Analysis and video of C-5 Galaxy crash at Dover Air Force Base
The C-5B Galaxy, operating for the 436th AW, took off from Dover AFB (DOV) at 06:21 on its way to Kuwait. About ten minutes after takeoff the aircrew observed a No. 2 engine “Thrust Reverser Not Locked” indication light. They shut down the No. 2 engine as a precaution and decided to return to Dover.
During the return, the pilots and flight engineers continued to use the shut-down No. 2 engine’s throttle while leaving the fully-operational No. 3 engine in idle. Also, the wrong flap setting was used. On final approach to runway 32, the airplane descended below the normal glide path. The C-5 stalled, hit a utility pole and crashed into a field. The tail separated and the C-5 crash-landed further on. The nose section separated due to the impact.
All 17 people on board the C-5 survived the crash, but three crewmembers were seriously injured when the aircraft stalled, hit a utility pole and crashed into a field about a mile short of the runway. The other passengers and crewmembers sustained minor injuries and were treated and released from local hospitals.
The video features a computer-generated crash animation along with cockpit recorder playing.
Details of Galaxy crash at Dover Air Force Base
on April 3, 2006, C-5 Galaxy aircraft took off from Dover AFB, De bound for Ramstein AB, Germany. The aircraft commander, “Golden Boy,” flew from the copilot seat, the assistant operations officer, an instructor pilot, occupied the pilot seat, and Golden Boy’s squadron commander occupied the jump seat.
Taking off in clear, unlimited visibility conditions just after sunrise, the aircraft turned toward the Sea Isle TACAN navigation station on the coast of New Jersey. As they approached that navaid about twenty miles from Dover, the “Thrust Reverser Not Locked” warning light illuminated on their instrument panel. This meant one of the mechanical locks on the reverse indicated unlocked. By the aircraft manual, the engine should be shut down expeditiously but in a normal fashion, not an emergency fashion. The crew shut down the engine and requested a return to Dover from air traffic control.
As they turned back toward Dover, many considerations had to be reviewed. First, the aircraft was very heavy with a full load of cargo and fuel. This meant a partial flap landing should be made to ensure enough power was available to use for both the landing or if a go-around was necessary. This is especially true with only three engines. Power would be limited, so the partial flap setting, which involves less aerodynamic drag, which would require less power, would be demanded.
Second, at this heavyweight, no more than a six-feet-per-second descent rate should be used so the stress of the heavyweight on the landing gear would be reduced on touchdown. This, along with the inertia of a heavy aircraft at high speed, would require more runway length to stop the plane.
So, the approved solution would be a partial flap (40%) configuration landing on a long runway to ensure sufficient stopping distance. While this would be the standard, other considerations came to the fore with this crew.
First, a previous emergency landing at Dover a year before seemed prominent in the crew’s deliberations. In that previous episode, a heavyweight C-5 had landed long, not been expeditious in slowing the aircraft initially, and had to apply maximum braking to avoid going off the end of the shorter Dover Runway (Rwy 01/19). This led to a massive brake fire and crew evacuation. As generals sometimes try to fight the previous war, this crew seemed worried about this previous accident. Most significantly, this crew had been briefed on this incident the day before this flight. This explains their overriding concern about hot brakes on landing, a concern that distracted them from much more imperative considerations they mishandled.
Factor #1: The aircraft commander flying the plane in the right seat broke off from an instrument approach to runway 19 in favor of a visual approach to runway 32, the longer runway available to provide additional braking distance. While it provided more runway, it took away a precision ILS glide slope guidance. Had they flown the precision approach, they may have recognized the problem over a thousand feet above the ground while trying, unsuccessfully to maintain the glide path.
Approach control put the plane on vectors for the visual approach to Runway 32. While on vectors, approach control told the pilot he would have to stay on vectors until the fire trucks, which had been delayed in their station, could arrive.
Factor #2 (In my view, the crucial, fatal error in the sequence) The aircraft commander became indignant at being delayed because of a ground problem with the fire trucks. He protested to the crew that they should have landing priority regardless of the fire truck problems. As related earlier, the pilot flying, Golden Boy, had explained this proclivity of his to me in his initial hiring interview six years before. He said he would get enraged at being cut out of the pattern by commercial jets while instructing students in his small prop plane. So, he would cancel instrument clearance and go visually instead, in retaliation. And, that is just what he did here, with catastrophic results. He told approach control he would cancel instruments and land visually on runway 32. He then turned directly toward the runway. He’d show them!
This threw the cockpit crew, the pilots and flight engineers, into confusion. Instead of running and briefing several emergency checklists, such as landing with an engine shut down, landing heavyweight, and estimating landing brake temperatures, they suddenly found themselves on final approach.
The copilot says, rhetorically and falsely, to the engineer, that they had run all the check lists they had not run. The rest of the cockpit crew did not dispute this error or complain this was happening too fast. Someone needed insist on more time to get their act together, but no one did. The plane was not on fire and there was no urgency to land except in the copilot’s mind. They could have orbited field for hours to prepare, if necessary.
The engineers were so flummoxed one did not realize they had the wrong flap setting, and did not catch the copilot briefing the wrong, full-flap, approach speed. The other engineer did not even know the plane was on final approach. They had several questions about the approach but had no time to bring them up or discuss them with the pilots, specifically on flap setting and approach speed. In this confusion, induced by the rushed approach, they did not catch that the copilot had brought up the throttle for the dead engine, and left the throttle for the good inboard engine at idle power (Factor #3, the second most crucial error caused by Factor #2.) This made them a two-engine heavyweight aircraft without enough power to save themselves.
The aircraft commander briefly directed the engineer to compute brake temperature limits, something apparently prominent in the copilot’s considerations. The left seat, evaluator, pilot warns immediately after the brake temperature discussion, forcefully, “Just put it on the ground, we don’t need a smooth touchdown.” (Factor #4) This demonstrates the hot brake landing of the previous emergency has crowded out the necessity of ensuring a gentle arrival onto the runway due to the heavy weight. This 12,000-foot-long runway had more than enough stopping distance, and this should not have been a primary consideration. Yet, it seemed to be the main worry of the pilots, instead of the power problems of a three-engine airplane and ensuring the proper flap setting.
A few seconds later, as the landing checklist is run, the engineer calls for “Flaps Landing.” The left seat pilot moves the flaps to 100% and says, “Flaps landing,” because that is the usual, rote, response to the checklist item. This error, too, is on the aircraft commander since he had briefed the full flap landing speed of 146 kts instead of the 162 kts for flaps 40%. As the left seat pilot confirms “Flaps landing” (100%) the copilot realizes the error of the wrong flap setting for the heavyweight landing. He calls for flaps to be returned to 40%. However, he does not have enough speed to raise the flaps to 40% without stalling the aircraft. He immediately corrects his command and says not to retract the flaps.The infamous Dover C-5 crash – Part 1: The lead up to the accident and crash animation with cockpit recorder playing
He is also low to the ground, I presume to make sure he gets the plane on the runway expeditiously for his unwarranted fear of hot brakes on landing. So, he is now low and slow and does not have the power to recover on two engines.
Golden Boy senses what is coming and says, in profound understatement, “Guys, I’m concerned.”
Silence now ensues as the plane slows and nears both a stall and the ground. The Ground Proximity Warning System (GPWS) starts its emergency call out: “Too low, terrain!” and “Don’t sink”. The last words are from the left seat instructor pilot: “Bring it out of stick shaker (that warns of a stall). That, unfortunately, can only be done by lowering the nose, which the pilot flying cannot do without crashing.
That angle of attack gauge suddenly leaps from 17 degrees to 20. A stall occurs at 18 degrees. Then the mechanical female voice alerts the crew: “Stall, stall.”
The tail strikes the ground, ripping the entire tail structure from the plane. The rest of the fuselage, from its very high angle of attack relative to the terrain, slams down on the muddy field with the cockpit hitting the ground with 35 g’s. The right wing drags on the ground turning the fuselage right until the momentum shifts and the plane leans over on to the left wing. Outboard engines are torn from the wings. As the plane slows, a fracture in the middle of the crew compartment breaks open as the fuselage continues turning right while the forward cockpit continues tracking toward the runway. This opens a gaping hole between the two sections with only a thin left side hinge holding the sections together
Effects of the crash varied with position in the airframe. Some of the few passengers and crew in the rear troop compartment initially thought the plane had merely made a hard landing.
Due to the whiplash effect of the tail contacting the ground first, the g-forces of the crash increased exponentially moving forward in the structure. This caused the fuselage to crack at the bulkhead that runs through the wall at the rear of the bunk rooms in the crew compartment.
As the plane slid across the field, the fuselage began to break open at the bulkhead. The instructor loadmaster was seated against the bulkhead facing rearward looking across the table at her two students. Suddenly the table and the students began retreating away from her as the fuselage began to open up. Eventually, the fuselage section with her students wound up at a right angle to her and ten yards away.
The cockpit, as previously state, hit the ground at 35 g’s. Although the airframe absorbed much of that force, the pilots all suffered back injuries. A passenger loadmaster in the troop compartment slid down the emergency slide and walked forward to the cockpit to help one of the pilots painfully come down the crew ladder to the ground. He helped him hobble to lay down against a tree on the side of the field. The loadmaster asked what had happened. The pilot said they were slamming the throttles against the firewall, but they couldn’t get any more thrust and couldn’t understand why they could not. They still did not realize they had the wrong throttle forward.
In some ways, our command training method contributed to this crash. When practicing for three-engine procedures, we always used the worst-case scenario which would be an outboard engine, either #1 or #4. When performing the go-around, we had to leave that engine at idle, obviously, or it wouldn’t be a three-engine go-around.
For this crash, however, an inboard engine was out. We never practiced for inboard engine failure, but the habit pattern of leaving the dead engine throttle at idle power prevailed. While it would be hard to confuse an engine failure between #1 and #4, it could easily be confused between #2 and #3, and it was.
In retrospect, the training procedure should have been either to use all four throttles with one engine actually inoperative or to place the dead engine’s throttle forward on the firewall. One of these techniques probably would have saved this crew. Not having thought of the possibility is attributable to us and the command for not preparing for such an event.
A second consideration was the instrument presentation. The new glass cockpit configuration for the new “M” model C-5 had been installed prior to adding the new engines to the airframe. These instruments presented a “Throttle position” indicator above an engine performance indicator. Viewing the throttle position gauge showed three engines advancing, but this meant only the throttle position advancing, not necessarily the engine. The lower gauge showed only engines #1 and #4 had been advanced in power. This confusion provided conflicting information to the crew and contributed to the crash.
As the C-5 crew turned back toward Dover AFB in response to their engine malfunction, they could little realize they were about to fly to controversy, aviation lore, and disaster. There were unusual considerations, to be sure. They were very heavy, and care would be needed to touch down gently to avoid gear damage or failure. They were also on three engines, so depleted power capability would be a vital consideration. Airspeed would have hoarded and given back only at the last moment. For a crew with several greatly experienced individual members, these obstacles should not have provided significant difficulty. Yes, the storied accident chain of events had begun, heavyweight coupled with engine loss, but need not have gone farther. Yet, as we know, it did.
Lurking subconsciously were two distracting factors, the crew’s recent briefing on a hot brakes emergency a few months before, and the aircraft commander’s personality quirk that would erupt at just the wrong time.
The hot-brake imperative that seized the crew could have been overcome during the discussions of the emergency checklists that, catastrophically, were not run. Instead, the previous incident they had just been briefed on the day before seems to become their inappropriate priority.
Although there are many theories of the failures in this episode, I zero in on the aircraft commander, my protégé I had hired and protected for many years before I retired. He had revealed his personality quirk to me on his recommend ride for an upgrade to aircraft commander in the descent to Hickam AFB, in 2004. As related earlier, when I chided him mildly for high airspeed in the descent, he stared me for a moment, then pushed the throttles up to max power before reducing them. He could not tolerate criticism or slights of any kind. His inexplicable throttle action provided a flashy maneuver to re-establish his prowess reputation despite the error he had made. When I finally, emphatically, confronted him about his airspeed a short while later, he shoved in full rudder to side-slip the plane, ludicrously and ill-advisedly, to decelerate and to make me forget his momentary airspeed lapse. I’ve suspected he knew he had this undesirable personality quirk. Perhaps this knowledge drove him to excel in all other areas to mask this chronic weakness.
His crucial mistake, as I have said, took place when approach control directed he would be placed on vectors to await the delayed fire trucks’ deployment. He protested to the crew that this was unacceptable, that, as an emergency aircraft, he should have priority! With that pronouncement, he canceled his instrument clearance to proceed visually and turned directly toward the runway. This allowed no time to run several emergency checklists, left the engineers scrambling to get the instructor to his station, failed to specify which flap setting would be used, and threw the cockpit into confusion.
There was no reason to force an expeditious landing, especially for the A/C’s pet peeve of supposedly being slighted by approach control. He had enough fuel to circle for hours to get his ducks in a row, but instead, in a fit of pique, made the fatal error to turn to final. Had he not done that, I estimate none of the resulting problems would have occurred. They would have briefed and established a partial flap setting, a corresponding approach speed, the need for gentle flare before touchdown, had all cockpit crew in position, and had everyone on the same sheet of music. Once the crew was tuned-up and attentive, the throttle mix-up probably would not have occurred, or, even if it did, would have been caught by an alert crewmember. The time compression of the turn to final left everyone scrambling to catch up. Then the flap confusion, followed by the throttle error, doomed them. Just before the crash, the engineers had just recognized on private interphone the several factors that had been confused but, as one of them said, “Well, it is too late now.”
Article originally appeared on Aviation Geek Club