Fusion Capsules Can Take a Hit, Until They Suddenly Can’t

IT’S the setup for a very expensive magic trick. Aim uneven pulses of radiation at a peppercorn-sized capsule of hydrogen fuel, squeeze it until atoms fuse, and expect the whole show to still work even when the squeeze is lopsided. That is the question hanging over any future fusion power plant, which cannot promise the pristine, single-shot conditions of a laboratory.

New simulations from Lawrence Livermore National Laboratory (LLNL) show something the plant’s engineers will want to hear: yield holds essentially steady as imperfections in the implosion grow, right up until a threshold, and only then does performance collapse. There is no gentle slide toward failure, just a plateau and then a cliff.

Real fusion reactors will not get the careful, one-off treatment of a single experimental shot. A power plant based on inertial confinement fusion, the approach used at the National Ignition Facility (NIF), would need to fire fresh fuel capsules into place and hit them with lasers several times a second, up to ten times, according to LLNL physicist Timothy Johnson, who led the study. Every one of those capsules is spinning into position at speed, and every one is a fresh opportunity for the laser to miss its mark by a little.

That’s where asymmetry comes in. An uneven pulse robs energy from the hot spot at the implosion’s core, the sliver of superheated fuel where the fusion reactions actually ignite.

Using two-dimensional simulations of a capsule design scaled up from N210808, the NIF shot from 2021 that first achieved a self-sustaining, burning-plasma fusion reaction in the laboratory, a full year before NIF’s better-known 2022 net-energy-gain milestone, Johnson and his colleagues modeled a target built to burn roughly 32 percent of its fuel and produce about 30 megajoules, more than twenty times N210808’s actual output. They then degraded that design with deliberately uneven radiation, pushing the imbalance higher step by step and tracking what happened to the yield.

“Asymmetries in the implosion tend to rob energy from the hot spot,” Johnson said. “But if it’s a good implosion, taking some energy away still results in a good implosion, and you’re still going to ignite.”

The reason for that resilience turns out to be a kind of race. Every implosion has a hot spot racing to heat itself to ignition before the surrounding shell of compressed fuel rebounds and blows the whole thing apart. Asymmetry drags on both sides of that race at once. It siphons mechanical energy away from the hot spot, so there is less fuel to work with, and it lets the shell start expanding sooner, so there is less time. In the simulations, the implosion still won that race, right up to a mode-1 asymmetry of about 2.4 percent, corresponding to roughly 18 percent of the fuel’s peak kinetic energy diverted from the hot spot. Past that line, the hot spot’s ignition was delayed by around 60 trillionths of a second relative to a symmetric shot, just enough for the shell to win instead, and yield collapsed to nothing.

A separate test made the margin visible in a different way. Holding the asymmetry fixed and instead dialing down the fusion reaction rate itself in three simulation runs, at 90, 92, and 94 percent of its normal value, produced the same kind of cliff. Ninety percent of normal heating never ignited at all. 92 percent limped to a marginal 2 megajoules. 94 percent, just two percentage points higher, reached roughly 10 megajoules and kept climbing. A sliver of extra heating rate was the difference between failure and a functioning implosion.

The team quantified this margin using a metric called the generalized Lawson parameter, essentially a scorecard for hot spot quality that crosses a value of one right around the point where ignition becomes possible. In the unperturbed design it starts out roughly twice that threshold, then declines steadily as asymmetry increases, tracking the hot spot’s shrinking margin even while the yield itself stays flat. Only once the parameter nears one does the cliff actually arrive.

The Margin Has a Price

That margin, it turns out, is something a designer can dial up or down, and the price is yield. Thickening the fuel’s ice layer by 20 micrometers and its ablator shell by 24 micrometers pushed the same design’s projected output from 31 to 45 megajoules, but shrank its tolerance for asymmetry from about 18 percent of peak kinetic energy down to roughly 10 percent. More fuel, less forgiveness.

“As you’re turning on the power plant, understanding the sources of asymmetry, over time you can tighten tolerances, solve problems and then switch to a higher-yield but less robust implosion,” Johnson said. It is a strategy borrowed from any new industrial process: start conservative, learn the machine’s quirks, then push for output once the quirks are known.

Nowhere Near a Power Plant Yet

None of this means a plant could be built tomorrow. The design modeled here has an estimated target gain of only about 9, far short of the roughly 50 most proposals say a commercial inertial fusion energy (IFE) plant would need to turn a profit, and independent assessments of the field put the gain needed for a genuinely competitive power plant substantially higher still. The simulations are also two-dimensional and treat the asymmetry as fixed in time, a simplification of the shifting imperfections a real, injected, spinning target would encounter, and they model the capsule alone, without a surrounding hohlraum, so the laser energy needed to drive it is estimated rather than directly simulated. The authors note that their simulation tools have been extensively checked against smaller, real NIF shots, which is some reassurance for a design that goes well beyond anything fired so far.

Johnson and colleagues are now working to connect this robustness framework to signatures already visible in real NIF implosions, brightness patterns in neutron images consistent with fusion burn starting inside the very jets of compressed fuel this study identifies as a byproduct of asymmetry. A related study, already underway, is asking whether an implosion’s tolerance extends to a different kind of flaw entirely: the mixing of fuel with the material of its own capsule wall, part of a broader push at the lab’s Livermore Institute for Fusion Technology to close the gap between ignition and a working power plant.

FAQ