The Reagan administration's office in charge of research for the "Star Wars" project has concluded that the dream of exotic space-based battle stations firing powerful laser beams to knock out Soviet missiles and warheads is beyond America's technological reach for the foreseeable future.

The Pentagon's Strategic Defense Initiative Organization (SDIO), formal name for the office conducting Star Wars research, has all but excluded use of lasers, X-rays and particle-beam weaponry, at least for the "first generation" of any space-based missile defense system the administration may decide to build. Instead, it intends to rely on kinetic energy weapons like "smart rocks" -- projectiles that home in on the heat produced by warheads -- and "railguns" -- devices that use electromagnets to accelerate a projectile along a rail to thousands of miles an hour, according to scientists directing the SDI research effort.

The implications of the SDIO's thinking -- that two separate generations of a space-based missile defense system may be necessary -- are likely to fuel the arguments of critics who have questioned both the huge cost and the technological possibility of meeting the goals originally enunciated by President Reagan, namely to "render nuclear weapons obsolete" and provide a shield for U.S. population centers.

The first generation of a Star Wars system "doesn't necessarily mean big battle stations in space," said Dr. Louis Marquet, director of SDIO directed-energy research. "It may be something more prosaic than that."

Exotic technologies like "beam weaponry" -- the futuristic systems that gave the president's strategic defense plan its Star Wars nickname -- would have to await a "second-generation" defense system, assuming the efforts eventually prove workable, Marquet indicated.

Interviews with five top SDIO officials suggest that pressures of time and political necessity are pushing proponents of the space-based defense system to resort to the most immediately available "off-the-shelf" technology. This is a tactic first proposed by High Frontier, a private group that was one of the early advocates of space defense against Soviet missiles.

The "first generation" of the space-based defense system, according to SDIO officials, would be an extension basically of the technology used in a successful experiment last June in which the Army fired a ground-based maneuverable, non-nuclear missile at an incoming dummy warhead and succeeded in hitting it more than 100 miles into space.

Lt. Gen. James A. Abrahamson, director of the program, indicated in a recent interview that several breakthroughs had been achieved in research on a space-based Kinetic Kill Vehicle, known as a "railgun," which would shoot high-velocity projectiles to destroy Soviet missiles in the first phases of their liftoff from Earth, or to kill warheads later in mid-course flight.

Despite the many unknowns about the practicality and cost of the Star Wars project, Abrahamson expressed confidence that research would be far enough advanced by the early 1990s for the next administration to make an initial decision at least on whether to go ahead with the development phase of a first-generation system.

But he also said that one of the three key factors in making that judgment would have to be what he called the "political dimension." This, he said, is where "people look at the world situation as a whole and see that, yes, this is the first of a series of steps that we're all confident now will indeed lead to greater stability."

The other two, he said, were the technical feasibility and cost of the project. But he made no mention of the issue of "survivability" -- that is, whether the components of a space-based defense would be vulnerable to attack themselves -- which has been central to the arguments of many critics of the scheme.

The growing conviction among top SDIO officials that the use of lasers will remain beyond U.S. technological reach for many years was most clearly reflected in the comments of Marquet, SDIO's directed-energy beam expert.

"I don't support directed energy weapons for interception of boosters," he said, referring to the initial boost-phase of a rocket. "We've got a simpler way to do it with kinetic-energy weapons.

"Right now, if you threw a 10 kilometer-per-second smart rock at a booster you are going to kill it. I don't believe directed energy is a competitor," he added.

Kinetic energy is the energy released when a moving object crashes into another. Directed-energy weapons are those that employ a tightly focused and precisely directed beam of intense energy in the form of light or atomic particles.

James Ionson, director of the SDIO's innovative science and technology office, concurred with Marquet, saying "smart rocks" were both possible and far cheaper than laser technology, though he cautioned that it is still not clear whether they could be used effectively against a rocket in the takeoff, or boost, phase.

This, he said, was because it is not yet known whether such high-velocity projectiles could survive being shot downward through the atmosphere without burning up.

"My guess is it's technically possible but how good it will work I don't know yet," he said. "We'll know in a couple of years."

The focus of hopes in the SDIO office for coming up with any workable "first-generation" system is a combination of the electromagnetic railgun firing some version of the "smart rock."

Abrahamson, in illustrating the Star Wars plan to a reporter, turned on a slide projection of the railgun. It is a 150-foot-long, cylindrical object consisting mainly of electrical coils used to generate an electromagnetic force. The force is capable of shooting a small projectile into space at the speed of 20 to 30 kilometers per second.

The Air Force officer said that 10 prototypes of the weapon have been built for experiments in the United States and that a "world breakthrough" in railgun technology was achieved in November when one of them fired multiple shots of tiny projectiles.

The main technological problem, he said, was building switches capable of handling huge jolts of electrical current needed to fire the weapon. Right now, he said, switches on experimental railguns were operating at 2 million amperes but, he added, "We have to able to demonstrate that we an get just about double" that.

"Some of our betters have to understand that's the issue," Abrahamson remarked. "It's not can you put that thing in space but the issue is can you fire this thing at these current levels of electrical power generation . So we'll have to explain that to people and that's complex business."

Abrahamson makes clear, however, that generating enough electricity to fire such a space-based gun is not the only problem to be resolved.

The Homing Overlay Experiment of last June fired an interceptor that weighed roughly 2,500 pounds. But the projectile that Abrahamson says the SDIO has in mind "to make it practical" would have to weigh seven to eight pounds. The general has a model of such a "smart rock" on his desk to show visitors.

"We will have to have been able to go far enough with the concept to test it in the laboratory," he said. "You might even take it out and shoot it at some different kind of targets . . . . But you wouldn't have to put the whole gun up in space and try to shoot down an ICBM," he added.

The Pentagon is planning 15 major experiments of various possible components of a space-based missile defense system. These experiments are being designed, the Pentagon says, so they will not violate the 1972 Anti-Ballistic Missile Treaty with the Soviet Union, which forbids tests of any components of an antimissile system. Critics of the administration's plans say that these plans violate the treaty, and that the Soviet Union will certainly consider them violations.

Another unresolved issue is the cost-effectiveness of a defensive system. Marquet said researchers must figure out how much it costs to "kill a Soviet booster."

He estimates the United States would need about 20 defensive rockets for every one of the 1,400 Soviet long-range, land-based offensive missiles, or roughly 28,000 "smart rocks." This advantage is needed largely because of what is known as the "absenteeism" of the satellites, or railguns, shooting the projectiles. Since they would be constantly rotating around the Earth, each one would pass over the targeted Soviet silo for a few minutes on each orbit.

High Frontier has argued that 432 satellites would be required to keep the Soviet missiles constantly under surveillance.

Marquet, unlike Abrahamson, talks less of railguns firing the "smart rocks" than cheap, light orbiting space "platforms," which he described as "a little rocket launcher with a telephone on it." The "telephone" would be a sensor and processor of information to guide the smart rock.

The cost ratio of a U.S. defensive smart rock to a Soviet offensive missile, he said, would have to be overwhelmingly in favor of the defense -- "not 1 to 1, or 5 to 1, but 100 to 1" -- so the Soviets would be convinced it was not worth a proliferation of new ICBMs to counter them.

In addition, the cost of putting the platforms into space, he said, would have to be reduced dramatically over present shuttle costs. Three-quarters of the expense of the space-based defense system will be from putting it into space. "It would take 50 years to put them up with the shuttle now," he remarked.

Col. George Hess, chief of research into the survivability of key technologies, concurred. He said that the cost-per-pound of payloads taken up by space shuttles today is $1,400 to $1,500 and that to make the space-based system affordable it would have to be brought down to $150 to $200.

"I would feel pleased if we get it down to $300 to $400 by the mid-1990s with the target of $150 to $200 in the post-year 2000 time frame," he said.