It was a darn shame to have pulled the cylinder head off only to discover that the problems were coming mostly from a pinched oil feed line to the turbo. However, some new findings revealed that I should be happy I'd replaced it with my original motor's 3.0L head.
First (as you will see in the next installment's dyno charts), airflow was down with the ported and polished Bavarian Engine-rebuilt cylinder head that had just come off the car, mainly because the exhaust ports were significantly larger than the ports in the exhaust manifold. Of course, this was our fault for not taking measurements when we first received this engine in mid 2002. We can't put the blame on Bavarian Engine Exchange, since it ported the head without knowing the actual diameters of the exhaust manifold runners. We should have spent more attention to detail.
As you'll also see next month, the resulting turbulent air in the exhaust cost us some serious early-boost spool-up and torque. So, going with the factory cylinder head, in my particular situation, was a good thing. But, let's not get too far ahead of ourselves.
While the cylinder head, exhaust manifold and turbo were off, I arranged other work to be done to improve airflow and longevity. And I had that burned-oil-encrusted turbo rebuilt.
The exhaust manifold was sent to Extrude Hone for its Abrasive Flow Machining (AFM) process, which is designed to smoothen inner surfaces. Extrude Hone not only machines such automotive products as cylinder heads, intake manifolds and turbine housings, it also handles specialized equipment and high-purity devices for the pharmaceutical, medical, semiconductor and aerospace industries. (Extrude Hone reports that every major aircraft turbine engine manufacturer worldwide uses Extrude Hone's products and processes).
When I received back the exhaust manifold, the once very grainy interior was now smoother than a baby's butt. Well, maybe there were a few pimples, but overall it looked beautiful.
As mentioned last time, the turbocharger had already been sent to Tom Rose of Performance Techniques, who completely rebuilt it using brand new parts, including bearings, thrust bearing and seals. Before putting it back together, though, he had the exhaust housing of the turbocharger thermal coated. This helps maintain heat inside the turbocharger for better response, and it helps limit radiant heat from traveling to other parts of the engine compartment. Since the Active Autowerke turbo manifold has a high nickel content for strength, I decided to contract Performance Techniques to thermal-coat the manifold as well for an even greater reduction in under-hood temperatures. It would also keep the engine compartment clean of excessive heat shielding and thermal wraps.
While the parts were out getting coated, Tom gave each turbine blade a 10-degree clip to increase its overall flow capabilities. Some of the AA Stage 3 cars have seen an increase in over 40 whp at the top end by doing this, at the same boost level (imagine being able to run up to 17 psi!). I still wasn't interested in going over 15 psi--I just wanted to flatten out the horsepower curve up top. Karl at AA also had reportedly seen a dramatic reduction in backpressure from clipping the wheel this way, helping to increase longevity.
Most would argue that simply going to a larger turbo would have been the better approach, but it wasn't that simple. As far as I know, I'm stuck with this Garrett/Mitsubishi hybrid turbo, because it's one of the largest turbos that will fit where it sits while still maintaining a proper gravity-fed oil drain from the turbo to the oil pan. Putting in a larger turbo, like AA's Stage 4 T-61 monster, requires not only a separate oil pump to drain the oil back to the pan but probably larger intake and exhaust tubing--and a larger intercooler would not be a bad idea, either. We're talking way more money than I care to spend, at least right now (remember, though; I'm impatient).
One of the things I had neglected to upgrade over the course of this project was the means of controlling fuel pressure. I had been relying on the stock fuel pressure regulator, but it was time to do it the right way for a little extra insurance, especially since evosport technician and BMW expert Ken Brightwell said he had seen a few factory pressure regulators go bad on some of the M3 race cars he'd built. I contacted Advanced Engine Management (AEM) and ordered a billet adjustable fuel pressure regulator, which is reportedly capable of supporting up to 1000 hp. In order to block the fuel passage to where the factory regulator once resided on the fuel rail, a special plug was ordered from BMP Design.
To properly feed the fuel going into and coming out of the AEM fuel pressure regulator, I ordered some Pro-Lite 350 hi-temp flexible lines with Earls AN fittings from TMR. (The two 90* fitting part numbers are 949006 and 985006, and the swivel seal numbers are 800106 and 809106. The tube adapters for the -6 Male to 5/16 part number is 165056.)
Since the turbocharger would now be retaining more heat for better spool-up and response, something had to be done to increase turbo longevity. "The biggest killer for turbochargers on street cars is hot shut-downs," said Performance Technique's Tom Rose. Realizing that my failure to diligently let the turbo cool down probably played a role in my turbo's eventual flow decrease, I decided to use the watercooling jackets.
The AA unit already comes with the water-cooling jackets in place. AA just leaves it open to be aircooled, with motor oil being the only lubricant running through it. To increase the bearings' life spans, I decided to use actual coolant this time in the watercooling jackets. I talked to AA, and per founder Karl Hugh's request, I had evosport use the water lines that run from the cylinder head through the throttle body and back to the cylinder head. Basically, instead of bypassing the cooling system through the throttle body, we'd be bypassing the coolant lines through the turbo (check your local highway laws before diverting the cooling system away from the throttle body). In order to do this, I purchased some Earls high-quality AN Fittings and 10 ft of silicone hose, all from TMR.
When the turbo and exhaust manifold came back from Performance Techniques, it was time for the cylinder head to go back on. evosport technician Steve Lee was set to finish the job, but what happened over the course of the next several days could only be deemed a serious test of patience. When the cylinder head went on the first time, one of the head stud threads in the block didn't hold past 60 lb-ft. This wasn't good, and the head had to be pulled again. Steve said it would be best to install TIME-SERTs, which are self-locking, solid and threaded bushing inserts used for refurbishing of stripped and damaged threads. I was surprised to hear this, because the builder who put this motor together had told me the block would indeed have these to accommodate the APR 11mm head studs I had going in. Steve's reply was, "No TIME-SERTS in this block, just one Heli Coil."
Steve ordered the TIME-SERTs, and over the course of the next 7 hours was able to install them in every thread. The Heli Coil was left untouched, because its hole would be larger than the others.
Now, listen carefully: As I have mentioned time and time again, learn from my mistakes. I decided to allow evosport to reuse the cutting ring head gasket I had ordered from BMP Design, since there were no more to be found anywhere in the country at the time. The head gasket had already seen 60 lb-ft, but I just wanted the car done and figured since the cutting rings hadn't been fully crushed that things would be all right (did I mention impatience?).
Steve began to install the cylinder head once again. In increments of 10 lb-ft at a time, he reached 55 lb-ft of torque per stud and had only 20 more lb-ft to go. Things were looking good for 60 lb-ft, too, until he reached the Heli-Coil. That's right--it spun. Now things were different: We needed to find something that would fit the hole and take an 11mm head stud at the same time. Steve found the solution by using an oversized KEEN-SERT, which is like a TIME-SERT but has four stakes that drive down into the threads to hold it better in place.
Finally, the motor got its noggin back on, and everything held to 75 lb-ft, just as we had hoped. Before Steve bolted the turbo back on to the exhaust manifold, he screwed in the new AN fittings I had ordered to run the new water cooling lines from the cylinder head through the turbo and back. To make his life easier, Steve connected the new silicone lines to the cylinder head prior to the intake manifold being installed.
With the fuel rail, injectors and adjustable fuel pressure regulator installed, Steve set the fuel pressure to 52 psi with no vacuum to the regulator. After the cooling, intake and exhaust systems and Karl's ECU were installed and the fluids replenished, Project M3's motor came to life. Things were sounding and looking good.
I put a few hundred miles on the car at a low 8 psi of boost before testing it on the dyno, and I noticed some funky hesitations at wide-open-throttle in first gear. It couldn't be the programming, I thought to myself. Karl's ECU programming has been smooth and spot-on with the air-fuel ratios every time!
It wasn't until the car got back on evosport's chassis dyno that we realized what was happening. One look under the car, at the outside of the motor, showed what appeared to be dried cooling crusted around the edges between the head and block, right where the head gasket sits. The head gasket didn't hold; we shouldn't have reused it. Like I said, learn from my mistakes.
The dyno testing confirmed it. At 7- to 9 psi boost, the car performed like a champ. There were noticeable gains in the low end, and the turbo came on hard. It wasn't until 10 psi and above that the graph showed a sudden drop, but not in a way that revealed timing retardation from detonation. When I pulled the plugs and looked inside the holes, it was evident water had made it into the combustion chambers. This caused the weird hesitations in first gear; there simply wasn't enough air to fully combust the watery mixture. Until I get the head gasket replaced, I'll be stuck driving Project M3 with no more than 7-8 psi of boost, though this shouldn't do any harm to the motor.
Good news is that the engine's response feels a lot crisper, especially unboosted. At 8 psi, the car was putting down over 311 whp with a highly conservative 11.0:1 air-fuel ratio and was showing no signs of detonation on pump 91 octane even at over 360-whp. With water cooling, I think the turbo should last considerably longer this time. We'll see what boost the motor can safely work with next time on pump gas (and perhaps with a race-gas mix).
Special thanks to evosport for its diligent and extensive work--once again. We're almost done, guys...I think.
|At a Glance|
|Estimated time:||20 hr @ $90/hr = $1,800|
|Extrude honing exhaust:||$500|
|Earl's fittings, hoses and lines:||$310|
|O.E. fuel rail plug:||$10|
|Cutting ring head gasket:||$230|
|Fuel pressure regulator:||$170|