by Orr Kelly
Compared to these two problems, the third one seemed relatively minor. In some cases, it was found, the pressure and heat in the afterburner were so severe that they caused the afterburner liner to collapse and the engine to lose power. The solution was a stronger liner.
“Warranties aside, GE considered it its responsibility to put the engine in good shape,” Larson says. “At no cost to the customer, we provided improved components and sent maintenance teams out at very, very great expense. We were saying, ‘you can rely on GE.’ ”
The three problems with the engine, serious as they were, turned out to be manageable and far less severe than problems that had plagued earlier jet engines. It now appears that the F-404 engine, in various modified forms, will be flying successfully well into the next century.
And, serious as the problems were, they were confined to the engine. They did not involve the airplane itself. If worst had come to worst, it would have been possible to adapt another engine to keep the F/A-18 flying.
Adm. Wesley L. (“Wes”) McDonald, who was DCNO for air warfare at the time of the crash in England in 1980, says he was at first concerned about the viability of the whole program, but he was relieved to learn the problem was confined to the engine. “The engine,” he says, “is not as critical as something that is weird in the airplane itself.”
As the F/A-18 completed its testing and went into service, the navy and the contractors were to learn just how weird and troublesome problems in the airplane itself could be.
CHAPTER SIX
When Weird Things Happen
Commercial airline pilots are not noted for giving up their places in line so other planes can take off first.
But, at Lambert Field, the St. Louis International Airport, they happily wait while F/A-18 Hornets, fresh off the assembly line, take to the air for the first time, putting on a free air show for the airline pilots and the passengers lucky enough to get a glimpse of the action.
Usually flying in pairs, the test pilots wheel out onto the field, run up their engines, release their brakes, and hurtle down the runway. For a few seconds, they hug the earth to be sure both engines are delivering full power. Then they pull back on their control sticks and zoom almost straight up in what is known as a “Viking Departure.”
Moments later, they level off at 10,000 feet, clearing the airport traffic pattern and permitting the airliners to resume their orderly operations.
As the new Hornets take to the air for the first time, they are dazzling creatures—sleek, powerful, agile—a seeming triumph of the science of aeronautics. But the design of these aircraft owes almost as much to the art of politics as it does to skillful engineering. Therein lie the secrets of the strengths of the Hornet and most of its weaknesses.
As the Hornet has moved through testing and into the fleet, it has encountered a series of problems—with its landing gear, its tail, its wings, its fuselage, its range, and its price—and all can be traced back to the series of compromises, involving both engineering and politics, made during its development.
It is at the sprawling McDonnell Douglas plant bordering the airport that all the components for the Hornet come together to form an airplane. Many of the parts of the plane have been slowly moving through the supply line for months, even years, because it is a forty-four-month process from the time a plane is ordered until it is delivered. McDonnell Douglas waits twenty-two to twenty-seven months for a landing gear, twenty-seven to twenty-nine months for the radar, and twenty-six months for an engine.
The biggest component, shipped to St. Louis in special railway cars, is the rear portion of the plane—from the back of the cockpit to the tail—which is assembled by Northrop in its plant at Hawthorne, California. That comprises roughly forty percent of the plane. It consists of the center and aft portions of the fuselage and the two vertical tails. It contains few of the high-technology parts of the plane that are relatively light and easy to assemble—and which are more profitable.
In its plant at St. Louis, McDonnell Douglas assembles the front end of the plane with the cockpit, the radar, the computers, and the avionics. This is the high-technology end of the plane and it is more profitable. If one is going to build half a plane, this is the half to build. McDonnell Douglas also makes and attaches the landing gear, the horizontal portion of the tail, known as the stabilator, and the landing hook, even though they fit on the Northrop part of the plane.
By the time a Hornet reaches the end of the assembly line in St. Louis, the two halves have been joined; the engines, computers, and other equipment have been installed; radios have been set to the proper frequencies, and everything has been checked to make sure that the test pilot will be able to make his Viking Departure without undue concern.
As the new Hornets scream almost vertically into the sky, their gray paint quickly merging with the atmospheric haze, they seem almost perfect examples of the airplane designer’s art. It is only when one looks closely at the plane, inspecting each of the parts that makes the whole, that one begins to see the results of the long series of aeronautical and political compromises of which this airplane was made.
First, go out to the end of the runway or, better yet, stand with the landing signals officers near the stern of an aircraft carrier, with the wind whipping over the deck at thirty knots, as a Hornet comes in for a landing. The two main landing gear, splayed out at an angle, dangle below the plane like the scrawny legs of some huge bird, groping for the deck. Surely those thin stalks will collapse when the weight of the plane crunches down on them. But as the plane touches the deck in its barely controlled crash landing, the gear folds at its “knees” and seems to crouch beneath the plane, cushioning the shock of its descent.
The gear was designed by the engineers at McDonnell Douglas as part of the process of navalizing the YF-17 that Northrop had built as a lightweight air force fighter. The gear does not look at all the way it would look if they had started with a clean sheet of paper. By the time they began their design work, decisions made by others had vastly complicated their lives.
Normally, the landing gear of a navy plane would be attached farther back than that of an air force plane, to help absorb the crushing impact of a carrier landing and prevent the tail from banging on the deck. But moving the entire gear would have forced a redesign of other parts of the plane. The decision was made to keep the gear where it was but to put in that “knee” that moves the wheels themselves further back.
The second compromise involved making room for two Sparrow missiles to be carried on the “shoulders” of the plane near the points where the gear is attached. Both the missiles and the landing gear were in competition for the same small piece of real estate on the bottom of the plane.
One solution that had a good deal of logic behind it was simply to leave off the Sparrows. The Sparrow was a relatively new missile designed to be guided to its target by radar. It filled the gap between the long-range Phoenix carried by the F-14 and the short-range, heat-seeking Sidewinder. The Sparrow had two advantages. It could hit a target beyond visual range, and it could hit a target head-on. The Sidewinders then in use had to be launched from behind the enemy in order to home in on the heat of his engines. The great disadvantage of the Sparrow was that it simply didn’t work very well. But a group of experts in the Pentagon thought the issue through, concluded that the new plane should carry the Sparrow, and prevailed with an unusual argument.
Robert Thompson, who had gone through the Navy Fighter Weapons School, the Top Gun air combat course at Miramar, as a civilian, made the case for arming the plane with a radar missile: “Whether it was good or not, the Sparrow was a threat. It is like you going after a drunk armed with a gun. Whether or not he can hit you, it alters your tactics. I felt, it alters the other guy’s tactics knowing you have a head-on weapon.”
The designers were told they had to find a way to get the landing gear in and out of its compartment in the bottom of the plane without banging into the Sparrows. It was a challenge
Rube Goldberg would have relished.
When a Hornet is launched from a carrier, it crouches on the deck with its “knees” bent. The heavy main portion of the gear—the equivalent, in a human, of the leg between the knee and the ankle—is parallel to the deck. As soon as the plane is airborne, the gear once again dangles awkwardly below the fuselage before it begins the strange minuet in which the knee bends and rotates, the wheel twists, and the whole assemblage tucks itself up into the bottom of the plane.
As might be expected of a mechanism so complex, things could go wrong—and did.
One of the most tragic incidents occurred on 3 December 1985, at the Miramar Naval Air Station.
Early that morning, Capt. Henry M. Kleemann took off from the Point Mugu Naval Air Station, on the coast north of Los Angeles, for the short flight to Miramar, which is in the northern suburbs of San Diego. Kleemann was one of the navy’s most experienced fighter pilots, with nearly 4,000 hours in the air. In August 1981, he was flying one of two F-14s over the Gulf of Sidra, when they were attacked by two Libyan planes, one of which fired a missile at Kleemann. Within seconds, the two U.S. planes shot down the Libyans with Sidewinder missiles. It was the first time the Tomcat had been involved in combat. Only a short time before the 3 December flight, Kleemann had been given one of the navy’s most prestigious posts as commander of VX-4, a special squadron based at Mugu, and assigned to do the operational testing of new fighter airplanes and weapons.
F/A-18 Main Landing Gear
This shows the complexity of the Hornet’s main landing gear, which has been responsible for a number of accidents.
Although Kleemann had flown thousands of hours in navy fighters, he had less than forty-three hours in the F/A-18, and he had hurried in five days through the four-week training normally required before an experienced pilot is considered qualified in the Hornet.
Piloting a nearly new plane, with only 327 flying hours, Kleemann, accompanied by another pilot in an F/A-18, took off at 8:30 A.M.
When the two pilots reached Miramar, they were held in the landing pattern for a few minutes and then the other plane was cleared to land first. Two minutes later, Kleemann made a relatively hot landing, touching down near the end of the 12,000-foot runway at about 140 knots (160 miles an hour). There had been half an inch of rain during the night. But, despite scattered puddles of water remaining on the concrete, he rolled normally down the runway for about twenty-six seconds. Then, while he was still traveling more than eighty miles an hour, the plane suddenly swerved to the left, then veered sharply to the right, ran off into the soggy grass, turned almost completely around, and flipped upside down.
The crash crew reached the scene within a minute. The canopy was broken off and Kleemann’s head was pushed down into the mud. The rescuers dug frantically with their hands to pull the mud from his face, but water, mixed with fuel and hydraulic fluid, ran into the hole as fast as they dug. They also groped unsuccessfully in the cockpit for the throttle controls to shut down the right engine, which continued to run. With the engine still turning over and the leaking fuel, the wreckage could have broken into flames at any moment. The danger to Kleemann and the rescuers was compounded by the possibility that he had tried unsuccessfully to eject. If the plane was moved, the rocket under the ejection seat might go off, driving the pilot down into the soggy ground and igniting the leaking fuel.
It took almost an hour to secure the ejection seat, raise the plane, remove the pilot and fly him to a nearby shock trauma unit. But Kleemann was dead. Members of the rescue team agonized over the possibility that he might have survived if they had been able to extricate him from the wreckage more quickly. But autopsy surgeons later confirmed that he had died almost immediately after the crash from a severed spinal cord.
What went wrong to cause a highly skilled pilot to lose control of his plane during a routine landing?
Accident investigators pinpointed the trouble in a short piece of metal called a “planing link.” The purpose of the link is to guide the landing gear’s intricate maneuvers during retraction in the few moments after takeoff. The best guess is that, when Kleemann took off at Mugu and retracted his landing gear, the wheel was still spinning rapidly as it moved up into the fuselage. When brakes were applied automatically, the spinning stopped suddenly, with a jerk that bent the planing link.
When Kleemann landed, according to this theory, the bent link pushed the right wheel slightly out of line. And then, as the plane slowed down and its weight settled onto the wheels, the plane veered so sharply that Kleemann was powerless to keep it under control.
Although they were not able to prove it, the accident investigators concluded that Kleemann may have received a warning that his landing gear was not properly aligned but that, because he was not as familiar with the plane as he should have been, he didn’t understand the danger. He may also have been in a hurry to land because he was already late for a meeting—on the subject of safety.
Ironically, Kleemann probably would have survived if he had experienced the same problem while landing on a carrier, even though that is normally more hazardous than putting down on a long runway. The cables on the ship’s deck would have stopped him before the cockeyed wheel had thrown the plane out of control. If he did receive a warning, he should have gone around again and called for a “trap,” in which cables at the end of the runway at Miramar would have been raised to catch the plane, just as though he had landed on a carrier.
The landing gear failure that cost Kleemann his life is one in a series of such problems that has plagued the plane since early in its test program. Not only was there the planing link problem, but there had also been cracks in the heavy section of the landing gear between the “knee” and the “ankle,” near the point where the main shock absorber is attached.
Bent planing links have been blamed for dozens of accidents, most of them much less serious than the one that took Kleemann’s life. A major reason there have been so many accidents is that the delicate link can be damaged on a single takeoff, as seems to have happened in the case of Kleemann’s plane, so there is no opportunity to detect the problem by inspection.
Cracks in the heavy portion of the gear, on the other hand, can be seen on routine inspection, before damage has progressed to the point where an accident occurs. However, the cracks were of deep concern to the designers because of the possibility of a basic flaw in the design of the gear.
As each weakness in the gear has been identified, the mechanism has been modified. On several occasions, where cracks pinpointed areas of weakness, more metal was added. Kleemann’s plane had come off the assembly line only ten months before his crash, and its gear had all the latest modifications. It also had a warning system that, while it wouldn’t detect a bent planing link, should have told him that his right wheel was not properly aligned for a landing.
The navy and McDonnell Douglas are continuing to keep a very careful eye on the plane’s landing gear. Pilots are also being given special training so they will be aware if there is a problem and know what to do about it.
Unfortunately, the stresses that cause cracks tend to “migrate” from the newly strengthened area to the next weaker section. Landing gear problems may still crop up in the future as Hornets build up flight time—and number of landings—in the fleet.
Such problems, expensive and dangerous as they may be, can be understood and fixed. But some of the other weird things that went wrong with the Hornet during its testing, and even after its introduction into the fleet, proved much more difficult to understand and more difficult to fix.
Marine Lt. Col. Peter B. (“Pete”) Field, military program manager for the test program at Patuxent, found one of those difficult problems in the worst possible way. It was a day on which he earned his pay as a test pilot, and then some.
The procedure for testing a new airplane is carefully designed to explore first the areas in which the plane is not expected to encounter any serious problems. Gradually, the test pilots
move cautiously further and further out toward the corners of what they call the “flight envelope,” into the areas where there is more uncertainty and greater danger. The first flights are flown below the speed of sound at relatively high altitudes, so there will be room to recover or eject if something goes wrong. As the test program progresses, the pilots go very fast, very slow, very high, and very low, and they experiment with maneuvers that put more and more stress on the plane.
This prudent practice of tip-toeing up to the toughest maneuvers is one reason the biggest surprises tend to come late in a test program.
The F/A-18 had been put through many of its paces and was performing well on the day that Field set out to see how the plane reacted as he went through a tricky high-speed, low-altitude maneuver. Testing at the potentially dangerous low altitude was essential because the plane’s performance might be different in the dense air close to the ground than it was at a safer high altitude. The test was especially crucial because it duplicated a tactic used routinely in bombing attacks, in which the pilot comes in very low and fast and zooms up sharply. Then he rolls upside down, pulls the nose down toward the target, quickly rolls right side up again, drops his bombs, and heads for home.
To one who has never executed such a maneuver, it would seem much simpler for the pilot to pop up and then push the nose back down toward the target without rolling upside down and back up again. There are two reasons for the quick roll. One is that the plane can stand about five times as much pressure exerted toward the bottom of the plane as it can in the other direction. This means the pilot can point his nose at the target much more quickly with the rolling maneuver. The other reason is that popping up and back down subjects the pilot to negative gravity and that is very uncomfortable. Blood rushes to his head, and his lunch heads in the same direction.
Field followed the standard bombing approach. He popped up, rolled upside down, and pointed the nose at the ground. Then he discovered what the engineers, in their dry way, labeled the “roll rate problem.” The Hornet was rolling back right side up more slowly than it should have—and Field was headed for the ground, less than 400 feet below him. Summoning all his skill, he managed to pull out before hitting the ground and, still shaken by his near miss, hurried back to tell the engineers about their problem.