Almost a year ago, we reported a potentially serious problem at ITER, the world’s largest nuclear fusion experiment. At stake was the heart of the machine: the giant central solenoid magnet (see diagram). Now, it seems that a solution may be within reach, although it is likely to cost the project more money.
The central solenoid (or “CS” to those in the know) is a 13.5-metre-high stack of six identical superconducting coils that are designed to generate a ginormous 13-tesla magnetic field. That field helps to confine hydrogen gas as it reaches fusion temperatures above 150 million °C.
Over ITER’s 20-year lifetime, the CS coil is supposed to cycle some 60,000 times, but tests at the SULTAN facility at the Paul Scherrer Institute in Villigen, Switzerland, showed that Japanese-made cable destined for the coil was cracking and failing after just 6,000 cycles.
Neil Mitchell, the head of ITER’s magnet division, says that a year later, the team now understands the problem. Mitchell believes that the original cable suffered from two issues. First, thin strands of superconducting niobium-tin and non-superconducting copper were braided together in a way that distributed currents and pressure unevenly throughout the cable bundle (more on this from Science Insider). Second, the niobium-tin strands themselves were manufactured in a way that made them unable to withstand the increased pressure. The combination of the two problems is enough to spoil the coil.
Mitchell says that there are a couple of possible solutions. First, they could use niobium-tin strands from another manufacturer who uses a different fabrication technique. Unfortunately, doing so would probably involve finding a manufacturer outside of Japan, something that nobody really wants to do because Japan is the nation responsible for the CS cable.
Alternatively, it might be possible to braid Japanese-made strands together in a way that distributes the load more evenly and solves the problem.
Additional tests in April and beyond should show whether the new Japanese cables will hold up. The cost increase could be in the range of €20 million (US$26 million), a drop in the bucket compared to ITER’s €15 billion-ish price tag. The additional work shouldn’t add any delay to the project, Mitchell says.
At present, ITER is scheduled to conduct its first experiments in late 2020.