Nuclear fusion, computer simulations, and the life not lived
The news out of Lawrence Livermore that the National Ignition Facility (NIF) has achieved a net positive energy output from nuclear fusion has been big, and rightly so. While it is not at the point where we could extract energy from it because it doesn’t take into account the energy efficiency of the equipment (only the energy that actually made it into the laser) it is far closer than we have ever come to producing free, sustainable energy from light elements. It is a star on Earth. Really, more than a star since stars are a very slow, lower temperature form of fusion, sort of like the difference between a compost pile and a gasoline fire.
There have been a lot of great articles written about the result, and I’m not sure I want to outdo them here. Instead, I want to talk more about my own personal journey, why this result is bittersweet for me, and a little bit about what the NIF’s purpose actually was and how it has grown from there.
About 8 1/2 years ago, in the Summer of 2014, I had a job offer in my hands from the NIF (really LLNL). I was ready to leave my then employer, and I had applied to several jobs that seemed interesting to me. I went through the lengthy interview process at LLNL and was given a nice offer. At the same time, out of blue I received an offer to interview at my current employer, Georgia Tech Research Institute. I had never applied to GTRI and I ended up at first getting rejected when I interviewed with them. My resume was passed to a different lab there, however, and I received an offer.
So, I had two offers on the table and I was a bit torn between them. The LLNL job was more interesting and more in line with what I wanted to do, but career-wise I thought working for Georgia Tech might offer more freedom. I was pretty naive at the time about research jobs in general and so I reached out to mentors for help. I thought that I would get more of a pros and cons list back, but I was warned away from working at a national lab at all.
I don’t think this warning had much to do with the actual science going on there. This wasn’t long after several national labs, including LLNL, had been rocked by mismanagement and completely reorganized with massive layoffs and a new oversight structure. So the warning was more along the lines of the feeling in the physics community that they had been treated badly.
This was in contrast to my meeting several physicists at LLNL of course who were absolutely in love with what they were doing.
I made the choice to turn down the LLNL offer much to their surprise. It was a competitive position with a good salary (relative academia). I would have been using the Sequoia supercomputer, one of the fastest in the world at the time, and I had made the case that it was well in line with my interests. My job would have been to simulate the experiments being carried out at the NIF.
I had worked in fluids, plasmas, and the like and published a monograph on my thesis subject, vortex filaments. In particular, I had published some work on magnetohydrodynamics, basically fluids that are electrically charged.
At very high temperatures or low pressures, electrons leave their host atoms and begin to migrate freely. The atomic nuclei become ions. This state of matter is called a plasma. Low temperature plasmas exist in the transition between Earth’s atmosphere and space where density of particles is very low and the bombardment from the solar wind creates beautiful displays at the poles called the Auroras. High temperature plasmas exist within the Sun where intense gravity causes nuclear fusion.
Magnetohydrodynamics is important in any fusion reaction, of course, because all of them require plasma. (So-called cold fusion is not known to be physically possible.) The key to fusion is to bring the atomic nuclei, the ions, together. In theory this shouldn’t be too hard to do using magnetic fields. You just confine them magnetically so strongly that they start fusing.
Under a confining magnetic field, the plasma organizes into columns of electrons and ions. While the ions are slow moving, the electron columns fluctuate rapidly, slamming into each other and generally causing chaos. This turbulence, as well as other sources of instability, is one reason why magnetic confinement based fusion like the ITER in Europe cannot sustain a reaction for long.
Another method of achieving fusion is simply to hit a pellet of material very hard from all sides. This is called inertial confinement fusion where, instead of magnets bringing the material together in a very hot ring, you push the atoms together with a strong force. This is the kind of fusion going on at LLNL. The amount of inertia is so high that they use a powerful laser to do it. Lasers produce light of course, but light has momentum just like matter. When light hits an atom, it confers its momentum onto the atom. This is how solar sails work and, if you have enough light concentrated on a small enough space, you can push atomic nuclei so close to each other that they fuse.
The NIF was actually expecting to achieve this milestone years ago, but they ran afoul of fluid dynamics. What they do is slam their X-ray laser into a target capsule which “is a deuterium-tritium (DT) sphere encased in an ablative shell.” This laser is supposed to squeeze the fuel and cause it to “ignite” releasing fusion energy. The NIF was built in such a way to achieve this with room to spare.
The problem was that their computer sims used to build the NIF showed that they would achieve ignition, but instead they ran into instabilities in the plasma fluid dynamics that the codes didn’t account for.
The NIF was actually built to test these computer simulations because those sims were being used in nuclear stockpile management. This is one of the biggest goals of the three big national labs, LLNL, Los Alamos, and Sandia and the reason the NIF was actually built.
So, yes, it is a military project, and, yes, America’s biggest fusion project was actually built to test software.
NIF showed those codes were wrong.
A big achievement early in the NIF project was to improve the software to take into account the instabilities in the fuel. These are, to use the technical jargon, Rayleigh-Taylor instabilities. Turbulence is the bane of engineering, except when you want to mix something.
This kind of turbulence happens between two fluids of different densities. They look like this:
See all those vortices and pockets carrying all that lovely energy away? That is the essential problem of turbulence. When you want energy to be directed in a particular way, turbulence comes in and steals it.
This happened because the ablative shell was mixing with the fuel itself, causing a loss of efficiency. The NIF had to work out mechanisms to dampen the instability which they did leading to their biggest successes, and their recent success is, likewise, coming from further damping the instabilities.
Oddly enough, NIF’s primary goal, to test computer simulation codes for nuclear stockpile management, was achieved early on because it showed these codes were wrong and led to their improvement. Yet, LLNL wanted more and spread the idea widely that they were trying to achieve sustainable nuclear fusion. That was never the intent of the facility and has led to some bad press for them over the years, but I’m glad it has finally become so. America needs serious investment in fusion energy.
These days my physics interests are far more in the realm of quantum theory, gravitation, and philosophy of science. While my physics work is technically part of my job as Georgia Tech faculty, it is unfunded, and my funding ensures that my days of writing code to simulate plasmas is far behind me.
That’s why, while I celebrate the NIF’s achievement, it is bittersweet to realize that I could have been part of it and now only watch it from afar.