Engine development isn’t rocket science, but a background in theoretical physics doesn’t hurt. In 2015, Rod Pobestekof PR Technology came to Oskar Elmgren of Elmer Racing with a need. Pobestek was in need of an extremely lightweight, rigid, high-capacity Porsche engine. To maintain compliance with WTAC (World Time Attack Challenge) regulations, it had to retain Porsche’s M44 water-cooled inline-four-cylinder architecture — meaning the engine’s bore-spacing and main housing bore still had to mimic the original 3.0-liter mill.
The result of Elmgren’s efforts is a 4.0-liter engine weighing just 233 pounds in long-block dress, capable of 3,000 horsepower configured as a drag-racing engine, or a 1,500-horsepower-capable circuit-racing engine. This lightweight beast would be essential to carrying corner speed in Pobestek’s brainchild — the production Porsche 968-based RP968.
Elmer Racing got its start in a very organic way. Elmgren grew up racing karts with his family in California. Speed was a family affair, and after moving back to Finland, he became enamored with not only speed but the “why” and “how” of going fast. A university education in theoretical physics augmented his need for speed and growing mechanical know-how.
The Elmgren family was racing Rover Minis, and as engineers do, Oskar was searching for advantages. One such advantage was a custom piston for the small British A-Series powerplant. Oskar realized that “no one was interested in small-quantity, one-off work like this.” This is where Elmer Racing got its start Before long it was CNC-prototyping a bespoke, billet-aluminum, eight-port cylinder head for the Mini 998cc A-Series engines.
Design symmetry of the engine block likely wasn’t just a happy accident on Porsche’s behalf. However, it did work to the benefit of Pobestek and Elmgren. In the case of the 16-valve 968 engine, the head is reversible on the block. Heat management is essential in racing, and even more consequential on turbocharged applications.
Pobestek’s team at PR Technology managed to reverse the intake and exhaust positions on the OE mill. This was the basis for creating a turbo system that allows room for exhaust plumbing to safely expel gasses while shedding unwanted heat opposite a right-hand-drive cockpit.
Elmer Racing tuned this layout to optimize the cooling system to function in this orientation without hot spots or steam pockets. Computational fluid dynamics modeling and simulation made many of these changes clear in a manner of what Oskar called “just putting in the time to do it.”
No Replacement for Displacement
As Porsche was building what it thought would be the successor to the 911, it manufactured a block with an increased 4.800-inch bore-spacing. Actually, it made a couple of blocks with a 4.800-inch bore-spacing — the M28 V8 and M44 inline-four. This modular design strategy allowed Porsche engineers to use a mass-production, parts-bin strategy where smaller parts could be interchanged, as in many areas, the M44 was literally half of the M28.
The crew at Elmer Racing were able to put another 10mm into the bore diameter without a hefty cooling penalty on this billet beast. The four-cylinder was now rocking a bore size much like a 572 Big Block Chevrolet at 116mm (4.567 inches). This increase created room for considerably larger valves and added a turbo-spooling 700cc to the engine’s displacement.
The extra capacity keeps the calibrator from having to run head gasket-eating, aggressive anti-lag, or unwieldy and unforgiving nitrous-oxide to light off a large turbine. Additionally, big cubic inches keep throttle response much sharper below 5,500 rpm.
The compression ratio is kept in check at a very conservative 9.0:1 ratio. Rod length wasn’t disclosed, but Elmgren asserts, “It’s always best to run the longest connecting rod that will fit. But, it’s not important enough to change engine capacity or deck height to accommodate a longer rod.”
We do know they’re Bill Miller Engineering 500-series aluminum rods, run at what Elmgren called “very conventional vertical oil clearances due to cold start constraints.” He adds, “The compression rigidity is the same as, or slightly more than, a steel rod due to the rod’s cross-sectional area.”
Crankshaft weight was kept in check with a billet unit, again from Elmer Racing. Modern design techniques using finite element analysis, and the “blank canvas” afforded from billet construction allows for weight savings through the use of narrow, profiled counterweights and piston-guided rods on a thrustless crankpin.
The oil pan is integrated into a girdled bedplate for torsional rigidity and harmonics while being easily serviceable. The bedplate features slightly larger fasteners (and more of them) to tie the main housing bore into the crankcase skirting, as well as evacuation ports for the modular dry-sump and evacuation pump. The engine is attached to the RP968’s chassis at both the front and back of the engine, causing it to be a stressed member.
Crankshaft and rod bearings can be inspected in-situ without the fluid mess associated with an engine-out service. The team believes this feature can make a difference at this level of racing. Cylinder bores or rod bearings can be inspected in 10 to 15 minutes versus 10-or-so hours in a GTR’s VR38DETT. If required, the cylinder head can be removed and the engine can receive an in-frame overhaul in the pits — not unlike a long-haul truck.
Gas in, Gas Out
The engineering challenges were not limited to just building a lightweight, stiff, inline-four that wouldn’t rattle itself apart. The team still needed a cylinder head to efficiently move enough air for a massive 4.0-liter engine to carry a redline from 8,500 to 10,000 rpm.
The billet 16-valve head uses optimized cooling channels and CFD-designed port geometry. 46mm (1.811-inch) titanium intake valves let the boost in, and 39mm (1.535-inch) exhaust valves spool the turbine. While the circuit-spec camshaft duration seems very conventional, with less than 300-degrees of seat-to-seat duration and 13.2mm (.520-inch) of valve lift, the ramp rates are very aggressive.
To compensate for the violent valve events, Elmgren designed the cylinder head with larger cam follower buckets. The smoothed pentroof chambers are kept as compact as possible to minimize the distance between the spark plug electrode and the cylinder wall. No flow-benches or pitot tubes were used in the development of the cylinder head because of the clean-sheet design involved.
Port shape and volume were simply created as needed based on airflow demand, valve size, bore size, and lastly, packaging. The ports’ cross-sectional area was able to be shaped with the appropriate taper from the manifold port entry to the valve, based on pressure pulse versus velocity. This is used in concert with valve timing to create a truly tuned cylinder head worthy of the displacement and optimized for the intended power delivery.
Thor’s block- and head-sealing methodology is very orthodox. According to Elmgren, the flanged iron sleeve is as thick as it needs to be to avoid cracking, blowing-out, or blow-torching the very ordinary multi-layer steel head gasket. Once again, the almost-clean-sheet nature of this engine development allowed Elmer Racing the ability to optimize deck rigidity and heat transfer at the same time.
This resulted in a deck design that one might refer to as a hybrid open-deck, where the tubulating ribs support the cylinder liners a bit farther down the water jacket. Proper clamping-force distribution is what keeps the gasket sandwiched between the decks.
Elmgren indicates he’s seen a plethora of deck inserts, closed-deck conversions, block guards, and the like on the internet, but most are lacking fundamental material science behind them. “It’s great if someone gets something like that to work but you’ll be destroying pistons before you meet cylinder-pressure levels that cannot be clamped by conventional means,” he asserts.
The Whole Package
The RP968 has won the WTAC overall two years in a row, owns the time attack track record at Sydney Motorsports Park, and is .135 second away from the outright track record previously set in an A1GP Ferrari. This has all been done in a vehicle that started out as Rod Pobestek’s road car. The team — consisting of Pobestek (owner), Elmer Racing as powerplant engineer, Dynamic Aero Solutions as the aerodynamicist, and PR Technology as the constructor — complement one another in equal parts of time, money, and expertise.
In the context of tire technology, mechanical chassis grip, and aerodynamic grip, the powerplant may as well have unlimited horsepower. Any more downforce might delaminate the tires; any more power might blow the tires off. Elmgren says the team saw an 860 kW (1,153 horsepower) peak value on the strain gauge data from the rear axles during Barton Mawer’s 1:19.142 lap.
The Bottom Line
If you need a 3,000-horsepower capable drag-racing four-cylinder engine, or a 1,500-horsepower circuit-racing engine, Elmer Racing has you covered. This (almost) entirely billet engine will set you back around $77,000 for the long-block assembly or $133,000 for the complete engine package — provided you can pass the vetting process, as they won’t sell this to just anyone. Remember, in racing, you pay by inverse weight, so it’s a good thing this 1,500-horsepower billet bullet weighs in at a scant 233 pounds.