25 years ago, NASA almost a lost a Shuttle • The Register

Twenty-five years ago, Space Shuttle Columbia launched the Chandra X-ray observatory and nearly ended in catastrophe. As the then-ascent flight director John Shannon observed: “Yikes. We don’t need another one of those.”

Space Shuttle Columbia was launched from Kennedy Space Center’s LC-39B on the morning of July 23, 1999. Two previous launch attempts, on July 20 and 22, were scrubbed because of a faulty sensor and bad weather.

The launch was third time lucky in more ways than one.

Unknown to the Shuttle’s crew and flight controllers, Columbia contained several flaws – as do all vehicles – some of which were about to make their presence felt during the launch phase of the mission. A bit of wiring within the payload bay had chafed against a burred screw head, a single gold-plated pin was slightly loose in a deactivated Liquid Oxygen (LOX) post in the main injector of the right engine, and the main center engine had a slight bias in pressure measurements on its B channel that would only show when the engine reached full throttle.

Oh, and there was a slightly loose connection on a hydraulic pressure sensor on the right solid rocket booster (SRB).

The team was blissfully unaware of any of this.

The countdown progressed normally, and by T-3 seconds, all the engines were up and running and operating at 100 percent power.

Launch of STS-93 (pic: NASA)

A former Shuttle flight director, Wayne Hale, described the subsequent events: “Exactly when it happened is not clear, but on the right engine, the gold-plated pin from LOX post 32 in row 13 came shooting out. Just like a bullet, it went through the narrow part of the converging nozzle and flew out into the nozzle extension.”

This could have been disastrous – the LOX post had been pinned for a reason and could have failed and let LOX flow into the engine, resulting in explosion. “Failure of the LOX post was considered a CRIT 1 failure – loss of vehicle and crew ‘promptly,'” Hale wrote.

Or the nozzle extension could have failed. Another CRIT 1 failure. According to Hale, it had been calculated that if five adjacent cooling tubes in the nozzle extension were split, there would not be enough cooling and a burn through would occur. As it was, only three tubes were breached as the bullet-shaped LOX pin hit the side of the right nozzle extension.

The immediate effect was a hydrogen leak from the nozzle. It was not huge, but enough for the engine’s controller to increase the oxidizer flow, increasing the turbine temperature approximately halfway to the point where an engine would be automatically shut down.

It took the booster officer and his team around a minute to realize something was amiss with the engine – not for want of attention but because they had their hands full with another problem. Remember that loose SRB connection? It resulted in an alarm on the console. There were two hydraulic systems on each SRB. If both failed, the SRB would not be steerable. Another sudden CRIT 1 failure.

And then there was that chafed wire and the potential short circuit. As the Shuttle lifted off, the commander, Eileen Collins, called “Fuel Cell PH.”

Hale wrote that the call indicated that one of the fuel cells might be failing: “It’s the Kaboom Case, Flight.” However, although the master alarm onboard Columbia was wailing, the fuel cell had not actually failed. Instead, one of the AC buses had shorted out. The affected part of the circuit had been automatically shut down, and the erroneous alarm was caused by suddenly unpowered instrumentation.

One effect of the short was a loss of power to the Space Shuttle Main Engine (SSME) controllers.

According to Hale: “The A computer on the Center SSME lost power, never to be recovered. The B computer (DCU B) immediately took control and the engine ran on normally.”

Except it wasn’t running normally. The A channel pressure transducer dropped offline, meaning that the B computer only had the B transducer, which was reading slightly high – in this case, 12 psi high. Automatically, the B computer throttled back the center engine. Not hugely, but enough to partially offset the shortfall of LOX caused by the nozzle leak on the right engine.

The engine had lost its B computer, but the A computer continued working, and the engine, with the leak, carried on running.

“How lucky we were,” Hale said. “Instead of being 200 or more fps short at MECO, possibly leading to an abort landing or requiring two tons of OMS propellant to make up, we wound up being only 15 fps short, well within the capability of the OMS budget.”

The mission itself was successful, and the Chandra X-ray observatory, which is now on NASA’s budget chopping block, was deployed. Columbia’s next mission would be STS-109 to service the Hubble Space Telescope.

As for the issues seen during the launch, NASA noted that the wiring problem was likely caused by workers “inadvertently stepping on it,” and the problem had likely been there since Columbia was manufactured. And the pin? Apparently, it had never passed any acceptance testing. STS-93 was the last flight of that generation of SSMEs.

“The next upgrade to the SSMEs was to build a more robust channel wall nozzle extension,” Hale said. “The shuttle program ended before that was done.”

While Shannon’s “yikes” will go down in spaceflight history, we’ll leave the last word to Hale.

“Be prepared. Spacecraft are complex and can fail in complex ways. Never, ever let your guard down. Practice for disaster all the time.

“And remember: Murphy does not play by the rules.” ®