Disaster 02 · Engineering disasters
The foam strike that was filed as routine
A piece of foam damaged the shuttle Columbia's wing at launch, and because foam strikes had happened before, requests to photograph the damage were not followed through.
- NASA
- Over Texas, United States
- 2003
- 3 min read
On this page
Before you readWhat would you do?
A spreadsheet model your team uses to estimate delivery times gives a reassuring answer for a new order. The order is ten times larger than any order the model was built on. A colleague is uneasy. What should you do?
Pick an answer now. The file tells you whether you were right.
01The file in one minute
1Foam hit the wing
About 82 seconds after launch, foam from the fuel tank struck Columbia's left wing.
2Pictures were not taken
Engineers asked for photos of the wing in orbit. Managers, used to foam strikes, did not follow through.
3Lost on the way home
On 1 February 2003 hot gas entered the damaged wing and the shuttle broke up. Seven crew died.
4Culture as a cause
Investigators said NASA's culture, schedule pressure and reluctance to hear bad news mattered as much as the foam.
02Timeline
16 January 2003
Columbia launches. Foam from the fuel tank strikes the left wing.
17 January 2003
Engineers spot the strike on film and later ask for images of the wing.
1 February 2003
Columbia breaks up over Texas on its way home. All seven crew die.
26 August 2003
The investigation board blames the foam and NASA's organisational culture.
03What happened
Columbia launched on 16 January 2003 on a 16-day science mission. About 82 seconds after liftoff, a piece of insulating foam the size of a small suitcase broke off the external fuel tank and struck the leading edge of the left wing.
Engineers saw the strike on launch films the next day. Some of them asked for the military to photograph the wing in orbit, so they could see whether it was damaged. Those requests were not followed up by the mission managers, who saw foam strikes as a known issue that had not caused disaster before.
On 1 February 2003, as Columbia came back through the atmosphere, hot gas entered through a hole in the wing and destroyed it from inside. The shuttle broke up over Texas and all seven crew died.
The Columbia Accident Investigation Board reported in August 2003. It found that foam had come off on many earlier flights and that NASA's culture, including pressure to keep to schedule and a reluctance to hear bad news, was as much a cause as the foam.
04Why it happened
The board drew a direct line back to Challenger. Seventeen years later, a known problem had again become part of normal operations. Because foam strikes had not caused a loss before, people treated each new one as an ordinary maintenance issue.
The analysis that did take place used a computer model built for far smaller pieces of debris. Its result looked reassuring, but it was outside the range the model had been tested for. The people who doubted it lacked the standing and the data to change the managers' view, and the one thing that could have given them data, the photographs, was never requested through the right channel.
The idea behind this caseLesson: Check what AI tells you →
05Lessons for decisions
- Ask whether a tool or model was built for a case as large or unusual as yours before trusting its answer.
- Make it easy for anyone to ask for more information, and make turning such a request down a decision someone has to explain.
- Review recurring problems on their own merits, even if they have never caused harm.
- When the people closest to the problem are worried, find out what would settle the question and get it.
07 — Sources
- Remembering the Columbia STS-107 Mission · NASA · 2023
- Space Shuttle Columbia disaster · Wikipedia · 2026