Postcard from nuclear fusion

As part of my first real job after university, I got sent into the wilds of Oxfordshire in the UK. It's a beautful area, and it's Fortran that got me there.

As an aside for the younger geek community, Fortran was the language of choice for any project that involved a lot of maths, and this one certainly did. This was years before PCs, let alone the sophistication of MATLAB.

In fields of Oxfordshire there was a group of three nuclear research facilities, the smallest of them, Culham Laboratory, was the home of an international project called JET to design a small scale nuclear fusion reactor. It wasn't to be a reactor in the sense that it would provide elecricity for the national grid. It was classed as an experiment, approximately one-third the size of a full scale fusion reactor, intended as a physics and engineering demonstrator.

My job was to take hastily written maths and convert it into working code. Before I got there, I'd considered myself to be a reasonable mathematician, but it took a morning to make me realise how ridiculous that view was. The physicists in particular were light years away from anything I understood. But, I knew the language well, and between us the plasma simulation code got written - we called it Icarus .

Interestingly, it was one of the best documented systems I've worked with because of the research paper that described it.

There's something called the "Q-value" which describes how well a fusion reactor is performing. A value greater than one means the reaction produces more power than it takes to create the plasma, a value of five means a self-sustaining reaction, and a full power plant needs Q=10 . JET achieved Q=0.7 in the late 1990s which was considered quite good, but it was never intended to be a power station.

Further reading:


© 2015-2016 Woodbrook-Wilson   |   Postcards   |   Archive   |   Home page