Scientists studied the formation of heavy elements in neutron star collisions

Astrophysicists at the Helmholtz Centre for Heavy Ion Research in Germany have succeeded in modeling the formation of heavy elements resulting from the collision of neutron stars in real-time using neural networks. This was reported by Zamin.uz.

This new method, named Rhine, opens a new era in utilizing the capabilities of artificial intelligence to study the most complex physical processes in the universe. The results of this research are considered an important step in understanding the secrets of the cosmos.

The emergence of gold, platinum, and many other heavy metals in the universe is directly linked to the collision of neutron stars and the process of rapid neutron capture. Until now, scientists have faced great difficulties in calculating this process.

The main reason for this was that even the power of the most powerful supercomputers was insufficient to simultaneously calculate the interactions of nearly three thousand different isotopes. Previously, such simulations were carried out in two stages.

First, the collision itself was modeled, and then the nuclear reactions were calculated separately. However, this approach was not without flaws, as it did not fully account for the effect of the energy released on the movement of matter.

The new method, however, integrated these two processes into a single whole using neural networks. The researchers created a complex consisting of sixteen specialized neural networks.

Instead of monitoring thousands of isotopes individually, they analyze several key physical properties of the environment, namely the proportions of neutrons, protons, and heavy nuclei. This increased the calculation speed several times, bringing the process closer to real-time.

Tests conducted using the new model showed unexpected results. It was found that when the energy released during the rapid neutron capture process is taken into account, the average velocity of the matter being ejected into space increases by forty percent, and its mass by twenty percent.

This energy helps the matter overcome the gravitational pull of the black hole, which is considered the central object. Additionally, the data obtained via the neural network allows for a more accurate prediction of the brightness of a kilonova—the flash that occurs after the collision of neutron stars.

According to calculations, ten days after the collision, such a flash is twice as bright as previously estimated. This will help astronomers obtain more accurate data when observing various cosmic phenomena through telescopes.

Currently, this new development and the neural network models have been released for open use. This will serve as a new tool for scientists worldwide to study the most mysterious processes in the universe and analyze data from future gravitational wave observatories.

This technology is a vivid example of the successful synergy between modern astrophysics and artificial intelligence, making a huge contribution to the development of science.