Nazanin De's Blog

Projects, Ideas and thoughts

Importance of engineering models in STS-107 investigation

On February 1, 2003, 18.9 seconds after the lunch of Colombia Space Shuttle, a piece of foam nearly at a size of a briefcase struck the left wing leading edge. The foam was originated from the left bipod ramp area of the external tank and hit the RCC panel 8 on the edge of the left wing. This wasn't the first time that foam pieces from the external getting released during lunch. NASA engineers were concerned and requested to take pictures of the shuttle while it is in the space, but their request rejected by managers since they assumed it is not necessary as this incident happened before and have not resulted in any serious issue. So, it postponed maintaining phase for a checkup. Unfortunately, STS-107 and its crew were lost during its re-entry into the Earth atmosphere due to the damage allowed hot atmospheric gases to penetrate and destroy the internal wing structure, which caused the spacecraft to become unstable and break apart.

Models Used

Transport Analysis models used to determine volume and velocity of the foam. These models are mathematical since they use physic based formulas to draw the foam trajectory, velocity, and volume. The investigation used sophisticated computer models to analyze the foam impact and to help develop an impact test program. Because an exhaustive test matrix to cover all feasible impact scenarios was not practical, these models were especially important to the investigation.
These models have to be good enough to replicate real environmental aspects and parameters to be able to help investigators find the root cause of Colombia shuttle accident.

In addition, STS-107 investigation models have to get accepted by stakeholders, since those are the one should agree to the root cause and get persuaded by model results. Since physical tests built to determine the foam struck was based on these models, they had to have high fidelity so those tests can perform near a real situation where the wing was struck. Well, I think those models definitely satisfied all those requirements.

STS-107 investigation models were really helpful identifying weaknesses of the space shuttle by testing the shuttle model in simulations in many different scenarios. I guess, one reason that many of the engineers of the STS-107 team did not take the foam struck seriously was lack of experimenting is simulation near to real environment and with different energy levels and angles.

Also, models evaluated how the testing environment needs to improve and what are the requirement for better testing data. Nowadays most of the spacecraft testing prior to building happens with using reliable models inside simulations. Although it is not the real situation it can enhance the process by identifying issues in a simulator before the build.

My take away

Although modeling helped revealed the physical RC of Columbia shuttle accident but in my idea the main RC was not physical. It was the mindset of some engineers and managers on the team who assumed a piece of foam will not be able to damage a Carbon panel. Assumptions are very dangerous in science and engineering when there are not supported by actual experiments.