Lack of training and inexperience with 'Dutch roll' tied to KC-135 loss in Kyrgyzstan
The loss of a U.S. Air Force KC-135 Stratotanker over Kyrgyzstan in 2013 has prompted significant changes to flight and emergency training for air force tanker crews.
Threepersonnel died when the KC-135R, call sign “Shell 77,” crashed just 11 min. after takeoff from Manas International Airport, on May 3, 2013. The aircraft was one of several KC-135s forward-based in the former Soviet republic flying aerial refueling operations for combat aircraft supporting ground troops in Afghanistan.
Accident investigators concluded in their March 6 report that while the primary causal factors were a factor, those problems had been exacerbated by crew inexperience and by subpar organizational training programs, crew composition, and cumbersome procedural guidance.
The problems began shortly after the afternoon takeoff from Manas. As the aircraft got airborne the crew immediately began to experience a flight condition known as rudder hunting, which prompted the aircraft's nose to yaw left and right by 1 deg., an issue caused by a malfunction with the aircraft's series yaw damper (SYD)—part of the KC-135R's flight-control augmentation system (FCAS). The crew identified the SYD as a potential source of the issue but instead of turning the system off, they attempted to correct the yawing motion by using rudder and aileron inputs as well as turning on the autopilot, causing the oscillations to worsen. Some yawing issues had been reported during the aircraft's transit flight to Manas fromMildenhall, U.K.
After the autopilot had been turned on a second time and failed to reduce the effect of the oscillations, the captain assumed control and began to use the rudder in a bid to correct the now-considerable Dutch roll effect—where the aircraft yaws several degrees to the left and right.
The report states that a series of alternating small rudder inputs caused by Dutch roll-induced acceleration forces was in reaction to the pilot's foot pressure on the rudder pedals, which sharply increased the oscillations. These fluctuating rudder movements, coupled with slight right-rudder use while rolling out of the turn, compounded the Dutch roll severity and produced what officials describe as extreme airframe stress.
“The cumulative effects of the malfunctioning SYD, coupled with autopilot use and rudder movements during the unrecognized Dutch roll, generated Dutch roll forces that exceeded the aircraft's design structural limits,” the report states.
The forces resulted in the structural failure and separation of rear fuselage bulkheads 1560 to 1440 with both vertical and horizontal stabilizers breaking away from the rear of the aircraft, immediately sending the KC-135 into an 83-deg. nose-down attitude. As the aircraft broke through the clouds, eyewitnesses on the ground saw the starboard wing separate, which caused the exposed fuel to explode. One of the aircraft's engines landed within 30 ft. of a bystander on the ground.
Much of the wreckage was strewn across a 2.5-mi. area across mountainous terrain near the village of Chaldovar. Some wreckage fell even farther.
Announcing the board of inquiry's findings at Scott AFB, Ill., Brig. Gen. Steve Arquiette, inspector general of the Air Force's Air Mobility Command (AMC), said that analysis of the accident showed that the crew lacked the experience to diagnose the phenomena of Dutch roll, [and] had been hampered in their attempts to recover because of the “cumbersome” nature of the emergency procedures.
“The crew encountered a condition that they had not realistically experienced in training, and when coupled with decisions based on their relatively low recent experience levels, were presented with an unrecognized hazardous and difficult situation to overcome,” said Arquiette.
Emergency manuals for the KC-135R contain no fewer than 21 emergency procedures discussing rudder control difficulties, but they are spread out over 177 pages, the report points out.
Furthermore, “knowledge and experience of Dutch roll is limited in the training program,” the report states. “The proficiency level required for Dutch roll recognition is familiarization, meaning each pilot must only discuss this topic and is not required to perform the maneuver.”
Post-accident, engineers and pilots involved in the probe found they could not recreate in the simulator the conditions experienced in the crash, an issue that is now being addressed. Prototype modifications already have been fitted to simulators at Scott AFB. Other KC-135 simulators around the U.S. will be modified following the award of a contract in February.
The changes to the simulator will improve the way KC-135 sims move, giving them a higher fidelity to such action on actual aircraft. Engineers have developed seven new rudder malfunction scenarios to allow more effective training to overcome the malfunction.
Arquiette says that since the introduction of the KC-135R, the need to familiarize crews with the issue of Dutch roll had virtually disappeared because the “assist systems” introduced to that version—installed to help crews in the event of an engine failure with the more powerful—meant that such issues were rarely experienced or reported. Crews flying the older KC-135A had been better trained to deal with the phenomena; even though that version was more unstable in flight, it did not need the assist systems because the asymmetric effect of engine failure on the older model had less effect on the aircraft's flight characteristics.
Since this accident, AMC has carried out checks on the SYD and FCAS across the 431 KC-135-class aircraft within the fleet.
AMC, Boeing and the Air Force's Special Programs Office say they are rewriting the sections of the flight manual regarding rudder-systems operation and malfunction identification that can occur during flight. KC-135 crews are now said to be aware of the issue and are expected to report any rudder anomalies they experience in flight.
According to AMC, since the Shell 77 accident there have been 36 events of aircrew-reported rudder anomalies, of which Flight Data Recorder analysis confirms seven instances of mild rudder oscillations that are said to be similar to those associated with the Shell 77 crash.