Ten Years After Columbia

A moment by moment review of how the Columbia space shuttle fell apart high above the western United States.

by Rand Simberg

It had been a successful mission, and the crew prepared to slowly start the long controlled fall home after completing the retroburn from the Orbital Maneuvering System. But unbeknown to them, the first hint of trouble appeared a few minutes after Entry Interface, defined as 400,000 feet altitude.

The vehicle was still moving at about Mach 24 — twenty-four times the speed of sound. The yaw moment had changed almost imperceptibly, with a gentle tug to the left — there was a slight asymmetry in the aircraft’s aerodynamics, but no one on the ground or in the cabin noticed it at the time.

Twelve seconds later, a temperature sensor indicated that a hydraulic brake line in the left wing was warmer than it should have been, and was slightly out of specification in that regard. As the vehicle was gradually slowing down, perceived gravity slowly growing to an indiscernible hundredth of a gee, fiery hot plasma was infiltrating through the hole in the leading edge, insinuating itself into the interior of the wing and starting to damage it, but none realized it yet.

A minute or so later, as they approached the California coast from three hundred miles out, an off-nominal rolling moment appeared, more evidence of subtle changes to the vehicle’s outer shape. A few seconds later, Mission Control received a signal that several sensors were starting to indicate problems, but no one on board or in the control room was yet aware.

Half a minute later, still going Mach 23 after crossing the coast line, observers on the ground in California saw an eight-pound piece of the left wing separate from the vehicle, creating a luminescent trail in the plasma, though they didn’t know at the time what they were seeing. No one in Houston or in the cabin was aware of this. About the same time, the side-slip angle exceeded any previous entry experience, as the vehicle was no longer moving in a pure forward motion. A little over half a minute later, the left elevon started to trim to compensate for the off-nominal forces, but no one noticed at the time.

A few seconds later, as the craft crossed the border into Nevada, someone in Mission Control finally noted the temperature-sensor anomalies from a couple minutes earlier.

A few more seconds after that, as the crew was doing a pressure check on their suits in preparation for landing, the brightest piece of debris was shed, but the sensors indicated nothing, and no one on board saw it.

The crew started to really sense the deceleration — about a third of a gravity — a minute or so later, as the dynamic pressure increased to forty pounds per square inch. This was a value that allowed the aero surfaces to take over from the small rockets that had been controlling the vehicle’s attitude. Thirty seconds later, they commenced the first roll reversal to bleed off excess energy, from right-wing low to left-wing low, starting the standard series of S-turns as they approached their eventual landing site. For the veterans on board, the entry seemed to be going normally.

But a picture taken from an infrared camera in New Mexico showed some bulges in the flow field on the left wing that couldn’t be explained by the vehicle attitude. No one saw it until later. The trim on the elevon was now departing sharply from a standard entry, as it fought to maintain the nominal attitude in the face of increasing asymmetric aerodynamic loads.

A little over half a minute later, alarms went off in the cabin as sensors indicated problems with pressure in the left main tire. A few seconds later, there was a false signal that the gear itself on that side was deployed and locked. After a minute or so, the crew received a call from mission control about the tire-pressure issue. Within a few seconds, the left elevon had lost control authority to maintain the proper pitch and roll rate. At that point, the yaw jets, which usually operated in pulse mode, started to fire continuously in a vain attempt to prevent the vehicle from turning to the left. The crew could have seen the indicator lights for this, and a rapid decrease in the propellant on the gauge, but we’ll never know if they did, because the signal was lost at about that time and mission control had received their last transmission. But the flight control system started annunciating its own master alarm, as actuators started to fail, which wouldn’t have gone unnoticed.

Within a few seconds, the hydraulics themselves failed, and the vehicle was out of control. It started to transition from a controlled glide to a ballistic trajectory, like a misshapen cannonball. What had been a forward-moving aircraft started to corkscrew, something that the commander and pilot couldn’t have failed to observe, both from the unexpectedly changing acceleration and shifting horizon in the windshield. The mission specialists behind them on the flight deck probably realized that things were going south as well. A few seconds later, the cover of the left Orbital Maneuvering System pod was torn off, and the top of the cabin started to overheat.

The acceleration was starting to build up, but the crew did manage to restore the mode of the digital autopilot to auto when it was accidentally changed to manual, a few seconds later. At that point, the reaction control system was out of propellant, having fought as long as it could to keep the vehicle under control. Seeing the loss of hydraulic pressure from the failed actuators, the flight crew attempted to restart the Auxiliary Power Units that drove the pumps, to no avail. It was likely their last conscious act to save themselves.

The left wing was tearing off, shearing the softened and melting aluminum, and the noise within the cabin would have been horrific. The vehicle started to pitch up.

Within another ten or fifteen seconds, attacked by the dynamic pressure from a direction never planned by the vehicle’s designers, the payload-bay doors blew off, exposing the payload bay itself to the brutal force. The hurricane of plasma separated the forebody of the vehicle, with the crew cabin, from the rest.

With all power lost, the cabin went dark other than sunlight and earthlight from the windows, but not silent, and started tumbling. The hull itself was quickly breached, and rapidly started to lose pressure. The crew had never closed the helmets of their flight suits, but it probably didn’t really matter. From either the rapid accelerations or the decompression, unconsciousness likely came very quickly, which is merciful, because the heat started to melt the structure itself, spattering molten metal within.

The fiery cabin, now briefly their airborne coffin, continued its long, thirty-four mile uncontrolled parabolic fall over northern Texas, with the charred remains of its inhabitants, until it came completely apart from the increasing atmospheric pressure forces and all took their own path to the earth. Ironically, and perhaps almost poetically, had it occurred an orbit later (as they had considered earlier that day for weather reasons at the planned landing site in Florida), Johnson Space Center would have been ground zero. The heaviest debris would have rained over the southwest Houston neighborhoods in which they had lived at supersonic speeds, and in which their co-workers continued to live. It might have even hit the Mission Control Center itself, where so many of the fateful, and ultimately fatal decisions had been made.

We will never know what thoughts went through the minds of Columbia’s doomed crew in that last minute in which they were becoming aware of their imminent peril. Perhaps, being the rational pilots and engineers that they were, they were simply focused on trying to save the ship. It could have been something as pragmatic as, “I wish that we’d had a better look at the leading edge of that wing.” What we can know is that they died doing something that they had worked years to be able to do, and they loved it, and thought it important. If we now fear to open up the high frontier because of what happened to them, they will have died in vain.

This article is the prologue of the forthcoming book, titled “Safe Is Not An Option: How Our Futile Obsession With Getting Everyone Back Alive Is Killing Our Future In Space.”

Rand Simberg is a recovering aerospace engineer and a consultant in space commercialization, space tourism and Internet security. He offers occasionally biting commentary about infinity and beyond at his weblog, Transterrestrial Musings. He is an adjunct scholar at the Competitive Enterprise Institute.