The best explanation for a Solid Rocket Motor Joint I could find came from The Online Ethics Center. It gives us a colorized diagram of all components. The “pink is the tang, which joins the clevis, colored orange. 177 huge steel pins (yellow) hold the joint in place. The O-rings shield the joint from 5800-degree gases inside the booster.

On the left scenario, hot gases (red arrows) are shielded from the joint by the zinc-chromate putty. On the right, immense pressure creates a blowhole in the putty, allowing the O-rings to move into the positions needed to seal the joint as the gap between tang and clevis expands. Through the blowhole, gases penetrate and wear away the O-rings.”

As seen in the diagram the Solid Rocket Booster (SRB) has two O-rings, working in series. This means that when the primary O-Ring fails the Secondary O-ring will protect the system from the outside, increasing the reliability of the system. Or at least that was the intent of the design; which worked during testing. The main problem was that all testing was conducted at warmer temperatures between 65 – 80°F and the launch day was about 50 degrees colder than the temperature during testing.

In preparation to the Challenger mission testing on the SRB system was performed and it was noted during testing that the O-rings eroded to an extent. It was also noted that the erosion was not to the point of failure, therefore NASA decided the risk was minimal. The problem with this approach was that the erosion noted during test was, as stated above, at temperatures higher that launch the ambient temperature. Temperatures as the one experienced at launch caused the O-rings to contract further compromising their sealing value.

With a good Risk Board NASA might have studied the O-Ring failure mode and extrapolate the behavior of the O-Rings at the ambient temperature of the launch. The graph below shows the predicted extrapolation (black curve) of the probability of failure for the lower temperatures at the time of launch and the confidence bands on that extrapolation (red curves).

The above graph would have shown that the probability of failure at low temperatures was quite high. On most programs of this nature a Risk Management Board will not accept a system with reliability lower than 97%. We can see here that the probability of success in the worst case scenario is zero. I am confident that no one at NASA or any company would have sent a shuttle with a crew to space if they would have seen how high the risks were, not even to keep up with the schedule.

After this disaster an investigation was conducted and the following recommendations from the report are some of the changes implemented by NASA: a) Redesign and recertification of the SRB. The redesign effort added an extra O-ring to the joints between the SRB segments which strengthened the physical connections between these segments. Heaters were also added to the joints between the SRB segments to prevent cold weather from affecting the sealing capability of the O-rings.
This would have helped with the contraction to the O-Ring caused by lower temperatures during the Challenger launch.

b) Numerous hardware, software and safety improvements were incorporated into the Space Shuttle. A crew escape system was added to allow astronauts to parachute from the Space Shuttle in certain conditions. Also, astronauts were required to wear pressurized flight safety suits during launch and landing operations.
This is a great safety measure for future flights. Not sure that it would have helped the Challenger crew, but it is good that NASA was thinking about crew’s safety in general.

c) A good risk identification and reduction programs were applied to all Space Shuttle operations. Also, all contractors’ quality control work forces were strengthened.
In my opinion this is one of the most important changes for NASA. A good risk management board would have assigned the correct weight to the O-Ring risk, instead of deeming it acceptable.
A good contractor’s quality control is always good. In the case of the Challenger the contractor built the O-Ring to specification. This change incorporated by NASA should have also included a better communication between the contractor and NASA. It was noted that the O-Ring contractor notified NASA the day before launch that even with the part built to print the O-Ring might fail under the conditions at launch.

d) Reorganization of the Space Shuttle program, placing experienced astronauts, where possible, in management positions. This new structure allowed for a better flow of information and provided the expertise needed to make key decision prior to launch.
Experience in key positions is always important. It is great to be book smart, but it is very important to have the “street smarts” necessary to take data provided and decide what more data we need. In the case of the challenger “how would the system is affected when it is launch 50 degrees lower than tested values?”

e) All prior flight safety waivers were revoked and stricter rules were put in place for environmental conditions during launch.
This was an important decision, putting safety of the equipment and personnel over schedule. If NASA didn’t have any test data in windy, cloudy and/or rainy conditions, to prove with high confidence the safety of their crew a risk of this nature should never be allowed. Canceling waivers and creating strict criteria will give NASA a guideline into what’s really safe.

f) If technical issues arise during preparation for a mission an independent government agency would review and rely information back to NASA.
During the Challenger mission the O-Ring was noticed as a technical issue but it was deemed acceptable to flight. There is a possibility that an independent agency would have come to a different conclusion.

g) A series of open reviews were introduced to discuss all significant and outstanding issues prior to a particular Space Shuttle mission. During this reviews NASA will have personnel from different areas like engineering, safety, and contractor. i.e CDR.

h) A mechanism was put into place that would allow NASA and contractor personnel to provide open and anonymous reporting of Space Shuttle safety concerns without fear of reprisal.
During the Challenger mission engineers notified the PM of risks to the joints that rendered them unsuitable for use. Would one of the engineers report this and escalate the issue? Maybe all it took was one person screaming loud enough without fear if repercussion to stop the mission.

In addition to the changes above, NASA made important changes to their “schedule driven” culture by adopting a more relaxed launch schedule and stop pursuing the installation of commercial and military payloads onto the Space Shuttle. Prior to the Challenger disaster, NASA had a pressure to launch the payloads on time, creating pressure on their personnel to quicken Space Shuttle missions, in some cases, compromising safety and/or quality.

Due to the decision to stop installing military payloads on their shuttles, NASA closed their launch pad at Vandenberg Air Force Base (VAF) that was scheduled to launch SLC-6 and host a program which contained multiple orbit military missions. After canceling the launch of SLC-6, several flaws were uncovered that would cost up to $1 billion and two years of work to repair. This revelation cemented NASA’s decision of not perusing any long-term military payloads.

References and Sources
Missed Opportunities and Graphical Failures Retrieved from (http://www.datavis.ca/gallery/missed.php)

What Went Wrong, Online Ethics Center Retrieved from (http://www.onlineethics.org/cms/12707.aspx

Cliff Lethbridge. The Challenger Legacy Retrieved from http://www.spaceline.org/challenger.html

Raheja, D. G., Alloco, M. (2006). Assurance Technologies Principles and Practices: A
Product, Process, and System Safety Perspective: Second Edition. Hoboken, NJ:
John Wiley & Sons Inc.

Rogers Commission (1987). Implementation of the recommendations of the presidential commission on the space shuttle challenger accident. Retrieved from
http://history.nasa.gov/rogersrep/v6index.htm