Closer to the Sun
By Cindy Moffett
A year and a half of painstaking experimentation came down to just one day in June—a day without thunderstorms, lest a power outage ruin the entire process.
Early that morning, the UT Magnet Development Lab (MDL) team would begin the vacuum pressure impregnation of resin into a sample solenoid coil—a 14-by-40 grid of superconducting cables.
The process would go on for eight hours under carefully controlled temperature and pressure. After curing for about thirty-six hours, the glass/epoxy composite would provide electrical insulation to the cables, as well as structural strength.
But why is this prototype solenoid so important? Because it was designed to prove the concept for the massive central solenoid that will eventually drive the plasma current for the world’s most ambitious nuclear fusion facility.
The quest for nuclear fusion energy—a clean process that produces ten times the energy it consumes—began in the 1940s. The sun is a natural fusion reactor controlled by the gravitational forces of the universe, but scientists have found recreating and controlling such power on Earth to be a monumental challenge.
In 1985, an international organization comprising China, the European Union, Japan, South Korea, Russia, and the United States was formed to make fusion a reality. After many years of research, the group known as ITER recently began construction of a $16 billion tokamak reactor in southern France that is scheduled to begin operation in 2020.
A tokamak uses magnetic fields to confine the reactor fuel (a hot plasma) in a doughnut-shaped vacuum vessel. As part of the consortium, the United States will supply the central solenoid—a giant electromagnet more than 60 feet tall and weighing more than 1,000 tons—that both ignites the fuel and steers the ensuing plasma.
In order to withstand the extreme temperatures inside the tokamak, the MDL developed a combination of glass and resin to insulate and support the six independent coil packs that will make up the central solenoid. They tested various combinations of temperature and pressure to find the optimum viscosity for impregnating the cables with resin. They tested their concepts at grids of 3 by 3, then 4 by 4, slowly working their way up to the 14-by-40 grid.
Still, it would all come down to this final test. If it worked, US ITER industry partner General Atomics in San Diego would use the technique to build the central solenoid coil.
David Irick and Madhu Madhukar, both faculty members in UT’s Department of Mechanical, Aerospace, and Biomedical Engineering, have been working on ITER projects since 2008. They are responsible for some very unique manufacturing processes to support the core of the magnet system.
About half of the project focused on developing the insulation system. Laying the groundwork for their practical solutions required fundamental research, such as characterizing the properties of the insulation materials, primarily fiberglass and epoxy. However, Madhukar says they have done far more applied research than fundamental.
For instance, one effort focused on developing a way to join the ends of the 800-meter-long superconducting cable so that the joint would not significantly impede the overall superconductivity of the coil. “They must be joined in a very low-resistance way so the energy flow can be maintained,” says Irick.
“This is the first time this had been done,” says Madhukar. “These new processes work well and are in the public domain because they were developed by the government. This is basic fundamental research, creating something that didn’t exist.”
Although the MDL’s work was done in support of ITER, the ramifications are far broader. “This has application in any superconducting magnet,” Irick says. “Some solenoid superconducting magnets are used to store electrical energy and help balance the power grid.”
Prior to the final test, Irick and Madhukar assessed their plan. “We’ve had eight or nine satisfactory runs. And we had sensors in place to check the temperature. When General Atomics builds the real thing, they can’t have sensors in place, and they’ll have about $6 million at risk.”
On the big day, vapor pressure impregnation was used to transfer the epoxy through a tube to the sample solenoid, the 14-by-40 array of coils nestled in a fully insulated tank about 6 feet tall.
The impregnation took about six hours. The array was then cooked for two days according to a specific temperature-time profile. Cross sections cut into the results proved what instrumentation had already told the team: The experiment was a success.
Now Irick and Madhukar will teach the engineers at General Atomics to replicate the process so they can produce and deliver the central solenoid to ITER in 2013. Four years of fundamental and applied research have paid off.
Whether or not the ITER project ever successfully realizes the promise of fusion energy, the advances in materials and processes spawned by the MDL will be essential to solving a variety of scientific challenges for years to come.