Space shuttle's main engine design comes up for improvements. Since the Challenger accident parts have come under review

While technicians and engineers in Utah pore over a recently tested booster for the space shuttle, similar teams in Mississippi are testing changes in the shuttle's main engines. Unlike the shuttle's strap-on boosters, which are packed with a rubbery solid fuel, the main engines burn frigid liquid hydrogen and liquid oxygen. The three main engines at the rear of each orbiter account for slightly more than half of the thrust needed to place the shuttle in orbit.

Unlike the redesign effort for the solid-rocket boosters, which is aimed at clearing up major problems with the original design, the liquid-fuel engine program is designed to make incremental improvements.

``The solids had a real failure that was indictable,'' says Joe Lombardo, manager of the shuttle main engine project office at the Marshall Spaceflight Center in Huntsville, Ala. ``The task with the liquid-fueled engines is to improve operating margins through design modifications.''

Though the main engines weren't responsible for the Challenger accident last year, they have given flight officials some frustrating moments. Twice, computers shut down the main engines during the few seconds between the time they ignite and the solid-fuel boosters are lit. The problem was traced to balky valves that controlled the flow of fuel. In another case, a faulty sensor shut down one of the three main engines while a shuttle was in flight. And all along, engineers have been concerned about cracks that develop in turbine blades used in the engines' high-pressure fuel pumps.

Currently undergoing acceptance testing, the engines to be used on next June's flight incorporate several changes. These include:

Changes in the computer that monitors and controls various engine functions.

New wiring for components that control fuel valves. This has apparently helped reduce the sticking valve problem that led to the two aborted liftoffs.

New temperature sensors for the engine's turbopumps, which ignite the fuel and then compress the resulting hot gases for further burning in the engine's combustion chamber.

Modifications on the combustion chamber.

Modified blades for the compressors in the turbopumps. These ``by far have received the majority of our attention,'' Mr. Lombardo says. At best, the cracking problems seriously restrict the turbine's lifetime. At worst, he says, a broken blade rattling around inside the turbine could lead to a catastrophic engine failure.

The need for many of these changes had been seen, and some made, before the Challenger explosion in January 1986, Lombardo says. But the hiatus since Challenger has enabled the changes to be made much faster than they might have been.

Meanwhile, technicians in Utah have removed and inspected the several segments of the solid-fuel booster they tested on Aug. 30. ``So far, the joints look just like we'd hoped,'' says John Thomas, manager of the National Aeronautics and Space Administration's solid rocket booster redesign effort. ``Hot gases never reached the O-rings. The insulation sealed [the joints] in every case.''

That assessment not only includes the joints between motor segments (known as case-to-case), but between the motor and its nozzle, which provides the escape route for the hot gases generated as the solid-fuel burns. While the failure of a joint between segments destroyed the Challenger, seals in the joints between the nozzle and the last booster segment have shown evidence of erosion on 28 percent of motors used, a higher figure than for the case-to-case joint that led to the Challenger explosion.

Indeed, the new nozzle joints have resisted efforts to introduce defects in them. Twice before, case-to-nozzle joints ``had our best shot at intentionally introducing flaws, but they healed themselves'' during smaller-scale tests, says Mr. Thomas. ``Now we're taking a different approach. ... We're going to use unnatural flaws to make sure the gas goes where we want it'' to test the joints during the next full-scale development motor firing later this year. He defines ``unnatural'' as flaws that would never make it past an inspection.

Although the new case-to-nozzle joint came through unscathed, H. Guyford Stever, chairman of the National Research Council's solid-rocket booster redesign panel, says, ``I'd feel happier if they would try alternatives.'' He says that if further testing spotlights shortcomings in the new design, having an alternative ready would reduce any additional delays in the booster program.

``We should be doing more,'' agrees Allan McDonald, who heads the solid-rocket booster redesign effort at Morton-Thiokol. But Mr. McDonald adds that the company is looking at alternatives. One promising design, he says, involves the use of steel seals instead of rubber ones. The steel seals are designed to expand inside the joint as the hot exhaust gases press against them.