The drawings and one-of-a-kind prototypes one sees of alternative "engines" like fuel cells, gas turbines, flywheels and fancy electric motors usually are pretty and impressive. That these alternative engines exist almost exclusively as drawings and priceless prototypes symbolizes that they are for now largely theoretical exercises in an industry that makes things. By the millions.

The only powerplant the auto industry has experience (and investment in) making by the millions is the internal-combustion engine (ICE). Given that, bet you'll see ICEs for some time to come.

For the newborn, turn-of-the-century auto industry, the ICE - both gasoline and diesel - proved to be reliable, durable, and, after accumulated decades of manufacturing, pretty cheap to produce. In simpler times basic ICEs were just fine.

But when a measly 15% of the energy in a gallon of gasoline finally gets to the road, that's not good enough anymore. The Partnership for a New Generation of Vehicles (PNGV) fingers the largest energy "waster" in every vehicle as the ICE (see chart p.89). And the developers of alternative power sources claim efficiencies that outstrip current ICEs. Some by a healthy margin. This troubles industry executives with massive engine-manufacturing investments centered exclusively on the production of ICEs. So these executives tell their engineers to find ways to improve the ICE.

It's happening. And has been for as long as there have been ICEs.

Electronic engine management systems have opened up new vistas for improving theICE. Most things that formerly were mechanically (and somewhat clumsily) controlled now are governed by intelligent, adaptable, high-speed electronics.

Just how much the ICE's basic thermal efficiency can be elevated, however, is a matter of contention. One Big Three engineer says "fully optimized" spark-ignited gasoline engines might reach 35% efficiency, diesels perhaps as much as 45%.

But how will ICEs best be "fully optimized?" There's an extensive spectrum of potential strategies, several already in production, others still on the development bench. Based on discussion with OEM, supplier and independent engineers, developers and researchers, here are the best ideas for the "ideal" future ICE - both gasoline and diesel:

BASIC DESIGN

n Spark ignition or compression ignition. Take your choice. Diesels offer potential for the greatest overall efficiency. Gasoline engines present a better NVH profile and their emissions are more easily catalyzed, but they're more complicated.

Our experts say go out on a limb - and consider future hybrid-electric applications - in choosing diesel.

n Ideas about the "correct" number of valves vary. General consensus finds four valves per cylinder the optimum arrangement to achieve the best intake and exhaust efficiency. Dual camshafts in the head - or heads, for a vee engine - should drive the intake and exhaust valves.

Mercedes-Benz engineers make a strong argument for a 3-valve design (two intake, one exhaust) and a single overhead camshaft (SOHC). Among other advantages, employing a single exhaust valve transfers more combustion heat to the exhaust; the catalyst then quickly reaches ideal emissions-scrubbing temperature.

Audi AG believes more is better when it comes to valves, and is moving to 5-valve-per-cylinder designs. General Motors Corp. thinks two-valves-per-cylinder and overhead valves work fine.

n Will we see electro-magnetic valve actuation that does away with camshafts, valve springs and the like?

We're told Ford Motor Co. has a system ready for diesels, and GM's been working on it for years. The electro-magnetic valve systems are reputed to be energy hungry, and GM engineers have in the past said that gasoline-engine applications pose problems.

For now, let's accept an I-4 or V-6 (number of cylinders dependent on vehicle size) conventionally operated DOHC 4-valve-per-cylinder arrangement for the future engine's basic architecture.

Aspiration

There's hardly a consensus here. Some engineers believe future ICEs will virtually necessitate some type of forced induction because the drive will be to reduce engine displacement - and the only way to ensure decent specific output and torque will be with turbo- or supercharging. Others think newer electro-mechanical enhancements make forced induction redundant.

Per Gillbrand, Saab Automobile AB's recently retired chief engine and powertrain engineer - and devout proponent of turbocharging - told Ward's in a 1995 interview that to meet future requirements for high torque, low emissions and fuel economy, turbocharging is the answer: "If you want to meet future demands in a normally aspirated engine, you have to use a lot of rather complicated mechanics. The turbo (can achieve those goals) without all these things."

Add turbos. Especially in diesels.

VALVE ACTUATION

Variable valve-timing is too good a thing to pass up, and we've seen variable valve timing popping up throughout the industry. It optimizes valve overlap, or lack thereof, throughout the engine rpm range.

That's important to emissions reduction and enhanced fuel economy. Probably as important to the customer, variable valve-timing helps engines to be better low-rpm "torquers" without sacrificing high-speed power production.

Hidetaka Nohira, Toyota Motor Corp. senior staff engineer, engine engineering division, solidly believes every engine of the future will incorporate variable valve timing designs. "Variable valve timing will become a major technology in the near future," he asserts.

Toyota's example is just one of many. Everybody raves about their variable valve timing setups, so we'll assume future ICEs must employ it.

FUELING

Two words: direct injection.

DI isn't ready to become a household term, at least for gasoline engines.

n For diesels, though DI currently is de rigeur. You'll hardly find a European passenger-car diesel that's not employing in-cylinder injection. Heavy-duty diesels have had DI systems for years.

Direct injection allows extremely precise, high-pressure injection of fuel directly at the combustion-chamber point of attack. Sophisticated air/fuel homogenization, rate-shaping of the fuel charge and two-stage "pilot" injection all come to the party, greatly enhancing diesel efficiency while simultaneously reducing certain emissions and calming the historic diesel "clatter."

Now add "common-rail" fuel delivery and your advanced DI diesel becomes a high-rpm torque juggernaut because injection pressures are consistent regardless of engine speed. These consistent pressures mean, in addition to enhanced injection, less emissions and noise.

Combine DI and common-rail fuel delivery and you've got the makings of serious efficiency. Mercedes, BMW AG and Audi all recently unveiled V-8 turbocharged diesels incorporating these features - and once they're introduced for production, luxury car drivers may never look back. In the U.S., the Big Three are engaged in serious programs to develop common-rail DI diesels for light trucks.

n Direct-injection for gasoline engines is in its infancy but has a following.

Currently, Mitsubishi Motors Corp. and Toyota have gasoline DI engines in production. Toyota is moving cautiously, but Mitsubishi says that by 2010 all passenger-vehicle gasoline engines will be DI.

Audi says its prototype 74-hp DI 3-cyl. gasoline engine, when used in its featherweight AL2 concept car, uses 40% less fuel than a similar conventional engine.

Robert Mull, Ford's director of PNGV activities, supports Audi's figures. "Our CIDI (compression ignition direct injection) engine will get to 40% (thermal efficiency)."

n DI will have to overcome emissions concerns. Because DI makes extensive use of extremely lean air/fuel ratios, excess NOx production is a problem and today's 3-way oxidizing catalysts aren't the answer. Troubling, too, are DI gasoline engines' sensitivity to high-sulfur fuel.

Expect all problems to be overcome. DI is here to stay.

THE SUM-UP

Want a cost-is-no-object ICE? Here are the specs for a midsize passenger-car application:

n 2.3L diesel V-6.

n DOHC 4 valves per cylinder.

n Common-rail fueling.

n Direct (in-cylinder) fuel injection.

n Variable valve timing.

n Turbocharged.

n Aluminum block/aluminum heads.

Don't agree? Give your thoughts (as succinctly as possible) at www.wardsauto.com. An update will follow.