Using Sound to Detect Shuttle Defects

Popular version of paper 1pNCd1
Presented Monday afternoon, October 17, 2005
ASA/NOISE-CON 2005 Meeting, Minneapolis, MN

NASA engineers have a perplexing problem, and we are using sound to help solve it.

The problem is that the spray-on foam used to insulate the space shuttle's external tanks comes loose, and pieces fall off during liftoff. That's what happened in 2003 with the space shuttle Columbia. A chunk of the foam, officially called SOFI (spray on foam insulation), hit the left wing damaging the fragile heat tiles designed to protect the shuttle during reentry.

Investigators determined that defects, including flaws such as air bubbles in SOFI, prevented the foam from being uniformly attached to the tanks. Following the Columbia disaster, NASA took steps to correct the problem. Critical areas of the external tank were redesigned, SOFI applicators received additional training, and videotape of the application process was reviewed to look for possible defects. The team is also evaluating several nondestructive evaluation (NDE) methods, some that require large and expensive equipment to check for defects.

In spite of these preventive measures, a one-pound foam slab fell from the external tanks as Discovery lifted off this past July. Fortunately Discovery wasn't damaged; but, with human life and the space program dependent on the shuttle's safety, NASA needs a reliable and efficient method to assess whether the foam is firmly attached to the tanks. Since any alterations or deformations would damage the foam, all testing must be noninvasive.

We devised a portable and inexpensive acoustic tool to measure sound absorption levels in the foam. Our goal was to determine if analysis of differences in these absorption levels could accurately identify the location and size of defects.

We were successful. We found that areas where the foam had detached absorbed more sound than areas where the bond was still good. Specifically, the peaks in the sound absorption spectrum were 25% to 50% higher in the disbonded areas.

We designed the tool and experimented with a number of variables to ensure the reliability of their tool and their testing methods. One area of concern was environmental noise. We conducted tests that verified their indicators would work as long as there is a 10 decibel difference between the background noise and sound properties of the foam. These results, as well as the portability of the equipment, offer a promising tool for NASA to be able to evaluate SOFI at various stages. They can test for defects while the foam is being sprayed on the external tank in the factory as well as do spot checks of certain locations during application or as the shuttle sits on the launch pad prior to liftoff.