CAPE CANAVERAL, Fla. -- Three hours south of the Kennedy Space Center, on the edge of the Florida Everglades, engineers are crafting a new rocket engine that one day might power the next generation of reusable launch vehicles.

Located on Pratt & Whitney's West Palm Beach campus, the spirit of early space pioneers such as Robert Goddard and Wernher von Braun are alive and well. There, rocket scientists are attempting to take a half-century of modern engine experience and figure out a way to make the powerplants more reliable and less expensive.

"This particular engine is much like the space shuttle main engine," said Rick Bachtel, Pratt & Whitney's vice president for booster propulsion systems and the manager in charge of the Co-optimized Booster for Reusable Applications (COBRA) engine effort. "We call it a booster engine, but depending on the application it could go all the way to orbit."

COBRA is competing with other engine designs produced by companies such as Boeing's Rocketdyne unit for the right to power a replacement for NASA's space shuttle that is being developed as part of the agency's Space Launch Initiative (SLI).

The goal is to increase the safety and reliability of operating the engine so that the chance of a catastrophic failure leading to loss of the vehicle and crew is reduced to five in one million, said Jim Snoddy, NASA's COBRA project manager at the Marshall Space Flight Center in Huntsville, Ala. This would be a significant improvement over the current shuttle main engine that has an estimated chance of failure rate of 258 in one million .

And if the safety and reliability can be improved, the overall program costs of operating the engine can be reduced, Bachtel said.

"It's as simple as the fact that if I've got a reliable engine, then I don't need a lot of people working on the engine between flights. I don't need a lot of inspections and I can reduce the standing army, and that brings the cost down," Bachtel said.

The key to meeting that goal is reliance on decades of previous rocket engine experience combined with at least 16 new embedded technologies. They range from new manufacturing techniques and a new powerhead configuration, to a health monitoring system that can sense when bad things are happening and take action before the situation gets worse.

"What's of paramount importance is if the engine has an issue, you want to be able to shut if off without blowing up," Bachtel said. "That then gives you the opportunity to either complete the mission or save the crew with some sort of abort mode."

The COBRA engine is largely based on the space shuttle main engine (SSME) and takes advantage of NASA's recent $1 billion Alternate Turbopump Design (ATD) program. This program was established to build new high pressure fuel and oxidizer turbopumps, which are literally the heart of the shuttle powerplants. The ATD pumps already meet the SLI goals for safety and reliability.

The differences between COBRA and the SSME start with the configuration of the engine's powerhead, which includes the turbopumps and the main combustion chamber. It's in the powerhead that a rocket engine's propellant is mixed together and ignited, resulting in the fire and smoke exhaust from the engine nozzle that leads to liftoff from the launch pad.

On the SSME there are two pre-burners, one attached to the top of each of the high pressure turbopumps. Liquid hydrogen and liquid oxygen are combined and burned in a fuel-rich mixture and the resulting hot gases are expelled out of the pre-burner and over the pump's turbine blades to power the turbopump, which moves the propellant through the engine's plumbing.

On COBRA there is a single pre-burner combining liquid hydrogen and liquid oxygen. Hot gases are expelled from the pre-burner and divided in two, with one stream powering the liquid oxygen turbopump and the other the liquid hydrogen turbopump.

The concern with the SSME's two pre-burners is if there is a problem with one side of the engine, the other side keeps going. Unless the computerized health management system picks up the problem fast and corrects it, the whole system can get out of balance. "If one of the two gets away then the engine can destroy itself pretty quickly," Bachtel said.

The COBRA advantage is that with this "powerball" configuration of pre-burner and turbopumps both turbines always see the same hot gas pressure, so "if something goes wrong they'll both come up or go down together," Bachtel said.

Another configuration change incorporates new manufacturing techniques and improves the way in which the COBRA engine is to start up.

On the SSME the walls of the exhaust nozzle are cooled by flowing liquid hydrogen through hundreds of tiny tubes that line the inside wall. The problem with the tubes, according to Bachtel, is that they add time and expense to engine manufacturing and are easily damaged. The COBRA solution is to mill grooves into one wall of the double-walled nozzle to create coolant passages for the supercold liquid hydrogen. A subscale nozzle with the milled channels already has been made, pressure tested and everything looks good so far, Bachtel said.

In a similar way, the walls of the main combustion chamber will be cooled using new platelet technology from Aerojet. Paper thin sheets of a mostly copper material are etched much like a circuit board and stacked up until they are about a half-inch thick. The plates are joined together in a process called diffusion bonding, which turns the individual sheets into one thick block -- except where the material was etched can now be turned into coolant passageways.

Bachtel said that with such technology the main combustion chamber can be cooled with less pressure drop and with greater heat transfer rates, which in turn will help the overall engine work easier and prolong the life of the combustion chamber -- making it more reliable and less costly to operate.

The same platelet technology also will be used to form the injectors that spray liquid hydrogen and liquid oxygen into the main combustion chamber like a shower head, Bachtel said. The goal with the injectors in any rocket engine is to get as fine a spray as possible so there is maximum mixing before the elements ignite, and the etched passageways available from platelet technology provide that requirement.

Officials say that what will make this engine work is that much of the technology isn't necessarily new, so there is a body of experience to draw from that will help drive up the reliability of the engine sooner than later. On the other hand, a lot of this technology hasn't been used in this particular configuration before.

"I think this is semi-revolutionary," Snoddy said. "We've got a well thought out plan to get us there, but it is not a 'just go do.' "

The COBRA contract calls for construction and initial test of a prototype engine in the 2005-2006 timeframe. Hotfire tests of the powerball powerhead configuration are expected in Florida in about one year. But the whole plan could change depending on which vehicle concepts NASA decided to pursue as the SLI program continues.

The next clue as to COBRA's ultimate fate could come as early as this October, Snoddy said. That's when the agency may decide to concentrate on an engine that uses liquid hydrogen and liquid oxygen for the SLI first stage, or kerosene and liquid oxygen - the propellant Rocketdyne's candidate engine uses.

In any case, a hydrogen-oxygen upper stage engine remains likely, so COBRA officials believe they will be working on this engine for many years to come.