebuilt 8.8 included a full Moser spool that spins the same 4.10:1 cogs as before, as well as 33-spline axles. Jim Britts at Southern Speed in Orange Park, Florida, installed the 9-inch-style ends to eliminate the C-Clips.

Once back at the track, the 9.98 was produced with a much improved 60-foot time, but the 133 mph was unexpectedly low, actually 5 mph down from an earlier pass that day. We discovered we had some belt slippage problems, signified by a haze of black dust covering the intake. At least we finally had a single-digit timeslip.

With a new belt in place and some additional changes in the transmission shift/lockup schedule, we bested the 9.98 with a 9.81 at 140.7 mph (1.47 60-foot), and backed that up on the next run with a 9.87 at 140.4 mph. Both passes were achieved with only 16 degrees of timing and 15 psi of boost. We added two degrees of timing on the next pass: The car left like a scalded dog and then shook the tires hard. After lifting out of the throttle, the blower belt promptly shredded itself. But even that doesn't diminish the fact that we finally accomplished our goal of making Bullitt 5094 a solid 9-second ride. Just as important, the new setup had completed 30-plus dyno pulls and approximately 25 quarter-mile passes with no sign of head-gasket problems. This is a huge improvement over the old Two-Valve. Our recollection is the car would make a maximum of 10 passes or dyno pulls before we would see water in the catch can. It's nice when great advice and hard work turns into positive results.

If you're interested in more detailed information on Bullitt 5094, Bob Watson's personal accounts of the build, or would like to see videos of some of the track passes, visit www.01bullitt.com. The site also includes a forum for the original '01 Bullitt Edition Mustang owners to discuss these rare cars.

Leave Well Enough Alone? Not Us.
Mission accomplished, right? Not quite. Since we had confirmation from Vortech early on that the standard aftercooler would be a flow restriction on the new engine, we were already researching alternatives. This being a track car, we decided that staying with an air-to-water setup is the way to go. We realized that on some of the most recent track passes, inlet air temperatures rose to the 160 to 180 degree range, so we knew there was more air passing through than could efficiently be cooled by our existing system. But where do you put a larger intercooler? It was already tight in the modular engine bay, and in most instances, replacing the front seat with said intercooler, along with adding some large holes in the firewall for the air tubing, would be next.

Our friend, John McGuire, had the solution on his 9-second NMRA Modular Muscle track car, as he had installed a Precision Turbo 1000 air-to-water intercooler offset in front of his radiator. It offered greater airflow and inlet air temps so low they would occasionally cause a misfire. Watson picked up the PT1000 intercooler, and during this re-engineering session, he also eliminated the SCT draw-through MAF configuration and used an SCT 3000 tonsil-style MAF in a blow-through configuration. Jim Britts' fabrication talents were once again employed to reconfigure the new induction design.

After making these changes, we realized that the larger intercooler, with the Vortech Max-Flow bypass valve in a blow-off configuration, and a blow-through MAF has actually improved driveability. When we got the new setup on the dyno, we only saw a horsepower increase of 3, to 696 rwhp, but torque jumped up by 37 lb-ft to 588. Inlet air temps at the track went from 160-plus degrees to 60 degrees under similar track conditions. Also, boost dropped to 12 psi, but airflow increased so much that we had to install a MAF extender on the 3000 to keep it from maxing out.

The drop in boost pressure is due to the huge decrease in the charge temperature. Recalling high school Science 101, generally when you cool a gas, it shrinks, and that's what happened here. This means that the 100-degree drop in charge temperature has made the air charge much more dense, and that in turn lessens the flow restriction-basically, more power with less boost. Another upside is that the lower temperatures lessen the possibility of detonation.

At the track, there was a little more powder in the Bullitt, as it shot down the track with great speed and accuracy. Our best quarter-mile elapsed time dropped further to a 9.69 at 141.4 mph. This was promptly backed up with another 9.69 at 142.9.