8
$\begingroup$

Is there any software which is able to simulate D-T interaction for example and get temperature-crossection curve without referencing to any experimental data?

Do we have quantum-level simulation working for anything more complex than hydrogen?

$\endgroup$
  • $\begingroup$ Could you elaborate on what specifically you mean by "without referencing to any experimental data"? $\endgroup$ – probably_someone Oct 16 '18 at 14:05
1
$\begingroup$

There is no software calculating this information, mainly because the of the limits of the theory. A gas of atoms can be relatively easy to model, even in plasma state, but connecting cross-section and temperature requires an exact mathematical form of the nuclear potential, which we don't have.

A more expanded version of the answer and why this code is not possible...

Codes / software available fall mainly within two categories: modelling particles interaction (like Monte Carlo codes), or modelling continuous media. I'll refer to them as DD (discrete description) and CD (continuous description) respectively.

You might have variations on these approaches, using a different approach for some quantities. Like in DD you can have continuous quantities like continuous energy deposition, or in particle in cell codes you might have macroparticles whose velocities spectrum can be folded with cross-section spectra to get reaction rates. But still these approaches limitations are rooted in their base conceptions: DD will always scale microscopic to macroscopic quantities linearly, and CD will not properly yield reaction rates of discrete phenomena.

On the other hand, our understanding of nuclei is still vague. We think quantum mechanics should provide the correct description, if we had the correct exact form of the potential between particles. However this potential is not known exactly, and the most successful models in nuclear physics provide only partial description of some features, while being way off for other features. With this situation is not possible to make any method, much less any code/software, able to provide ab initio reliable predictions.

| cite | improve this answer | |
$\endgroup$
1
$\begingroup$

It is possible to perform first principles calculations of light ion reactions relevant to nuclear fusion (like d+t), see here. However, this kind of work is at the frontiers of computational physics and requires supercomputing resources and sophisticated codes. More phenomenological approaches (optical potentials etc) are described in text books on nuclear reaction theory, for example here.

The basic problem is indeed that even if the microscopic nucleon-nucleon interaction is known, and the Hamiltonian in the few-body sector can be diagonalized (for low-lying states), solving the time-dependent Schroedinger equation (with scattering states as in and out states) is very difficult beyond three or four nucleons. Something smaller, like $n+p\to d+\gamma$, is easily done.

| cite | improve this answer | |
$\endgroup$

Not the answer you're looking for? Browse other questions tagged or ask your own question.