It was a humid morning on February 25th of 2008 in the Andersen Air Force Base in Guam.
That day more than $1,400,000,000 were lost, two lives were saved, and an engineering work of art was scrapped, the B-2 stealth bomber “Spirit of Kansas” crashed and burst into flames seconds after the lift-off.
The first accident of a B-2 stealth bomber since it passed its testing period, 20 years ago, was caused by humidity that affected the aircraft’s air data system, providing erroneous lectures of airspeed and angle of attack to the Flight Control System.
The erroneous voting and consensus between the different Air Data System’s sensors feeding the “fly-by-wire” control system led to a catastrophic combination of large airspeed over-reading and angle of attack under-reading errors.
This combination of air data system’s measurement errors derived in erroneous pitch-up corrections generated by the aircraft’s Flight Control System, which rendered the aircraft completely uncontrollable in the pitching axis, and left the pilot’s with only one action to consider, the emergency ejection.
Through this article, we will unveil the root causes of the accident from a technical point of view and will go through the unique stealth characteristics of the B-2’s Air Data System and Flight Control System (FCS) to understand why Guam’s high humidity was a key factor which led to the catastrophic accident of the iconic aircraft.
A Design Rendered Around Stealthiness
The B-2 Spirit is a very special aircraft from every point of view. It was conceived to be “the stealthiest aircraft of its time”.
It belongs to the third generation of low-observable aircraft, which were designed using computer technologies of the 1980s, and has an estimated Radar Cross Section (RCS) of less than 0.1m². In other words, the B-2’s radar echo intensity is the same as that of a bird. That is really incredible, considering that the B-2 has a wingspan of 52m.
Every detail of the B-2 Spirit has been designed to maximize the aircraft’s survivability in hostile environments, with stealth as the main driver to neutralize the enemy’s radar systems.
It was this stealthiness requirement that shaped most of the B-2’s systems. Maybe the most notable one is the Air Data System of the aircraft, one of the main responsible for Guam’s mishap.
But why B-2’s Air Data System is so special?
A Sophisticated Air Data System
In every aircraft, the Air Data System (ADS) provides crucial information to the Flight Control Computers and to the Engine Control Units. Its main objective is to measure the following parameters:
- Total Pressure
- Static Pressure
- Barometric Altitude
- Dynamic pressure
- External Temperature
- Mach number
- Angle-of-Attack (AoA)
- Angle-of-Sideslip (AoS)
In standard aircraft, all of these flight parameters are measured by specific sensors like the pitot tube, the static pressure port, and the AoA and AoS vanes. The ADS’ sensors can be easily identified in the following image of an Airbus A350 XWB, where you can spot the following sensors:
- Ice detector
- Multi-function probe
- Static port
- Angle-of-Sideslip vane
But all of these sensors that are indispensable for the ADS, especially the vane sensors which measure the AoA and AoS parameters, are completely counterproductive for the aircraft’s radar signature, and this is of paramount importance for a stealth bomber like the B-2 Spirit.
So, how did the engineers at Northrop corporation solved this problem? How did they manage to measure the B-2’s angle of attack and angle of sideslip without using any vane sensor at all?
They relied on a technology developed during the aviation era of the first hypersonic flights, a new way of measuring the external airflow incident angle using only information of the local static pressure measured at different locations in the external geometry of the aircraft.
It’s all a matter of apriori knowing the aerodynamic pressure distribution on these spots for certain flight conditions to estimate then the angle of attack and sideslip using the differential pressure measurements provided by the static pressure ports.
This technique was first used in the North American X-15, the pioneer of the manned hypersonic flights, and later on, used in the Lockheed Martin F-117A and curiously, also in the Space Shuttle, but in this last case for other technical reasons not related to radar signature reduction.
So what is the specific implementation of this technology in the B-2 Spirit?
Can you see the white circles within the red-highlighted areas in the image below?
These small dots are static pressure sensors, sometimes also called Port Transducer Units (PTUs), distributed along the upper and lower surface of the aircraft. In a similar way to that shown below (NASA’s original sketch), these little-holes are connected to redundant pressure transducers that provides the differential pressure measurements.
These transducers provide the required information to the Air Data (A/D) conversion unit to compute the estimated Angle of Attack and Angle of Sideslip for the Flight Control Computer and the cockpit displays.
In this photo, you can see with more detail the location of the static pressure ports on the aircraft’s lower surface.
As you can imagine, humidity can affect the readings provided by the PTUs, and thus pre-flight calibrations need to be performed to account for any system’s deviation from nominal performances.
Nevertheless, if this calibration is done only once during the pre-flight checks, and later on water and humidity in the PTUs disappear, the system will be again uncalibrated prior to takeoff, and this was a concern for the most experienced B-2’s mechanics. They knew the system’s calibration was meant to be performed in the same conditions where the system was expècted to operate (in-flight), where no humidity (or stagnated water) was expected, and thus, the calibration automatic procedure was performed (usually) with the pitot heater switched on.
Guam’s Mishap Sequence
During the B-2’s 2006 deployment in Guam, the extremely humid environment required frequent Air Data System calibrations. Line maintenance technicians telephoned B-2 manufacturer technical representatives to ask for advice on this new environmental issue.
Support engineers recommended using the aircraft’s pitot heater to dry the static pressure ports before calibrating the ADS. This is key for the next sequence of events.
But this technique was not formalized in a technical order change or captured in a “Lessons Learned” report. Only
some of the ground crews and pilots working with the B-2s during their 2007-2008 deployment were aware of the ADS’ sensitivity to moisture and the pitot heat workaround.
What Really Happened?
In the morning of February 25, 2008, the “Spirit of Kansas” and another B-2 were preparing to return to Whiteman AFB, after a four-month deployment at Andersen AFB in Guam.
The mishap aircraft would follow the lead aircraft. The sequence of events that led to the mishap was as follows:
- 9:29 AM: During the pre-flight check, the mishap flight crew received an “AIRDATA CAL” message, indicating that the Air Data System needed to be calibrated. The pilots and flight control specialist working on the aircraft were not aware of the pitot heat technique, so they calibrated the Air Data System without turning on the pitot heat to dry the PTUs. The calibration procedure created a significant bias in three of the twenty-four PTUs.
- 10:29 AM: As the crew prepared for takeoff, they turned on the pitot heat (as indicated in the checklist), which dried the moist sensors. Skewed air data caused an altimeter error of 136 feet above actual airfield elevation, but the crew did not notice the error because there were no field elevation placards in view near the runway.
- 10:30:12 AM: The B-2 began its takeoff roll, but approximately 19 seconds after the brakes were released, the Master Caution Light illuminated, along with a Flight Control System caution on the status display. The crew noted air data fault indications, but approximately six seconds later, the FCS rescinded the caution lights. The pilots continued the takeoff because all caution lights had cleared and their instruments indicated that airspeed was well above the B-2’s go-no-go airspeed of 100 Knots Indicated Air Speed (KIAS).
- 10:30:49AM: Pilot #1 rotated the bomber’s nose for takeoff at an indicated speed of 142 KIAS (actual speed was later estimated to be 132-134 knots) with a normal stick force, attempting to establish a standard pitch attitude and climb -rate. As the nose gear lifted off the runway, the FCS calculated a negative angle of attack (estimating a severe nose-down attitude) based on skewed ADS data and pitched the aircraft nose up 30°. Pilot #1 tried to regain control of the aircraft, but the low energy condition (high angle of attack, high gross weight, high temperature, low airspeed) proved unrecoverable. The aircraft yawed and rolled to the left; as the left-wing scraped the ground, both crewmembers ejected.
The Lessons Learned
1. Configuration Management
This incident illustrates the crucial need to document changes and workarounds developed in the field. The pitot heat technique would have pre-empted this accident if it had been widely-known and incorporated into the standard calibration process. Every procedure—no matter how simple or mundane—must be treated as if mission success and the safety of the crew depends on properly performing every detail of the procedure. This requires an open line of communication between team members and between various teams.
2. Systems Engineering
This incident illustrates the importance of developing a comprehensive understanding of the systems and hardware we work with. The accident board’s report implies that, had personnel fully understood the significance of the air data system calibration, the pitot heat technique would have been written into the formal calibration procedure.
Even complex sustainable systems should not require profound knowledge in the field. From a design standpoint, the relationships between all the elements in a system should be transparent.
Another lesson to draw from this incident is a reminder that equipment tests and calibrations should simulate field use as closely as possible. Here, the process called for equipment calibration without pitot heat even though pitot heaters would be used during the flight. This was probably intentional— pitot heaters were known to overheat if left running on the ground—but as we design calibration and testing procedures, the discrepancy between moisture in the PTUs on the ground and the dry PTUs in flight is a reminder that we need to duplicate field use conditions in testing and calibration whenever possible.
Photos of the Accident
Although images from the wreckage are sad, they are a reminder of what you don’t know can kill you. At the end of the day, the excellent airmanship demonstrated by the aircrew saved their lives, no one was hurt in the accident, but the mishap cost more than a billion dollars to the American contributors.
Did you like what you have just read? Recommend this post by clicking the heart button so others can see it!
Rodney Rodríguez Robles is an aerospace engineer, cyclist, blogger, and cutting edge technology advocate, living a dream in the aerospace industry he only dreamed of as a kid. He talks about coding, the history of aeronautics, rocket science, and all the technology that is making your day by day easier.