31-5 Polymer properties: polymer tragedy (Challenger disaster 1986) O-ring, glass temperature

Uploaded by JUSTANEMONERD on 17.12.2012

ANNOUNCER 1: One. Liftoff of space shuttle Atlantis, featuring
the crest of its historic achievements in space.
JEFFREY HOFFMAN: The Challenger disaster occurred in January of '86.
Probably most of the students in this class weren't born at
that time, but it is important to understand how
traumatic it was. I'm Jeff Hoffman, professor of Aerospace Engineering
at MIT in the AeroAstro Department.
Prior to coming here, I was a NASA astronaut for 19 years,
made five flights on the space shuttle. My second flight, as it was originally scheduled,
would have been the very next flight after the Challenger
disaster. Before going into the simulator, I took a
look at the pictures from the launch pad, saw all
the icicles hanging, and said to myself, oh well, no way
they're going to launch today, it's much too cold.
Unfortunately, the managers who were making the decision
of whether or not they should launch on that day, they
didn't really appreciate the temperature sensitivity. I came out of the simulator when they began
the final T-minus 9 countdown, and we were expecting
to see the shuttle get launched and then we'd go back
into the simulator to continue our training.
ANNOUNCER 2: We have main engines start. Four, three, two, one, and liftoff.
Liftoff of the 25th space shuttle mission, and it has
cleared the towers. RICHARD COVEY: Challenger, go at throttle
up. DICK SCOBEE: Roger, go at throttle up.
ANNOUNCER 3: At 11:39 eastern, at twice the speed of sound,
the Challenger's fuselage breaks apart from the inside out.
America's space program suffers its first fatalities in flight.
JEFFREY HOFFMAN: What first went through most of our minds
is maybe it's a main engine problem. It wasn't until several weeks later that all
of the discussions about the O-ring came out.
There was a groove which the O-ring, both the primary and
secondary O-rings, fit in those grooves, and there was a
little flange and tang that they came together. And in order to stop the gas from going around
and escaping, when the initial bit of gas hit
the first primary O-ring, that pressure would push the O-ring
into the little space between the flange and the tang,
and that would make the seal.
In other words, the O-ring was not in a sealing configuration
when it was just sitting statically on the launch pad.
It has to move dynamically at the moment of ignition to be
able to make the seal. And the requirement for it to move quickly
is why it was dependent on temperature, because lower temperatures,
less flexibility and not so rapid reconfiguration of the O-ring.
MICHAEL RUBNER: My name is Michael Rubner and I'm a
professor in the Materials Science Department, and my
specialty is polymers. These engineers should have known, and probably
most likely knew, that if they drop below a certain
temperature, they were going into a temperature regime
that was going to be quite dangerous in terms of the
performance of that material.
If we think about the crystalline state and we try
to capture what's going on in terms of phase transitions
during that state, if we have polymer chains, those polymer
chains, at least segments of them, are going to be aligned
and organized into a highly ordered segments. So this is what you'd normally expect for
the crystalline state, a three-dimensionally ordered structure.
If we now take the temperature so that it's greater than the
melting transition, and we did a snapshot of the chain
organization, what you would see, in fact, is that the
chains are now highly disordered, it becomes like
spaghetti-like structure. And in fact they're able to move around and
slide past each other very easily.
That's a very clear phase transition that's easily
discernible. If we're in the amorphous state, which is
the state that is characterized by the glass transition temperature,
if I were to take a picture of the chains below
the glass transition temperature and take a picture
of the chains above the glass transition temperature, and
I were to draw them just on the board as I did here,
you wouldn't be able to tell any difference.
There would be no change in volume between these two
different states, very little change. There would be no change in the average distance
between the molecules.
So when we go to a temperature T greater than Tg, from this
perspective, you can't really see a phase transition of the
type that we see in crystalline phase transition. However, if we were able to take a movie picture
of these chains, I'd automatically be able to see
what's going on there. Here below the glass transition temperature,
the chains would be in this frozen state.
They'd be basically locked in place. When you get above the glass transition temperature
however, if this were a movie, we'd be able to see that the
chains are sliding past each other. Segments--
20, 30, 40 carbons long-- are activated and moving around
and that provides them the flexibility they need to, for
example, behave as a good O-ring. In terms of the space shuttle, I think the
glass transition temperature is an easily characterized parameter.
One can use a technique called differential scanning
calorimetry, for example, to know exactly where that
temperature is. Therefore you have a pretty good idea of what
temperatures you want to keep away from in terms of this
material transitioning from this very resilient, flexible
state, which is obviously needed for an O-ring, to
this very brittle state where the material is not going
to respond and change very well.
JEFFREY HOFFMAN: To find out that safety had been
compromised in the way it was, through a lack of
understanding and a lack of communication, among many
other things, was deeply-- it hurt us, it hurt us.