Like two tunneling teams digging a railway route through a mountain from opposite sides and aiming to meet “close enough” in the middle, groups of space shuttle accident investigators have created two very credible chains of evidence — but the critical juncture still eludes them.
From one end, investigators have dissected foam insulation on recently manufactured external fuel tanks and found foreign objects, undesirable bubbles and other fabrication flaws. They have meticulously examined previous debris-shedding events to map out the pattern of inadequate quality. Analysis of digitized imagery from the incident during Columbia’s final ascent on Jan. 16 shows a 2.5-pound (1.1-kilogram) hunk of foam tumbling off the tank and impacting just below the leading edge — not on the tiles below and behind that edge, as NASA once thought — midway out the left wing.
From the other end, investigators have used recovered pieces of that wing, radar returns from another piece shed in space, data radioed back to Earth in the spaceship’s final minutes, and the still-barely-believable good luck of a recovered onboard data recorder — all to pinpoint the existence of a small breach in that very same section of the wing’s leading edge. Through that breach poured super-hot ionized air as the Columbia descended from space on Feb. 1, and the cascading collapse of the ship’s structure is now spelled out in horrifying detail.
But at the critical juncture — the original moment when the insulation struck the wing and mortally wounded the Columbia, dooming its seven astronauts — there remains baffling obscurity and unresolved disputation.
The exact way in which the impact could have wreaked such havoc, and the exact nature of that havoc, still eludes investigators. So additional tests and analyses are now being prepared for next month.
On Thursday, NASA investigators visited the offices of the Columbia Accident Investigation Board (the CAIB, or the “Gehman Committee,” named after its chairman, retired Navy Adm. Harold Gehman). The two groups formally presented their findings and their speculations to each other. Individual members of both teams had always been in close touch, but this was the first formal bilateral conference.
Later that day, two experts from the Gehman Committee held a background briefing for journalists who had been covering the investigation. They discussed a recent analysis of data from the Columbia’s recovered flight recorder, and the CAIB released a statement specifying several areas needing more analysis.
But despite repeated questioning from the journalists present, the experts would not offer a “best scenario” for the shuttle’s destruction. When asked if NASA had offered them such a scenario, they clearly indicated they did not feel it was appropriate to answer at the time.
Sources close to the NASA investigation have provided MSNBC.com with precisely that suggested scenario which NASA experts discussed with the independent board. It is in fact very close to theories which board members have discussed on the record at previous press conferences.
The NASA presentation to the CAIB was based on several internal reports. On April 15, the team reviewed four scenarios and settled on scenario No. 2, involving leading edge damage to the reinforced carbon-carbon material between panel 5 and panel 9. The next day, the hardware forensic team laid out the case for damage to the T-seal between panels 8 and 9. On Monday, experts delivered a management-level summary of the previous two briefings, and this in turn became the basis for the presentation to the Gehman Committee.
The leading edge of each shuttle wing is armored with those curved panels, known as RCC panels. Between each of these high-temperature armor plates, to allow for thermal expansion, is a strip also made of reinforced carbon-carbon material, which is called a T-seal because its cross section is T-shaped.
NASA's foam theory
In NASA’s theory, the foam hit the T-seal between RCC panels 8 and 9, fracturing it in at least two places. Air pressure during launch held it in place, but once in orbit, it worked itself loose and drifted away.
This opened up a long narrow gap, perhaps an inch (2.5 centimeters) wide and 20 to 30 inches (50 to 75 centimeters) high. Since the gap was on black material, with a dark cavity behind it, these sources told MSNBC.com that they doubted any long-distance imagery could have detected the damage. Only close-up eyeball inspection by a spacewalking astronaut would have been able to notice the crevice, they said.
Pieces of RCC panels 8 and 9 have been recovered, and they show severe high-temperature erosion — sharpened almost to a knife edge — along the gap between them. What may be the lower stub of the T-seal was also recently identified, and it shows severe erosion along its upper edge as well.
The patterns of aluminum slag splatter on the inner surfaces of recovered nearby RCC panel fragments, and the sequence of wire bundle burn-throughs revealed on the recovered data tape, also support this sequence, sources told MSNBC.com.
“The plasma weakened the support of neighboring panels,” said one source, speaking on condition of anonymity, and panels began falling off in sequence away from the 8-to-9 boundary. “It burned through into the wheel well, and may have burned back out through the top of the wing.”
Airflow over the left wing was also severely distorted. MSNBC.com was told that the left side of the tail and the Orbital Maneuvering System pod on the left side of the shuttle show scorch marks. The progressive disintegration of the wing created increasing air drag, which ultimately pulled the shuttle into the sharp tumble that led to its breakup.
'Connection remains unclear'
This consensus, sources said, was not a “conclusion” but merely the best-guess scenario so far. “We’re in violent agreement,” a senior CAIB source joked during the background briefing, which was conducted on the condition that he would not be named.
However, the source reminded journalists that only last week, the leading scenario involved the in-space loss of a tiled access panel (the “carrier panel”) below and behind RCC panel 6 — and identifiable pieces of that very panel showed up the same day the theory was disclosed.
Even if the debris impact is identified as the precipitating cause of the wing breach, just how it did so remains unclear. MSNBC.com has learned that some NASA analysts still insist the white spray seen departing the impact point is not pulverized insulation at all, but scraped-off tile fragments from severe damage behind the leading-edge impact.
“The connection remains unclear,” journalists were told during the Gehman Committee’s background briefing. “We need the high-impact testing [scheduled for mid-May], and there is a lot of disagreement over just how strong the RCC material really is.”
One of MSNBC.com’s NASA sources agreed: “What’s going to be harder is to decide on what the contributing causes are.”
Some data still don’t fit even this consensus scenario. “There are some puzzling high temperatures way up front,” on Columbia’s left fuselage, a CAIB source said. “It’s very difficult to fit it all into a working hypothesis that covers all the facts.”
Nevertheless, he was optimistic: “We’re close to having one,” he told the press briefing.
“It’s going to take the tests we already have lined up,” his associate added, “so that I’m willing to say that the foam caused the breach.”
Only when the full cause-and-effect chain is completed — when the two separate tunnels join in the middle of the mountain — will NASA deserve to have the confidence to believe that its already-in-work fixes to the shuttle hardware will prevent future recurrences of this tragedy. The connection may be near, but it is still not yet here.
James Oberg, space analyst for NBC News, spent 22 years at the Johnson Space Center as a Mission Control operator and an orbital designer.