We started with a 1966 Porsche 912 engine, type 616. The case and heads believe it or not had never been touched. That could not be said for the crank, rods, pistons, and cylinders, all of which were thrown away. My engine builder and longtime friend, Jake Raby, of Aircooled Technology, had never seen such a tired engine. He said that it was a miracle that I had made it from Portland to San Francisco and then back to IL. That said, the heads looked wonderful and the case didn't need to be align bored.
The heads went to Len Hoffman, of HAM Inc. in 2005 I'll come back to the heads later, as they were a long time in the making and were not ready until days before the engine was scheduled to come together in April, 2006.
The last part (and granted a very important one) needed was a camshaft, and not until we had flow numbers did we know what cam to run. We ended up choosing an SX-3 grind. Many say that such a cam is not streetable, I beg to differ. With sufficient displacement (and proper head work), you can use more lift and duration when you add more displacement. In our case, we started with a 82.5mm bore and finished with a 90mm bore, adding almost 300ccs of displacment. This allowed us to have a powerband with peak hp figures between 4000-6000 rpm with peak toque from 4000-5500 rpm, and 70% of peak torque from 2000 rpm. We discovered this from William Noblitt's 1800, which had a very flat and wide torque curve considering he was running a fairly hot cam with more compression. Our engine on the other hand, has very little compression, just 9.25:1, designed to run on even the crappiest 91 octane gas found on the West Coast. A nice rule of thumb we've noticed for the 356 and 912 is that for every 80ccs of added displacement over 1720, you can run a cam with about 15 degrees more duration and a proportional increase in lift and still keep the rpms down, torque curve flat, and engine driveable. I can drive my 1883 in traffic and do so fairly often. I also can drive the car with my wife on a weekend outing without hearing any complaints about noise (or my poor stickshift driving). You don't need a heavy clutch- I got away with using an unsprung type 1 ceramic clutch and stock pressure plate. They've been proven to hold even 200hp without slipping and have wonderful wear characteristics.
Back to the heads. The factory made the 912 heads with roughly a 1:1 E to I ratio, or the ratio of exhaust to intake flow. Typically you want the intakes to flow more than the exhaust. Our porting focused on the intakes and we were able to get the intake to effectively flow 25% more, correcting the E to I ratio to 75% across the board. Another neat trick was the re-shaped combustion chamber, allowing a larger intake valve while not unshrouding the valves such that you could run a stock bore with our heads. This also preserved the total chamber cc's, keeping the squish area more pronouced and chambers tighter. Our heads were 58.5ccs, but we've been able to creatively add material to get chambers even smaller without flycutting, which weakens the head substantially. We reduced the valve stems to 8mm and while increasing the intake valve size, we still managed to save 14 grams on the intake and a whopping 24 grams on the exhaust. This allowed up to reduce the spring pressures while increasing the redline at the same time- the valves do not float even at 7000rpm. Lower spring pressures do wonders to improving cam and lifter life, especially when running large durations on very reduced base circles. It also helped with the switch to a tapered thin wall chromoly pushrod, saving some 30+ grams over traditional pushrods.
Balancing is very important. Our goal was to get the whole assembly down to Raby's race specs of .01 grams, which we managed to get pretty close. The engine runs silky smooth and idles smooth as low as 650 rpm, very uncharacteristic of an engine with so much cam. It also helps that we used a standard weight Scat crankshaft and an unlightened 215mm flywheel. The extra mass didn't affect how quickly the engine revs but does make it much smoother and easier to drive, for those who are not the best stick-shift drivers, like myself. Also, the Scat crank was 8 grams out of balance out of the box, so make sure to have your whole assembly dynamically balanced! Jake charges between $200-400 depending on what level of balance you are looking for. The Carrillo rods were perfect out of the box by the way.
In hindsight, a Webcam 86b would have been a better choice even though it has more duration. For a slightly more torquey engine which would probably only sacrifice a few hp and bring the rev range down 500 or so rpm an 86a would be great. All of these cams, including the SX-3 are straight pattern cams and are best suited for the corrected 75% E to I ratio. For those with more traditional porting and/or un-modified or otherwise stock heads, a split duration cam is needed to best accomodate the 1:1 E to I ratio. Some great cams to consider are the Norris 337 and 356 and from Elgin, the 7010-17 and 7208-19 are excellent choices, the latter being a hotter cam and the one that was originally used by William Noblitt in his original 1800. He has since put a Webcam 86b and upped the compression from 10.75:1 to just over 11:1, which has netted him a 15% increase on the dyno in torque numbers, which puts him in the 180ish HP range, which is the highly sought magical 100hp per litre highly.
Even with reduced base circles, it gets very tight inside the engine. In fact, the excess threads of the rod bolts were hitting the cam lobes and had to be shaved off after the rods had been torqued and installed to provide sufficient clearance.
Basically everything else with the assembly went according to plan. We ended up with .060" deck, which is on the upper limit of our recommended deck height of .040-.060", which netted us 9.25:1. We could have shortened the cylinders .020" to get us up closer to 10:1, but on a tight schedule, we decided to go with what we had.
As far as rings are concerned, you can pretty much get away with almost any ring. We do however recommend STD or ligher tension rings, say 10-12lbs on the low side to 20 lbs on the high side. The rings fitted to my engine were high tension, well over 20lbs and may contribute to shorter ring life. That said, due to the expansion characteristics of Nickies, you can run pretty high tensions and very small ring gaps while still having normal ring life or in some cases, longer ring life. We have even devised a gapping procedure that involves setting the minimum gap cold, in this case we set it at .003" at 28F, which comes out to .0032" per inch of bore ring gap. Jake Raby, my engine builder, as experimented with as little as .001" total ring gap at 50F in his race engines with Nickies cylinders. In every case however, that involves having a ring gap no larger than .0042" per inch of bore. Any larger and you get great cold leakdown figures, great compression numbers, so-so hot leakdown numbers, and large amounts of blowby. You cannot gap rings per JE's recommendations, the Porsche specifications, or like a cast iron cylinder, or you will be very sorely disappointed. Another thing to contend with is piston to cylinder clearance- it should be tight, say .001-.0015", and more performance will suffer- our cylinders are designed to run rediculously tight clearances with JE's high strength forging. In most cases, you get quieter operation than even high-silicon pistons in cast iron cylinders.
For the exhaust, we used Skirmant's race header fitted with a turbo muffler 3" in and out, modified to allow us to fit Precision Matters oil pump cover full flow filter.
The engine ran smooth, cooler, and made more torque with the muffler on. If you're running an engine like this on the street, it's worth using a race header with a little bit of fabrication to fit a muffler. That said, this race header will not fit with the Precision Matters oil pump cover full flow filter. You will need to extend the primaries about 3 inches for it to clear and not be right on the heat shield. While you are at it, you should also weld an o2 sensor bung in the collector for later tuning. It is well work the extra effor to fit a full flow filter, as it will reduce engine wear, increase oil and engine life, and reduce maintainence. It is a must for all 356 and 912 rebuilds, hot rods or otherwise stock.
We chose against the added complexity of an external oil cooler or the reduced ground clearance of a deep sump. We did however use a billet aluminum oil cooler in place of the stock oil cooler, which has a higher temperature delta. The only drawback is that it increases the air going to cylinder number three significantly, so monitoring of your CHT is very important. The temperature differential between cylinders can be excessive, so check it on #3, if not all cylinders. Monitoring EGTs are equally important for tuning, but to some extent an LM-1 can be used (more on tuning later). If your engine is going to see real track time, an external oil cooler with the stock location delete would be a better move. My 912 is for street use, so it's not so important to have the extra cooling or even an accusump, for those long sleeper curves where there might be oil starvation. We used Mainely Custom by Design's oil temperature dipstick to verify oil temps and calibrate the oil temperature gauge in the car.
Now to the fuel system. We were going to do EFI, which I still highly recommend. Alan at the Stable likes Motec. My engine builder uses SDS, since it's very easy to set up and program even for the most novice of uses. Due to time constraints, we decided to use 44 IDFs. We probably could have squeezed another 10-15% more power from EFI, but we're happy with the results nonetheless. 40 Solexes would have been a superior setup, but we had new Webers available to us from another enigne I was building. I'll talk about tuning issues another day...
We chose to go with a single plug ignition and used a Mallory Unilite distributor. We set the timing at 28 degrees max advance, without any concern to the static or idle timing. We actually ended up with about 0 degrees at idle and it runs great. We also used Nology capacitize discharge wires with plug gaps opened up to .035".
There are a few other things that I thought I would mention. When using the Scat crank, you will need to narrow the center main shells because of the larger journal radius. Don't remove the tangs per what one particular "expert" instructs- gently sand the shells on the sides on very fine sandpaper until you have clearance.
The tangs are very import to indexing the bearing, and if you remove it, the bearing might as well be floating. We also learned the hard way that you do not use the o-ring gaskets on the head stud fasteners. Use a non-hardening sealant on the face of the fastener that seals on the head. Second, you absolutely need to get yourself a set of Skirmant's valve cover gaskets. We kept sucking in valve cover gaskets until we fit a set on the engine. Make sure that all the engine tin fits tightly- the gaps in the tin, when sealed, dropped head temps about 10F. Lastly, the breathers need to be set up and both heads should be vented. We chose to put a breather to each head, and although this may contribute to some oil consumption rather than condensing the vapors back into the crankcase, it's for the better.
The heads will need to be retorqued after break-in for sure. The heads should be torqued to 24lb/ft and no more than 28lb/ft. You will also need to adjust the valves a few times through the break-in process. Set the valve lash at .008" with chromoly until the engine is fully settled and no more changes in valve lash are measurable. I know that .008" lash cold is loud and your engine might sound like a diesel, but it's better than burning a valve if your valves get tight. I still run .008", but you can probably get away with .004 on the intakes and .006 on the exhaust say after the first in service oil change at 3000 mi. We used an aircraft piston non-detergent engine break in oil (exxon) for break in along with a pint of GM EOS, which is a ZDDP engine oil suppliment to reduce cam/lifter wear and scuffing at break-in. In lieu of non-detergent, Castrol GTX with the ZDDP engine oil suppliment is recommended. After a whole day on the dyno, the engine had 1-2% leakdown on all cylinders, which lent me to believe everything was nice and broken in, so we switched to synthetic for the second day of dyno testing. More important is frequent oil filter changes during the process. We chose Royal Purple's Max Cycle synthetic, as we have seen this oils ability to drop oil temperatures significantly over most conventional (dino) and other synthetics. More on oils later... under normal break-in conditions, you should wait until at least 1000 mi or even 3000 mi to switch to a synthetic.