ENGR 482 Lecture 9

« previous | Monday, September 23, 2013 | next »

Challenger Disaster

O-Ring seals on Solid Rocket Booster

Recommendation for no-launch:

• leaking o-ring seals in past launches
• seal function depends on seal temperature
• predicted cold weather at time of launch

BUT: "take off your engineering hat and put on your management hat"

Viewpoint should have been: "We don't launch until you're absolutely sure nothing will go wrong."

Decision makers were

• Unaware of history of o-rings
• Unaware of initial recommendation of contractor advising against launch at temperatures below 53°F
• Unaware of continuing opposition of engineers at Thiokol after management reversed its position
• No clear understanding of Rockwell's concern regarding ice on the pad

The decision would have likely been reversed had the decisino makers known all of this information.

Normalization of Deviance

"The same risk was flown before without failure is often accepted as an argument for the safety of accepting it again. Because of this, obvious weaknesses are accepted again and again," until you get (un)lucky

Proper Engineering Decision

Involves technical matters that require engineering expertise Judgment or involve ethical standards in engineering codes, esp. with respect to protecting the public.

Proper Management Decision

Factors affecting well-being of general public.

Lessons

• Know and understand limitations of equipment or product
• Learn to recognize external pressures or conflicting interests cause deviations from usual process.
• Be wayr of incremmentaly increasing risks by normalization of deviance. Know past problems and anomalies
• Be mindful of big picture when worrying about details.
• Understand basic process used to make decision; challenge assumptions and basis for analysis
• learn to differentiate between Proper Engineering Decision (PED) and Proper Management Decision (PMD)
• Become effective communicators in support of your decisions. Don't be afraid to speak up or ask questions.

Columbia Disaster

• Foam strike detected in launch videos on day 2
• Engineers requested inspection by crew (unscheduled spacewalk) or remote photo imagery to check for damage
• Foam strikes had been observed before, but the damage would be analyzed upon return.

Everyone was asking the right questions whether a foam strike contained sufficient kinetic energy to cause a burn-through, but they didn't follow-through.

Result of Damage

5000° plasma enters damaged RCC, sensors show damage progression, and the shuttle broke up during reentry.

Similarities and Summary

Someone failed to prove it was not safe to launch... Only needed to say that the launch could not be proven safe.

Groupthink is not a good idea.

What NASA did not do:

1. Maintain sense of vulnerability
2. Combat normalization of deviance
3. Establish imperative for safety
4. Perform valid/timely hazard/risk assesements
5. ...

Too much information condensed onto too few slides: info is lost, and remaining info can be misinterpreted.

Emphasis on production over safety

Primary Causes of Engineering Disasters

• Human factors (ethical and accidents)
• Design flaws
• Material failures
• Extreme conditions
• Combination of all of the above

Exam Review

There is a lot of material to study. Look at review.