In 1981, the shuttle Columbia became the first to successfully reach space. From 1981 to 2003, Columbia carried dozens of astronauts on missions, spending a total of 300 days, 17 hours, 40 minutes, and 22 seconds in space. Columbia’s last flight was STS-107, a 16-day research mission focused on scientific experiments. The crew was a diverse group and included the first Israeli astronaut, Ilan Ramon, and the first Indian-born woman in space, Kalpana Chawla.
As Columbia went through the final minutes of its re-entry on February 1, 2003, NASA lost contact with the shuttle over Texas. Controllers spent several minutes trying to make contact with the shuttle. At the Kennedy Space Center, the families of the astronauts waited to welcome them home, but contact was never made again. Video footage emerged of Columbia breaking into pieces as it re-entered the atmosphere. All seven crew members were killed.
What caused this disaster?
Immediately, investigators rushed to determine what had caused the accident.
The Columbia Accident Investigation Board (CAIB) was formed to review the causes of the disaster in the hopes of preventing it from ever happening again on a future shuttle mission. Harold W. Gehman Jr., former commander-in-chief of the U.S. Joint Forces Command, chaired the board, and members included former astronaut Sally Ride.
The shuttle’s main fuel tank was covered in thermal insulation foam intended to prevent ice from forming when the tank filled with liquid hydrogen and oxygen. Such ice could damage the shuttle if shed during lift-off. At 81.7 seconds after the launch from the Kennedy Space Center, a suitcase-sized piece of foam broke off from the external tank and struck Columbia’s left reinforced carbon-carbon (RCC) panel. According to CAIB, this likely created a six-to-ten inch diameter hole that would allow hot gases to enter the wing when Columbia re-entered the atmosphere.
The foam debris problem was well documented in the years before Columbia’s doomed mission in early 2003. NASA referred to this as “foam shedding.” Indeed the insulation had been observed falling off on four previous flights in 1983, 1990, and 1992 and in-flight STS-112, which occurred just two launches before the ill-fated Columbia mission. The STS-112 had actually been the first flight with the “ET cam,” a video feed mounted for the purpose of giving insight to the foam shedding issue. During a review of the video feed, a chunk of foam broke away and hit a ring near the bottom of the left solid rocket booster, causing a dent. Yet even after STS-112, NASA leaders decided to move forward, citing no new concerns.
Video taken during STS-107’s lift-off was routinely reviewed two hours later and revealed nothing unusual. The following day, higher-resolution film that had been processed overnight revealed the foam debris striking the left wing, potentially damaging the thermal protection on the Space Shuttle. At the time, the exact location where the foam struck the wing could not be determined due to the low resolution of the tracking camera footage.
In a risk-management scenario similar to the previous Challenger disaster, NASA management failed to recognize the relevance of engineering concerns for safety. Engineers had requested to review higher resolution imagery from the Department of Defense, but NASA denied their requests.
Much of NASA’s risk assessment hinged on damage predictions to the thermal protection system (TPS). These fell into two categories: damage to the silica tile on the wing lower surface, and damage to the RCC leading-edge panels. The TPS included a third category of components, thermal insulating blankets, but damage predictions were not typically performed on them.
Before the flight, NASA believed that the RCC was very durable. Damage predicting software known as Crater was used to evaluate possible tile and RCC damage. The Crater predicted severe penetration of multiple tiles in the event of a TPS tile area impact, but the NASA engineers downplayed this. The model overstated damage from small projectiles, so engineers believed that the model would also overstate damage from larger spray-on foam insulation impacts.
Despite concerns about the energy impact by the spray-on material, NASA managers ultimately accepted the rationale to reduce predicted damage of the RCC panels from possible complete penetration to slight damage to the panel’s thin coating. The NASA Mission Management Team felt there was insufficient evidence to indicate that the strike was an unsafe situation, so they declared the debris strike a “turnaround” issue (not of highest importance) and denied the requests for the Department of Defense images
In addition, many senior NASA managers were influenced by their belief that nothing could be done even had the damage been detected. This affected their stance on investigation urgency, thoroughness, and possible contingency actions. They decided to conduct a parametric “what-if” scenario study more suited to determine risk probabilities of future events, instead of inspecting and assessing the actual damage.
NASA official Wayne Hale noted that the mindset of operations was, “If it has been damaged, it’s probably better not to know. I think the crew would rather not know. Don’t you think it would be better for them to have a happy, successful flight and die unexpectedly during entry than to stay in orbit knowing that there was nothing to be done until the air ran out?”
In response to the Columbia disaster, the investigative team performed foam impact tests and used a compressed air gun to fire a foam block of similar size and mass to that which struck Columbia at the same speed. RCC panels from NASA stock, along with actual leading-edge panels from shuttle Enterprise, were used to represent Columbia’s left wing. Investigators released dozens of foam blocks over many days, but these produced only cracks or surface damage to the RCC panels. Through further analysis of information from Columbia’s flight data recorder, which was found in Texas, investigators could isolate the probable impact site to a single panel: RCC wing panel 8. In one final test on July 7, the team fired a foam panel at RCC panel eight, which created a hole 16 by 16.7 inches.
CAIB issued its report and confirmed the immediate cause of the accident was a breach in the leading edge of the left wing caused by insulating foam shed during the launch. The report also blamed cultural issues that led to the accident. The CAID was highly critical of NASA’s decision-making and risk management.
CAIB was also able to determine the final moments of the astronauts with some accuracy. The report found that Columbia depressurization occurred so rapidly that the crew did not have time to prepare their suits for full protection from the loss of cabin pressure. Some crew members were not wearing their safety gloves, and another was not wearing a helmet. In response, NASA developed new policies to give the crew more time for the descent. Also, the crew’s harnesses malfunctioned during the descent. The harnesses in the three remaining shuttles were upgraded as a result.
On February 4, 2003, President George Bush led a memorial in honor of the astronauts at the Lyndon B. Johnson Space Center. After Columbia’s loss, the space shuttle program was suspended for more than two years. The actual cause of the foam loss was not determined until 2005, when x-ray photographs of another tank showed that thermal expansion and contraction during filling caused cracks that led to foam shedding.
Hubble, EVA, many options to confirm damage. Proximity of breach to hypergolic Orbital maneuvering engine (OMS pods). During this era there was a Delta launched ever month from the Cape repurposing one Delta with a thermal insulation, patching and more tiles to be applied by and EVA could have easily prevented this. It would have meant asking others for help which NASA never does. 100% on the Engineers at Houston and not their finest hour.
It is best not to underestimate findings when engaged in designs with such critical roles
Tough when evaluating low frequency/high consequence risk! Particularly hurtling man made objects into and back from space. Budgets and management vs. over engineers. The bosses have to recognize the cultural creep that takes place in this process and not embed it beyond a tipping point. Boeing is a good example of going too far. From a human perspective so is the Jan 6 terrorist attack on the US Capital.
My goodness you folks missed a couple of hitches in your politically correct giddyup:
1. Suits are irrelevant. When the burn thru occurred, a wing fell off giving an instantaneous 40 G (approx) roll. The Astronauts were killed instantly.
2. You can’t do anything – horsefeathers!! Columbia had a Spacehab module attached. With the extra oxygen and hydrogen, a rescue mission could have been attempted – it was just that NASA was in FULL POLITICAL DENIAL.
3. The space program had been severely crippled because of a 1995 brain drain. In 1995, NASA led a pogrom to eliminate all senior contract engineers because ‘they cost too much’. As a senior engineer with a professional license and over 15 years experience I was cut loose and replaced by a fresh out of college engineer from the LOWER 50% of the class of engineers from a college that was parallel to a Division II or III NCAA entity. Thus, there was NO ONE with the ability and/or confidence to stand up to NASA Political Stupidity.
4. An additional part of the problem is that our current engineering schools focus so much on CAD and computer solutions that it is truly a wonder that new engineers are able to flush the toilet without a set of computer instructions and a CAD drawing.
I’ve always been leery of the foam strike as the root cause of this incident. Low density foam exfoliated from the storage tank would have to have a very high velocity to do the damage identified, much higher than the expected relative velocity in this case. The July 7th final test I suspect may not have been done with an unrealistic relative velocity (a push to show a cause and get it behind them). I suspect the damage may have been from a layer of ice attached to the foam rather than the foam itself. Sub-zero tank temperature, high humidity Florida weather, any shrinkage crack in the insulation, and you have ice formation. Just my pet theory.