Undersecretary of the Army James R. Ambrose has begun an investigation of Ford Aerospace & Communications Corp., the prime contractor for the Army's controversial $4.5 billion Divad antiaircraft gun.

The investigation centers on reports that the company has hired at least four retired military officers who had roles in the gun's development, and that four other officers have been hired by Ford Aerospace's Divad subcontractors, including the main one, Westinghouse Electric Corp.

Divad, for Division Air Defense, an elaborate weapons system that relies on radar to detect enemy aircraft and helicopters in the dust of battle and to aim its rapid-firing guns, has had a succession of serious problems since the Army decided to build it in the mid-1970s. Last year Defense Secretary Caspar W. Weinberger set aside the scheduled purchase of the next 117 of the units -- 146 already have been bought -- pending further tests this May.

Ford Aerospace won the contract after a disputed shoot-off in 1980 with a General Dynamics Corp. design.

An investigation by The Washington Post has found that before the 1980 competition between the companies' prototypes, Ford and two of its subcontractors had hired three former Army lieutenant generals, all of whom helped in preparing its proposal. Before they retired from the Army, all three had been involved in the Divad research program, and one helped formulate the Army's requirements for the system while he ran the Air Defense Command.

Another five lower-ranking Army officers who were involved in running the 1980 shoot-off were later hired by Ford and Westinghouse. One, the Army test manager at the trial, was hired to be Ford's chief test manager three months after the contract award.

"One has to be sensitive to perceptions," Ambrose said in an interview. "Whether there is any impropriety or not, there is always the perception there might be one, or the perception the actions of the individuals might have been influenced by his hopes to get a job."

Ambrose, who worked for Ford Aerospace 25 years before he came to the government, said he initiated the inquiry because such hirings "put a cloud on government operations," which are carried on "in a goldfish bowl. We have to accept that operating in a goldfish bowl is different."

The Divad hirings are part of a familiar problem for the military and military contractors, the so-called "revolving door."

It is not a simple issue. Even critics acknowledge that former military and civilian defense employes have important expertise for companies seeking to do defense business. The question is whether the companies, in hiring such people, are buying not only expertise but influence as well.

There is no evidence of any such exercise of undue influence in the Divad case. But it is nevertheless a major example of an intensifying problem for the Defense Department.

Defense research and procurement expenditures have risen sharply the past five years. So has pressure in the military services for many officers to retire -- and the revolving door today is spinning faster than at almost any time in the past.

The latest reports on file at the Pentagon show about 2,200 retired officers working for companies with $10 million or more apiece in Pentagon business.

Federal conflict-of-interest law does not bar such work, nor even work on the same systems retired officers handled while in uniform. Retirees are barred only from working in sales and marketing for companies on programs for which they had contract authority while in government service.

Some members of Congress want to further restrict retirees' activities. A bill to do so was introduced last year in both houses of Congress.

The Pentagon inspector general's office also is said to be reviewing the issue, although David W. Ream, who handles revolving-door questions for the Defense Department general counsel's office, said in an interview that "it is largely a myth that the retirees have influence over people they worked with" or that while in service "they roll over in hope of getting future employment" with contractors.

The General Accounting Office, at the request of the Senate Government Operations Committee, is now investigating the Divad program, including how the contract came to be awarded to Ford Aerospace in 1981.

Army officials as well as Ford and Westinghouse executives say all the Divad officers were hired because of their expertise and not because of any influence they could exert.

Lt. Gen. Louis C. Wagner Jr., the Army's deputy chief of staff for research, development and acquisition, who supervises the Divad program, said in an interview that although he was not aware of all the Divad revolving-door hirings, it was "innuendo rather than fact" to imply that anything was wrong with the employment of the retired officers.

Louis F. Heilig, vice president of Ford Aerospace for the defense group and supervisor of the Divad program, said his company hired the former Army officers for the knowledge they brought to the job and not for their influence.

Retired general John R. Guthrie, now head of the Association of the U.S. Army, who three years ago selected the Ford Divad model over the General Dynamics alternative, said in an interview that no influence was brought to bear by the former officers. The current Army managers of the Divad program, such as Wagner, say the same.

One officer hired by Ford was Lt. Gen. Howard Cooksey, who retired in 1977 after serving as assistant chief of staff for research and development, the same job Wagner now holds. Cooksey said in an interview that he set himself up as an independent consultant and approached only companies with which he had not been directly involved in a contract decison.

He said he approached Ford Aerospace because, although he handled the money for the Divad development and "overall policy" while at the Pentagon, the choice among the competing companies was handled by Guthrie and the Army Materiel Command.

Cooksey said he retired at the end of 1977 and was hired by Ford "at the beginning of 1978." He said he helped Ford on "choices of ways to do things" in its Divad proposal and assisted on "programmatic" approaches. But he said he "did no buying, no advertising," and avoided acting as a social intermediary between Ford executives and his Army colleagues.

Another retired officer was Lt. Gen. C.J. LeVan, commander of the Air Defense Center at Fort Bliss, Tex., in the early 1970s and one of the leading advocates of the Divad gun program in the Army. He also was the author of some of the system's requirements.

Cooksey described LeVan as "the guy at Fort Bliss who was very strong" for Divad during his Army days.

LeVan retired in 1980 and went to work for R&D Associates, which had been hired by Ford to help prepare its proposal for the Divad shoot-off with General Dynamics.

LeVan said he would not comment on the work he did for Ford on the Divad contract. Ford's Heilig said, "LeVan was known as the highest authority on air defense at Fort Bliss" and is "a very sophisticated analyst." He was employed as a consultant, Heilig said, so Ford "could pick his brains on how the Divad would be used . . . and how to do the test equipment."

Another retired officer, Lt. Gen. Robert Baer, joined XMCO, a McLean, Va., consulting firm in August 1980. By that time, XMCO had already been made part of the Ford Divad team, hired to do work on the logistical support Divad might require.

Baer, who had been deputy director of the Army Materiel Development & Readiness Command for research and development and a deputy to Guthrie, said, "I had no role to play" in the Divad contract, because "I recognized the sensitivity."

When the shoot-off took place, five military officers were closely associated with running the tests:

* Col. James C. Crosby was president of the U.S. Army Air Defense Board and manager of the test. In July 1981, two months after Ford was awarded the Divad contract, Crosby was hired by Ford to run its follow-up testing.

* Col. Gary C. Mahan was Divad system manager for the Army training command, developing the tactical use, logistics and personnel requirements for the system. One month after Crosby went to Ford, Mahan joined Westinghouse to assist that company's Divad manager in testing and production of the system's fire control system.

* Lt. Col. Walter M. Burch was Crosby's executive officer and managed the next Divad test. In August 1982, he, too, was hired by Ford, although not to work on the Divad program.

* Lt. Col. Lawrence J. Keller managed distribution of the test system results at the shoot-off and later Divad tests. In July 1983, he went to work for Ford as a test engineer, again not on the Divad program.

* Lt. Col. Clyde L. Humphrey was deputy and then chief of the Army's operational test and evaluation agency at Fort Bliss and analyzed the data of the Divad shoot-off and later tests. In August 1983, he joined Westinghouse as a research engineer, dealing with future air defense systems.

Crosby, who supervised the Fort Bliss range for the shoot-off, was hired by the company to be "responsible for managing the test program for the Sgt. York Divad gun system," according to his statement filed in 1982 with the Pentagon.

Ford's Heilig said Crosby was chosen because of his "depth of knowledge of short-range air defense and operational testing." Before joining Ford he checked his proposed job description with the Army judge advocate general branch and was told there would not be a conflict, Heilig said.

"I understand why people think there might be something funny there," Heilig said, but "the people at Fort Bliss couldn't buy us that test" because it was evaluated elsewhere.

Charles French, a retired Army civilian technician from Fort Bliss, said Crosby was hired because "when Ford got the contract they had to expand quickly in order to fulfill it . . . . I don't think there was anything wrong."

Mahan, who was Divad program manager at the shoot-off for the Army's training command, joined Westinghouse in August 1981 to "assist the Divad department manager in the production and testing of fire-control equipment," according to his Pentagon filing.

General Dynamics also offered Mahan a job, as did several other defense contractors, a Westinghouse spokesman said. He chose Westinghouse because he liked its offer best and ended up working on the Divad contract, the spokesman said, after looking at other areas of company business.

Although Ford won the test, the data disclosed more than a dozen "shortcomings," according to Ford's Heilig.

From November 1981 to January 1982, a "check test" was held to "validate if Ford had corrected the shortcomings," Heilig said.

He said Crosby designed that test. Wagner said the Army designed the test.

In any event, Crosby's former deputy, Burch, his successor as head of the Air Defense Board, managed the check test for the Army. Crosby was on the scene, according to French.

The results of that test are classified. But at a hearing last year before the House Appropriations subcommittee on defense, the Army reported that there still were system deficiencies. The Defense Department inspector general's office has been looking into that test, as is the GAO.

Six months after the check test was completed, Burch retired from the Army and nine days later was employed by Ford in advance systems engineering. He did not work on the Divad program.

Keller and Humphrey continued to work for the Army on the next set of contractor engineering tests of Divad from April to December 1982. On July 18, 1983, Keller, who had retired 18 days earlier, went to work for Ford as a test engineer, according to his filing with the Pentagon.

Slightly more than a month later, on Aug. 29, Humphrey went to work for Westinghouse in the radar systems development department. He did not retire from the Army until Sept. 30, according to his Pentagon report, having started work while on terminal leave.

A Westinghouse spokesman said Humphrey had been recommended to the company by Mahan and chose the company after looking over several others. "He had made up his mind not to go to work on a job with direct involvement with Divad," the spokesman said.

Ambrose represents a different kind of revolving-door situation, the former business executive who has to deal with his former company as a government official.

As the vice president for technical operations of Ford Aerospace, Ambrose had reviewed the company's proposal for getting the Divad contract before his retirement in 1979.

When he came to the Pentagon in July 1981, the contract had been awarded.

"Recognizing this town operates on perceptions," Ambrose said in an interview, "we had discussion at the time on how to deal with this." At the time, he said, "air defense was one of those areas where there were serious problems" and Ford Aerospace was involved in several beside Divad.

Army Secretary John O. Marsh Jr. sent him a memo on Aug. 4, 1981, exempting him from "any requirement that might otherwise appear to exist for you to disqualify yourself from participating in official actions affecting Ford Motor Co. or its subsidiaries by virtue of the identified interests" he had with Ford.

Since he was sworn in as undersecretary in October 1981, Ambrose has been involved in Pentagon deliberations on the Divad project, which has become the largest defense program undertaken by his old company.