Los Alamos, New Mexico—Behind a guard shack and warning signs on the sprawling campus of Los Alamos National Laboratory is a forested spot where scientists mimic the first moments of a nuclear detonation. Here, in the Dual-Axis Radiographic Hydrodynamic Test (DARHT) facility, they blow up models of the bowling ball–size spheres of plutonium, or “pits,” at the heart of bombs—and take x-ray pictures of the results.
In a real weapon, conventional explosives ringing an actual pit would implode the plutonium to a critical density, triggering an explosive fissile chain reaction. Its energy would drive the fusion of hydrogen isotopes in the weapon’s second stage, generating yet more neutrons that would split additional fission fuel.
This fission-fusion-fission process eats up some of the atoms’ mass and, according to E=mc2, Albert Einstein’s famous equation, releases ferocious amounts of energy. That’s why a warhead about 1 meter long can explode with the force of a megaton of TNT. If dropped on a city like Washington, D.C., it would instantly vaporize an area more than 2.5 kilometers across while crumpling buildings much farther out with its radioactive blast. It would kill nearly half a million people and injure or sicken almost as many.
DARHT’s experiments take place within a steel vessel shaped like a diving bell. The mock pits, made of dense metals such as lead, tantalum, or depleted uranium, have properties similar to plutonium—minus its tendency to fission. As the explosive charges are detonated, two perpendicular beams of x-rays document the pit’s implosion like high-speed cameras. Weapons scientists compare those pictures with classified supercomputer simulations of the bomb blasts to see how well the real and digital worlds match.
Facilities like DARHT have been important since 1992, when the Department of Energy’s (DOE’s) three weapons labs—Los Alamos, Lawrence Livermore National Laboratory, and Sandia National Laboratory—stopped full-fledged tests of nuclear weapons. By 1996, the United States had signed the Comprehensive Nuclear-Test-Ban Treaty—credited not only with stopping the environmental damage of nuclear testing, but also with disincentivizing new weapons designs.
Without tests, however, the only things ensuring that warheads work are facilities like DARHT, computer simulations from “weapons codes,” and a cache of data from the old days of nuclear testing. For relatively minor changes to old weapons—new fuses, fresh top-ups of the hydrogen isotope tritium—that has been enough. Every year, DOE’s National Nuclear Security Administration (NNSA) and the Department of Defense have certified the stockpile, an assessment that means they are convinced the weapons will work when they’re supposed to, as they’re supposed to—and not do anything when they’re not supposed to. “Because we’ve blown up so many of them, these things are incredibly reliable,” says Geoff Wilson, director of the Center for Defense Information at the Project on Government Oversight, which argues nuclear weapons spending should be reduced.
But now the stockpile is getting an overhaul, the biggest in decades. This fiscal year, NNSA has a record $22.2 billion budget. Much of the money will go to producing new plutonium pits to replace those in the arsenal and to modernizing four warheads. A fifth weapon, dubbed the W93—a submarine-launched warhead—is a new design program. “It’s really the first warhead program we’ve had since the end of the Cold War” that isn’t a life extension or modernization of an existing weapon, says Marvin Adams, NNSA’s deputy administrator for defense programs.
The work has become more urgent, with the post–Cold War calm turning stormy again. Russia has backed out of its only remaining major arms-control treaty with the United States, while making regular nuclear threats during its invasion of Ukraine. China is thought to be expanding its stockpile, while Iran and North Korea continue to bolster nuclear programs. “Everybody went to sleep for 25 years,” says Charlie Nakhleh, Los Alamos’s head of weapons physics. “I think we’re awake now.”
Wilson worries that the international dynamics and the U.S. overhaul could ultimately lead to a revival of bomb tests, bringing back their hazards and stoking a new arms race. “It is not unfathomable to me, which is scary to say.” It’s one thing to tweak weapons with a deep heritage. It’s another to infer functionality for modified weapons that have never been fully tested, he says.
Weapons physicists at the labs are confident they can improve existing weapons and design new ones without tests. Their computer simulations are vastly superior to those of the past, and experiments like DAHRT’s are more powerful. “Would you design a new Formula One car without taking it on the track? Or would you design a new Boeing jetliner without flying at first?” asks Rob Neely, Livermore’s program director for weapons simulation and computing. In the case of nuclear weapons and their plutonium pits, he says, the answer appears to be, “Actually, yes.”
As the simulations and experiments have improved, they’ve also revealed gaps in nuclear knowledge, and approximations in the codes that haven’t been updated in decades. Despite the doubts, Neely brims with confidence. “Not only will these things work, but they’re going to work better.”
SIMPLY REPLACING the bombs’ plutonium pits poses a science challenge: understanding how subtle changes affect their behavior. They aren’t easy to make, in part because plutonium, a metal only in existence since 1940, is mysterious and hard to handle. The last time anyone made pits at scale—in the 1980s at Colorado’s Rocky Flats plant—DOE’s contractor was shut down for environmental violations and forced to pay an $18.5 million fine.
This time, NNSA is splitting production between Los Alamos and the Savannah River Site in South Carolina. It has tasked them with making 80 new pits per year by 2030, a deadline NNSA admits it will not meet.
Los Alamos’s pits will be made at a facility called PF-4, a set of high-security buildings surrounded by cyclone fences with razor wire. Inside PF-4 are glovebox enclosures—radiation-shielded workstations where workers use thick gloves and peer through glass windows to manipulate the exotic metal. The lab is hiring thousands of workers, and its first pit is likely to be ready for the stockpile next year.
The gargantuan effort is motivated by a simple fact: many current pits are more than 40 years old, and plutonium behaves in confounding ways as it ages and radioactively decays. A green, fuzzy coating grows on it as its surface oxidizes. Atoms in its metallic lattice are knocked out of place as it spits out uranium isotopes. Its dimensions shift when it slips between six different solid phases. And the pits do not necessarily degrade smoothly. “We know at some point there will be a nonlinear piece,” says David Clark, director of Los Alamos’s National Security Education Center and editor of the Plutonium Handbook. “We just haven’t seen it.”
So far, the silvery spheres seem to be holding up. Internal and external assessments have vouched for their integrity, suggesting the pits could have decades of viability left. “We haven’t seen any issues,” Clark says.
But Jason, a secretive group of physicists who advise the government on national security matters, raised concerns that galvanized DOE. In a 2019 report, the group urged the agency to reestablish pit production “as expeditiously as possible” to “mitigate against potential risks posed by Pu aging.”
One might think the new pits would make it easier to certify the stockpile, by avoiding the uncertainties of aging plutonium. But they come with uncertainties of their own. The new pits won’t be twins of their predecessors, so weapons scientists will have to understand how the alterations change pit behavior. They are being manufactured using recycled and purified plutonium from old pits, not fresh material, unlike the originals. Moreover, they will be made with different processes, and in some cases designed to slightly different specifications. “If you look at a new requirement,” Adams says, “you often will find that the old pits we have available to us are really, really suboptimal.”
IN SOME WAYS, understanding the behavior of nuclear weapons has grown harder as scientists have gotten better at their jobs. The higher quality simulations enabled by ever more powerful supercomputers, for instance, have sometimes revealed new problems. This was the case with “boost physics,” or the processes at work in the first stage of a thermonuclear bomb, where fissioning plutonium triggers fusion reactions in a deuterium-tritium booster, which releases neutrons that spark more fission in the weapon’s pit.
For a long time, the simulations couldn’t reproduce what physicists saw in data from underground nuclear tests without the application of digital fudge factors. In 2006, scientists increased the simulations’ resolution. “And, lo and behold, we found obviously some interesting things that got a bunch of people scratching their heads,” Neely says. That helped spawn years of research in a program called the National Boost Initiative, which aimed to understand the fundamental physics of thermonuclear burn and to incorporate more basic physics into simulations, rather than relying on calibrations and approximations.
Pesky approximations rear their fuzzy heads throughout the weapons codes, Neely says. One is inherent to the nature of the simulations. They are all “meshed”—simulated in gridlike parts, like pixels in a digital image. Within each mesh element, physical properties are assumed to be the same. The mesh is getting more refined, but it’s still not a precise representation of reality. “You’re just able to capture better approximations,” Neely says, “but still an approximation.”
There’s also fuzziness in the physics that governs the simulations. To make the simulations run more efficiently, scientists often rely on math tricks and approximations rather than explicit, first-principles solutions.
Christopher Fryer, head of Los Alamos’s Center for Nonlinear Studies, has found that the weapons codes still contain computational tricks conjured up decades ago by Manhattan Project luminaries such as Hans Bethe and Richard Feynman. “Instead of relying on them to be the clever people, we’re going to have to be clever again,” he says.
The three U.S. Department of Energy weapons laboratories are getting billions of dollars to upgrade four weapons. A new design program, the W-93, could end up fielding the first new weapon since 1988.
| Name | UPGRADE |
|---|---|
| W93 | Will be put in service by 2040 and launched from submarines |
| W88-Alt-370 | An alteration to submarine-launched W88 weapons will replace fuze assemblies, add a lightning protector, and replace the conventional explosives. |
| W87-1 | A replacement to the land-launched W78, the W87-1 will have enhanced safety features and use insensitive explosives. |
| W80-4 | This weapon will extend the life of the air-launched W80-1. It was engineered with the Air Force, which designs the delivery systems. |
| B61-12 | A life-extension program to replace all four variants of the air-dropped B61. It will have maneuverable fins, enabling better targeting that will allow designers to reduce the yield. |
One of Bethe’s recipes, still stirred into some fusion simulations, involves the movement of charged particles. The recipe assumes that if a particle travels X distance, it loses Y energy—a kind of average scattering that isn’t always realistic, particularly in reactions that happen quickly. “It fits the data so well until you find out it doesn’t,” Fryer says. Replacing it could mean simulating each particle and its particulars—too tough a task even for the latest supercomputers. “This is why we haven’t done it,” Fryer says.
But other approximating physics could be replaced by better or truer formulas. Los Alamos theoretical physicist Mark Paris is working on a numerical approach to solving the nonlinear differential equations that pulse throughout the weapons codes. “You’re actually solving the system of equations that govern the system,” he says, “not the pastiche of physical mechanisms that are approximately derived from the system of equations.”
Simulation can’t be the only tool used to understand the bombs, however. All humans, even weapons physicists, are storytellers, Nakhleh says, and the simulations help them create confident narratives. But that can only go so far. “At some point,” Nakhleh says, “you have to step into the unknown—walk into the dark room, and see, ‘What did the experiment have to say?’”
THAT IS THE POINT of expensive, high-powered efforts like DARHT. To help illuminate the inner workings of bomb primaries, Los Alamos wants to increase the number of DARHT tests per year, currently seven, and improve its imaging abilities so it can take more x-ray pictures during any given test.
A second facility, an underground complex in Nevada called U1A, is also being revamped. It will soon be home to the Enhanced Capabilities for Subcritical Experiments (ECSE), a setup in which scientists will implode real plutonium, tiptoeing toward a chain reaction without actually triggering one. In ECSE, scientists will take x-ray pictures of scale-model pits as they collapse and investigate how neutrons behave during those crucial instants before the nuclear detonation. Because there is no nuclear explosive yield, such experiments technically adhere to the test-ban treaty.
Perhaps the most famous experimental site is Livermore’s National Ignition Facility (NIF), which focuses 192 laser beams onto a thimble-size target containing hydrogen isotopes to spark tiny fusion explosions. NIF creates temperatures and pressures that don’t exist anywhere else on Earth, says Laura Berzak Hopkins, associate program director for integrated weapons science at Livermore. “These conditions are those of astrophysical bodies—the center of Jupiter, the core of the Sun,” she says.
NIF drew headlines in 2022 when it produced more energy from the thimble than the lasers put in, a milestone relevant to civilian efforts to generate fusion power. But the achievement came about a decade and billions of dollars later than scientists first expected. And they still can’t accurately predict how much energy they’ll get out of a given fusion shot. “We don’t know the physics,” Fryer says. That physics is important for understanding the fusion components of the weapons—and also for how they would themselves hold up to a nuclear blast, a branch of research dubbed “weapon survivability.”
NO MATTER HOW GOOD the combination of theory, simulation, and experiment gets, it will probably never fully represent what happens in a weapon, Nakhleh says. “Omniscience is going to be a ways away,” he says. “The idea is to push that boundary of knowledge as far as possible.”
That knowledge isn’t only important for maintaining an arsenal. It’s also important for broadcasting to the world that the country knows the weapons will work. Nuclear deterrence—the idea that one country can prevent attacks by threatening an attack of similar magnitude—only holds up if the other country actually finds your threat credible.
In the era of explosive nuclear testing, conveying that message was simple. Other countries could pick up the seismic signal from an Earth-shaking blast half a world away. That left “no doubt in the minds of our adversaries or allies,” Adams says. Without tests, the United States has to signal confidence in a quieter way. “If you can prove to them you understand this physics well enough, you’re not bluffing,” Fryer says.
That’s one reason why the national labs also work on unclassified, fundamental science that overlaps with weapons science, subjects like star formation and supernovae. Lab scientists can publish that work, talk about it, stick it on a poster at an international conference. In that sense, Fryer says, Los Alamos’s Center for Theoretical Astrophysics “is a deterrent.”
Some still doubt, though, whether that physics-based storytelling will continue to adequately substitute for testing as the weapons overhaul progresses. “It’s all well and good for the engineers to go, ‘Boom, here’s this new warhead. We’re super sure it works,’” Wilson says. That might not be enough certainty for the military. At a certain point, Wilson says, someone might say that “a cheaper way to do this would be ‘Let’s just blow one up.’”
And there’s some political appetite for testing: Senator Tom Cotton (R–AR), for instance, has suggested the country withdraw from the test-ban treaty. In 2020, he proposed an amendment to the National Defense Authorization Act that would provide funding to prepare for potential nuclear tests. It passed the Senate, but the House of Representatives’s version of the bill prohibited such spending.
Holding off any push for testing motivates Fryer to dig deeper into the physics, he says. “For me, it comes down to ‘I don’t want to resume testing,’” he says. If the alternative is understanding the physics better, so be it.
Source: Science Mag