Is There Free Software Capable of Simulating Fusion in Large Atomic Systems?

[l0st]

In a nutshell: is there any free software, which can take as input full description of a relatively large system (20-100 light atoms + photons), simulate its evolution over a period of time, and tell the probability of fusion over that time?

If there's no such thing available, how close could one get to it with existing software?

 

[phyzguy]

I don't think anybody simulates fusion at the level of atoms. It would take a huge number of atoms to simulate even a small fluid volume, I think most simulation is done using fluid dynamics codes where you treat the plasma as a fluid characterized by a composition, temperature, pressure, magnetic field, etc. You can then write the rate of fusion as a function of the fluid variables. There are several open source MHD codes that can simulate fusion plasmas. Most of these were developed for astrophysical applications. One example is Athena.

This paper gives some details on simulations using fluid dynamics codes at NIF.

 

[raining Pete]

There are a few options for free software that can simulate the evolution of a system and determine the probability of fusion over time. One possible option is the Monte Carlo N-Particle (MCNP) code, which is designed for simulating the transport of particles through matter. This code can handle systems with up to 100,000 particles and can calculate the probability of fusion events.

Another option is the OpenMC code, which is also based on the Monte Carlo method and is specifically designed for simulating nuclear systems. It has been used to simulate fusion reactions in various systems, including light atoms and photons.

Both of these options are free and can handle relatively large systems, but they may require some programming knowledge and expertise to use effectively. As for how close one could get to simulating fusion with existing software, it ultimately depends on the specific system and parameters being simulated. However, with the right software and input parameters, it is possible to get fairly accurate results.