Técnicas tradicionais e modernas para simulação em física

Abstract

Computational physics emerged in the general context of physics as a new way of thinking and building models about nature, which takes its place not as a part, but alongside the already traditional theoretical and experimental physics. This third way has surprisingly advanced in recent decades, due to the developments of computing and the technologies associated with it. Nowadays, the use of alternative methods in physics (frequently the ones coming from computation or mathematics) in the development of methodologies for computational physics is getting common. The most common idea behind these methods is solving problems using traditional techniques adapted for high-performance architectures (such as clusters or video cards). However, space has gradually emerged for other techniques to be used in the treatment of physical systems, techniques such as bioinspired algorithms or modeling via dynamic systems. In this thesis, we will review some of the more traditional techniques in computational physics and introduce some relatively new ones. We will review the concepts of molecular modeling and micromagnetic systems, then move on to the study of dynamic systems and some bioinspired optimization algorithms. After the methodology, we will introduce some of the several works carried out in parallel to the writing of this thesis. They demonstrate the applicability of our proposed methods inside and outside the normal scope of computational physics.