The rotary community as a whole is plagued by a problem with engine reliability issues, namely the lack thereof. Generally speaking, this does not apply to non-turbo applications. It’s mainly a turbo thing. I believe the blame is a combination of things — one, a false sense of security that we’ve sold ourselves into thinking we can run lots of boost and make lots of power out of this dinky motor even though we’ve not given the proper attention needed into engineering the system as a whole, looking at many parts and finding reliable, solid ways to “balance” this all out.

With every piece of my being, I am utterly sold on the notion that these engines fail due to pre-ignition during or at the tail-end of the compression stroke because the fuel that’s mixed with the then-compressing charge, as it builds up heat (pressure is proportional to temperature), reaches a point of temperature and autoignites, firing the combustion event early, before a timed spark from the spark plug. It’s not detonation; it’s pre-ignition due to spontaneous, auto-ignition of the fuel. It makes perfect sense. This is why the old style turbo rotors (8.5:1 compression ratio) can “take more boost”. It’s why it’s generally dangerous to run a turbocharger on a higher compression engine, say with 9.4:1 or 9.7:1 NA rotors. The higher the compression ratio, the higher the effective pressure generated during the compression stroke, therefore the higher temperatures. At some point, barring any heat-dispersion problem related to “too hot” of a spark plug, the fuel just blows up on its own because of the surrounding heat in the combustion chamber — the temperature of the two side plates, the rotor housing, the temperature of the oil that’s filled the rotors to pull heat out at the faces, as well as the temperature of the intake charge at the beginning of that compression stroke. Auto-ignition seems to be the devil’s finger here. Aside from any additive benefits such as lead, race fuels that are within the 100-117 octane rating have substantially higher auto-ignition temperatures over pump fuels. Pump fuels, based on what I’ve read and while I’m not sure if it’s specific to a particular octane, have an average auto-ignition temperature of 491*F which is approximately 9* cooler than the average temperature of the rotors under load based on one SAE paper I’ve read. Peculiar.

It’s *all* in the fuel!

hange that one key factor, and magic happens. Carl Byck’s road race car, tuned with a 5* split and a 17* leading advance, 25psi of boost, 506rwhp on a series 5 turbo 4-port block. Runs 20lbs of boost for 20 minutes at a time on a road race course with oil and water temperatures through the roof. Does the motor let go? Nope. How does it do it? 111 octane leaded race fuel. Unwaveringly reliable compared to setups 250hp under running nowhere near the load.

We modify these cars the wrong way. We’re looking at it wrongly when we should be looking at it like a complex array of simple systems and parts, all balanced together to produce a certain output. If one is out of whack, then the whole thing falls apart at some point. We’re putting humongous, front-mounted intercoolers in our radiator dams, effectively blocking the air needing to get to the oil cooler and radiator, therefore raising base engine temperatures with respect to the water in the jacketting as well as the oil running throughout the system. Although the air temps are lower, the engine internals are much hotter, therefore challenging the knock-resistance of the fuel mixed in the charge.

The density of the O2 molecules per volume that’s entering the compression stroke and being fully compressed does not change the temperature of combustion — only temperature based factors do. Running more boost in as much as the density of that charge has doubled doesn’t yield higher temperatures; the higher temperatures come from the challenges incurred when pressurizing air at the turbo’s compressor in the first place — temperature is proportional to pressure, as I’ve said before. The typical compressor can belt out air temperatures in the 300* – 400*F range. Intercooling helps but potentially comes at the cost of starving the radiator and any oil-cooling heat exchangers for high-pressure, ambient-temperature “ram” air, thereby dramatically reducing their efficiency.

Air-to-water intercooler, built in late 1998 and tested throughout 2001… the underdog, Unorthodox setup that “wasn’t supposed to work” according to the then well-known “experts” belts out over 420hp on several runs, back to back, on the dyno along with 320ft/lbs of torque. 14-15psi of boost, standard-shaft Turbonetics 60-1 HIFI compressor w/ an HKS cast manifold (undivided) and undivided P-Trim 0.96 A/R turbine housing. Running 17* of leading advance with a 7-8* split and it never knocked one time. Manifolds still cool to the touch after a pass on the strip or dyno. Intercooler core frozen over, producing condensation, water temperatures in the 40-50*F range. Bouncing off the 8100rpm rev limiter, highly reliable setup, EGT’s in the high 1200’s to approximately 1300* range, EGT probe post-downpipe. Oh, and one more thing.. A2W core converted from stock core, complete with 1 7/8″ intake piping. Impossible, huh?

And recorded tonight…
10″ vacuum, 3300rpm, 80mph, EGT’s 1300*F while cruising at light load.
15.5psi of boost, 6500rpm, ~95mph, EGT’s 1300*F while under heavy load with a mixture of 80% gasoline and 20% methanol, sustaining a re-calculated AFR of approximately 10.40:1.

Dealing with a problem of having too volatile of a fuel in the combustion chamber by throwing more of that same volatile fuel at it is ridiculous and oxymoronic. We don’t stop a murderer by overwhelmingly-smothering him with targets to shoot. There’s no reason for us to run such rich mixtures in the high 10’s to mid 11’s:1 with fuel injection, thereby forcing us to use higher capacity fuel injectors and more capable fuel pumps. We have to get the intake air temperatures down, we have to focus on oil temperatures and water temperatures in the motor (the core, critical stuff; oil cools 1/3 of this motor!), and we must use a quality fuel that has a relatively high auto-ignition temperature so as to prevent pre-ignition knock.

Water injection, for the purpose of raising the anti-knock index of a fuel, is worthless. It also must be atomized to a very high pressure (in an already pressurized system) to even have its specific heat and latent heat of evaporation properties be of any use at all. It is inert in the combustion chamber, providing 0 BTU’s as it does not combust. It can also hydro-lock a motor if the solenoid that’s keeping the high-pressurized system from hitting the nozzle fails. The only benefit of water in all practical application is its ability to pull some heat out of the intake charge, and even then its arguable how much it can based upon the amount of water entering the chamber thereafter.

Alcohol (ethyl and methyl alcohol specifically), on the other hand, does not require atomization via a pre-pressurized system as its flash point is approximately 54*, making it turn to a vapour instantly in the intake system even at very low amounts and also removing “staging” problems. It provides BTU’s (about less than half of gasoline, lb per lb), burns “cold”, dramatically lowers intake air temperatures, can literally act as an intercooler on its own (and does on alcohol-fed race setups), has an extremely high latent heat of evapouration as well as an extremely high auto-ignition temperature, giving an effective octane rating as high as 140.

EGT’s are the key. Air/Fuel ratios are temporary; EGT’s are what need to be looked at like a hawk when tuning methanol and pump gas under load.

Gasoline — Auto-ignition temperature of sub 500*F.
Various race gasolines — auto-ignition temperature ~660*F.
Methanol — Auto-ignition temperature 878*F!!!!! The motherload of all internal combustion fuels!