Info Recovered From Scorched Hard Drive Helps Fill In Blanks

When the shuttle Columbia tragically broke apart on reentry in 2003, killing its crew of seven, few believed that much, if anything could be recovered. The intense heat of reentry combined with a massive debris field over the Texas and Arkansas border made the chances recovering anything nearly impossible.

Five years later, the impossible was made possible when scientific data was successfully recovered from a hard drive that barely survived the disaster.

"When we got it, it was two hunks of metal stuck together. We couldn't even tell it was a hard drive. It was burned and the edges were melted," said Jon Edwards, an engineer at Kroll Ontrack Inc., a data recovery company to the Associated Press. "It looked pretty bad at first glance, but we always give it a shot."

The drive was sent to Kroll Ontrack six months after it was found in Texas like most of the Columbia debris. The company is used to recovering information from computers involved in fires, floods, or other disasters... but the drive from Columbia presented a unique challenge.

The drive's metal and plastic elements were scorched, and the seal on the side that keeps out dirt and dust had melted. With it gone, the drive was vulnerable to scratches and damage from particles entering the case and destroying the ability of the drive to retain data.

Luckily the core elements of the 340mb drive, the metal disks that store data, were not warped. There was damage to certain parts of the disks, but most occurring in areas where data had not been stored since it was only half full.

Additionally, since the computer used in the experiments was running the comparatively-ancient DOS operating system, data was easier to recover. Unlike more modern operating systems, DOS doesn't scatter data over multiple sectors in a drive.

After cleaning and reinstallation into a newly built drive, Ontrack's Edwards was able to recover 99 percent of the data within two days, start to finish.

Proving how difficult the recovery process was, Edwards remarks he attempted a similar process on two other drives recovered from Columbia, but they were damaged beyond recovery by the inferno of the reentry.

The data was from an experiment on the properties of liquid xenon. Though most of the data from the experiment was radioed to Earth during Columbia's final voyage, the data within the drive was critical to fill in the final details. As a result, researchers were able to publish the experiment in the April issue of science journal, Physical Review E.

FMI: www.science.nasa.gov, www.krollontrack.com