Believe it or not, the Challenger disaster happened 30 years ago last month. It’s another one of those “where were you when?” moments. I’d decided not to get up early to watch the launch covered live on TV. A 3am wake-time seemed a bit extreme. This was now the 25th shuttle launch, and these launches were getting to be very quotidian. The space shuttle was proving to be the “space truck” that NASA had promised. NASA estimated a risk of 1 in 100,000 of a launch failure with this “safe and routine” technology. That’s why everyone thought it was safe to put a teacher on board this flight as part of the “Teacher in Space Program” (or “TISP” as NASA had inevitably called it).
Of course, when I woke up the news was very different. The shuttle had exploded. And after the national mourning, the Rogers Commission was set up and went to work. And what it found was that key people knew the launch was a big risk. One engineer writing in his diary the night before the launch: “I sincerely hope this launch does not result in a catastrophe”. And yet as an organisation NASA – an organisation dedicated to the safety of its people – still made the decision to launch. What had gone wrong?
“I sincerely hope this launch does not result in catastrophe”
This week we’ll look at the physical reasons the Challenger failed. Next week, we’ll look at what went wrong in the thinking and conversations leading up to the launch. That’s much more interesting from my viewpoint, but to understand the thinking we need to first understand about O Rings on solid fuel rocket boosters and why they don’t work when it’s cold.
What went wrong – physically
The Challenger exploded 73 seconds after launch. One of the solid fuel rocket boosters burned through one of its joints and into the external liquid hydrogen fuel tank of the main engine, causing the massive explosion. The picture below shows the two solid fuel rocket boosters (they look like white rockets) attached to the orange external liquid hydrogen fuel tank. The whole assembly is attached to the underside of the shuttle:
You can see that if there is a leak in either of the solid fuel rocket boosters, you’re going to have a problem sitting next to all that liquid hydrogen in the orange external fuel tank.
And that was the problem. The solid fuel rocket boosters weren’t made of one piece with no joins. Instead they were made of short cylinders that were then joined up in Florida to make the booster. The diagram below shows the make-up of the boosters (you can see the different segments in the “exploded” booster on the right):
Why weren’t the boosters made of one piece of metal, so there weren’t any joins that could be a weakness? Well, that depends on who you read. It was cheaper to make them in small pieces and then fit them together. But I have also read that Thiokol, the contractor, needed to make the boosters in Utah for political reasons. They then had to be transported to Florida, where the launch site was. So they had to be made in separate pieces to allow for transport before final assembly in Florida. A single piece booster (with no joins) would have been too large to transport from Utah to Florida.
So how did Thiokol plan to make the joins (the “field joints”) in the diagram above, secure against leakage. Remember, a solid fuel rocket is exactly what it sounds. It’s a rocket with a solid fuel inside (in this case aluminium mixed with other stuff). Once you light the solid fuel it’s just going to burn, you can’t turn it off. Once it’s lit it’s going to keep burning until all the solid fuel is used up. So, Thiokol needed to make sure that none of this burning fuel was going to leak out through the joints and go anywhere near the liquid hydrogen in the external fuel tank.
So they sealed the joints with large O Rings, with a circumference of 11.6 m (also shown in the diagram above). Two of them. Those O Rings had to fit into the grooves in the field joint to create a seal:
The above diagram from Anthony Cutler shows a cross section of the joint in the booster. The O Rings are shown in cross section and appear as two black dots. As the pressure built up from the burning of the fuel the joint would flex. It was the O Rings’ job to flex with the joint and to “seat” so that burning gases couldn’t rush round them and out around the joint.
The problem was, the launch was the coldest one ever (the temperature that morning got as low as 19 degrees F – a full 34 degrees lower than the previous coldest launch). And the problem with the cold is that it stiffened the O Rings. And the problem then was that one set of O Rings on the right booster didn’t “seat” quickly enough. And therefore the burning gases escaped and hit the liquid hydrogen. If you want to see Richard Feynman showing how O Rings stiffen when cold, have a look at this famous video from the Roger’s Commission.
This was not a new problem for NASA. Before the launch of Challenger on that day in January 1986, they had been getting plenty of warning about burn-through in O Rings. They knew it was a problem. In fact the engineers at Thiokol recommended against launching in such cold weather. So, why did NASA launch?
Next week – what went wrong with the thinking? – a fateful teleconference
So why did NASA launch when they knew there was a risk? It seems that something happened in the pre-launch teleconference between Thiokol and the Marshall Space Flight Centre. Thiokol, as the contractor that had built the boosters, needed to give its OK to launch – this was standard practice with all the contractors for every launch. Thiokol’s engineers were dead against the launch. In fact after the teleconference, late on the final night, Roger Boisjoly made that prescient note in his private diary: “I sincerely hope this launch does not result in a catastrophe”. And yet Thiokol had given its OK.
What went on in that teleconference? That’s next week’s discussion – how can an organisation’s thinking and conversations get so mucked up?