This article is taken from the monthly journal Sciences et Avenir – La Recherche #913 of March 2023.
Daniel Vanderhagen is Director of the Military Applications Direction Simulation Program at the French Atomic Energy Commission (CEA).
Sciences et Avenir: How are the experiments of the National Ignition Center (NIF) in the USA and the Laser Mégajoule (LMJ) in France being used by the military?
Daniel Vanderhagen: Since the end of nuclear testing in 1996, the United States and France have used computer simulations to ensure the reliability of their arsenals and to develop new weapons systems. They are based on a chain of equations that model the physics of a thermonuclear warhead – what we call simulation codes. And it is to continuously improve the accuracy of these codes that NIF and LMJ conduct laser fusion experiments on a much smaller scale than during a nuclear explosion, but as representative as possible.
Do these simulations require powerful computing resources?
The equations describing plasma flows during fusion or the behavior of materials under pressure of several billion atmospheres are extremely complex. To solve them, we use some of the most powerful supercomputers in the world. In particular, the Exa1 machine, which has a computing power of 36 petaflops (36 million billion operations per second, approx. ed.) and almost 900,000 processor cores, was opened in 2022 by the Minister of the Armed Forces.
Can LMJ reach, like NIF, the flammability threshold?
We have been conducting high energy density physics research since 2014 and nuclear fusion experiments since 2019. Eighty laser beams are currently in use, with eight more coming in the next few months to deliver 350 kilojoules of energy. When 176 beams are finally installed in 2026, the power will reach 1.3 megajoules. We will begin in 2030 a gradual ascent to the threshold of ignition and self-sustaining synthesis. The more we incline towards this regime, the closer we come to the combustion phenomena in the operation of the hydrogen bomb.