Search Dryden


Text Size

Volume 46 | Issue 1 | January 2004


photo: X-31 in attack position
The X-31 performs a high-angle-of-attack maneuver during the aircraft's mission over Dryden, which included 406 research flights from 1992 to 1995. The X-31 was the first international X-plane and recorded a record number of flights - a total of 580.
NASA Photo / Jim Ross

X-31's loss

Jay Levine
X-Press Editor

On a brisk winter morning, German pilot Karl-Heinz Lang climbed into the cockpit of the X-31 and, as he had on so many other mornings, began the checklist that would allow him to ease down the runway and into the skies.

The 43-minute mission was nearing its conclusion when ice began forming on the X-31's pitot tube. The ice caused incorrect airspeed information to be sent to the craft's flight control computers, which were responsible for reconfiguring the aircraft for lower speeds. The result was a series of sudden, uncontrollable oscillations in all axes. The aircraft then pitched to 90 degrees angle of attack, and Lang was forced to eject as the X-31 crashed near the northern boundary of Edwards Air Force Base.

That was the scenario Jan. 19, 1995, when Dryden lost one of two X-31 research aircraft. Six men who had key roles in the X-31 program at the Center returned for a Make Dryden Safer event Jan. 13 to discuss the accident and to detail some of the reasons for its occurrence. The idea was to give employees who are new to Dryden a chance to learn from the past, as a hedge against repeating it.

"I was in charge. I cannot blame anyone else," said former Center Director Ken Szalai in the opening remarks of the session. "But lessons from that might be applied to current projects."
Szalai noted that the X-31 team was a highly skilled and experienced one, saying that, "If it happened to them, it can happen to you."

Root causes of the incident were outlined in the presentation, which served to illustrate that not just one factor caused the accident, but rather that an entire set of circumstances aligned to doom the aircraft.

"Many people involved had the opportunity to break the chain of events that led to the mishap," Szalai said. The warning signs are far easier to recognize in retrospect, he said, as he recommended several actions for Dryden flight research staff to consider that would minimize the potential of a similar accident in the future.

A primary contributing factor to the incident involved installation of an experimental pitot tube, which replaced the X-31's original pitot tube. The new pitot-static pressure system of an X-31 utilized a small tube called a Kiel probe at the nose of the aircraft to provide air speed data both to the pilot's instruments and to mission control center monitors. It is a key source of information in flight and was being used to evaluate the system for high-angle-of-attack maneuvers. The original pitot tubes had heaters, but the experimental system did not. This was not considered a safety risk by engineers.

Because the experimental pitot tube did not have a functional heater, X-31 mission rules prohibited flights in any type of precipitation or in the clouds. They did not, however, include a separate stipulation specifically prohibiting flight during potential icing conditions, despite simulations that had showed icing of the pitot static system could lead to loss of control under certain circumstances.

photo: Discussion about the loss of the X-31
Current and former employees connected with the X-31 project at Dryden gave a Jan. 19 presentation to discuss the loss of a research aircraft and how lessons learned during that flight can help researchers avoid similar situations. The panel included, from left, former Center Director Ken Szalai, Dryden X-31 project manager Gary Trippensee, research pilot Dana Purifoy, acting Research Engineering director and lead X-31 controls engineer Pat Stoliker, X-31 operations engineer Brad Neal (one of two X-31 operations engineers) and Dryden X-31 chief engineer John Bosworth.
NASA Photo / Tom Tschida

Additionally, the experimental probe installed on the X-31 was more susceptible to icing - a fact discovered through wind tunnel research performed after the accident. Information had been distributed among crew and engineers explaining the pitot tube change, but no formal closed-loop system (an internal system of checks and balances) had been in place to ensure that everyone had read and understood the change - in retrospect, a critical lapse in communication procedures.

Lang noticed the abnormality reflected in his instrumentation and told the control room that he was switching on pitot tube heat. In reconfiguring the controls for landing shortly afterwards, he indicated that he was leaving the pitot heat on. The ground controller then said that the switch was nonfunctional. That information, combined with other conditions being observed in the control room, could have provided critical warning signs of the pending disaster.

Dryden acting Research Engineering director and the Center's lead X-31 controls engineer, Patrick Stoliker, told panel attendees that the discrepancies between the X-31's airspeed and altitude had been observed in the control room, but that information had not been shared with the entire control room staff. In addition, he said, the significance of the unusual airspeed change had not been well understood at the time of its occurrence.

Stoliker also explained that the aircraft parameters for the research flight included a mission rule not to fly in conditions where visible moisture was present, but said that, in retrospect, officials now understand that a separate prohibition against flight during icing conditions should have been specified as well.

Dryden X-31 chief engineer John Bosworth focused his remarks on the nature of flight research work and how the X-31 accident served to illustrate the complexity of the process.

The accident footage, Bosworth said, which was replayed at the panel discussion, still affects him.

"When I see the tape I get angry. [At his control room station] there was 10 (decibels) of static and it was not clear [to Bosworth] what the pilot was saying," he said of the transmission of Lang's comments to mission control staff indicating that he was "turning on" the heat switch.

"In retrospect, I feel I should have called for a stand-by so that everyone could clearly state what was going on," he said, referring to the option of directing staff to halt activity while an evaluation could be performed.

The aircraft had redundant air data, Bosworth continued, and the outcome of the X-31 incident might have been very different had communication been better at several key moments.

For example, he explained, air data is redundant electronically, but not mechanically. A pilot-selectable control mode was available that, had it been used, could have saved the vehicle. Better communication about the nature of the problem might have generated discussion and consideration of selecting that control mode, Bosworth said.

"The best we can do is find and fix as many weaknesses as we can. We approach flight test with options knowing that we can never get to zero risk," he said.

Dryden pilot Dana Purifoy, who flew chase during the X-31 incident, confirmed that better communication could have made a significant difference in the outcome.

"As the chase, (air-to-ground) communication became a main problem," Purifoy said. "I was hearing only half of what was happening (because only control room transmissions were heard as a result of how the air-to ground communication system had been configured). As pilots, it's our responsibility to be part of the team.

"(Chase pilots) add value to projects," he added. "It's not cut and dried. Mission rules need to be evaluated in the context of the moment."

Brad Neal, one of two Dryden X-31 operation engineers, said he had performed the installation of the new pitot probe on the X-31. The absence of a functional pitot tube heater had not seemed unusual to him, said Neal, who had come of age as an engineer during the era of F-104 aircraft. Production F-104s did have heated pitot tubes, but research 104s had unheated pitot research airdata systems.

During the circumstances leading up to the accident, there was ample opportunity for staff to question superiors and procedures, communication that could have led to more questions and possible ways of averting the accident. But in flight research, complacency is the enemy, as Dryden X-31 project manager Gary Trippensee pointed out; adequate consideration of potential problems can become less likely when, as in the case of the X-31, accepted procedures had to that point resulted in 523 successful research missions.

"It's human nature to back off, but you have to stay on top of things all the time," Trippensee said of conducting a flight research project.

photo: X-31 in flight
The x-31 flies in a flight research mission. The first international X-plane flew 406 sorties at Dryden.
NASA Photo / Jim Ross

"It all unravels quickly. Flight research is a difficult environment, and unless you're prepared for situations, sorting out a problem in a few seconds and taking the correct action is unlikely. Failure situations must be practiced in the control room by the control room staff to be prepared and primed to act," Szalai added.

The accident investigation board reviewing the X-31 crash recommended that training be conducted on the system safety analysis process, that procedures be implemented to assure all test team members receive configuration change notices, and that improvements be made in the remaining X-31 to prevent similar single-point failures.

"It's things that you would consider routine that cause trouble," Szalai said, "because there is no such thing as 'routine' flight research." As an example, he noted that it was, ironically, ice in the desert that ultimately brought down the X-31, and also that it was a piece of foam that damaged the Shuttle's wing and led to the loss of Columbia.

However, Szalai reiterated that he is a strong believer in flight research.

"Complex, high-risk programs can be done safely," he emphasized.

The bulk of flights for the two X-31s were conducted under the auspices of an International Test Organization (ITO) based at Dryden.

This spring, the team was honored with the prestigious International Council of Aeronautical Sciences (ICAS) von Kármán Award.

The ICAS-von Kármán Award for international cooperation in aeronautics is usually granted biannually. However, in conjunction with the Centennial of Flight year, the organization bestowed a special von Kármán in 2003 on the U.S. and German team, in which NASA was a partner. The award was presented "for over 20 years of successful Trans-Atlantic R&D (research and development) teamwork producing the first-ever International X-plane and significant breakthroughs in thrust-vectoring control."

The award was founded in 1980 in memory of Theodore von Kármán, a leading figure in the world of aeronautics and a founding member of the ICAS. Von Kármán is considered one of the great aeronautical scientists of the 20th century and was an outspoken proponent of international scientific cooperation. The award recognizes the team of countries that contributed to the outstanding technical achievement of the program for which the team is nominated and ultimately awarded the medal.

The X-31 program researched thrust vectoring - using engine thrust to control the aircraft - and cutting-edge technologies and flight controls as a means of enhancing the maneuverability of future fighter aircraft. The project was managed by the Defense Advanced Research Projects Agency (DARPA), with participation by NASA, the U.S. Navy, U.S. Air Force, Rockwell Aerospace (now the Boeing Company), the Federal Republic of Germany and Deutsche Aerospace (now Daimler-Benz).

The initial flight phase of the highly successful program, which began in October 1990, was based at Palmdale, Calif., where Rockwell Aerospace assembled the aircraft. The ITO was based at Dryden in February 1992.

The X-31 team twice achieved as many as five flights per day at Dryden and recorded 406 flights during the aircraft's flight series at the Center from 1992 to 1995. In all, the first international X-plane recorded a record number of flights - more than 580 sorties.

After the accident, the remaining X-31 was brought to flight status and in June 1995 appeared in a spectacular performance at the Paris Air Show.

As part of flight research in the program conducted at Dryden in 1994, software was added to simulate tailless flight at supersonic speeds using thrust vectoring.

Five years later, a follow-on program called Vectoring ESTOL Control Lift Tailless Operation Research (VECTOR) focused on improved operational performance. That program, as well as the most recent demonstration flight series on use of thrust vectoring as a means of achieving Extremely Short Takeoff and Landing capabilities, was flown at the Naval Flight Test Center at Patuxent River, Md.

The remaining X-31 is currently on display at the Deutches Museum in Germany and will be returned to the U.S.

X-31 Photo Gallery

X-31 Movie Gallery