<p>A torsion pendulum can probe a variety of phenomena such as photon spin or gravity, uncover the properties of spider draglines or examine stochastic thermodynamics of driven systems. Here we use such a device to probe the state of an assembly of macroscopic self propelled rod-like robots, a system that is out of equilibrium and very sensitive to the surrounding boundaries. When the persistent motion of the active particles plays an important role or when the particle assembly becomes dense, the probe dynamics shows non Markovian behaviour, i.e. memory effects, and a Gaussian but exponentially correlated noise arises. The correlation time of this active noise increases with the number of particles and undergoes a large jump when a transition to cluster formation occurs. The system allows direct tests of fluctuation theorems for second order Langevin equations with correlated noise. Effective temperatures are then obtained for a wide range of particle densities including the range where clusters dominate the dynamics.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Non-Markovian dynamics of a torsion pendulum immersed in an active bath

  • Jean François Boudet,
  • Thomas Guérin,
  • Clément Tiffon,
  • Thomas Barois,
  • Hamid Kellay

摘要

A torsion pendulum can probe a variety of phenomena such as photon spin or gravity, uncover the properties of spider draglines or examine stochastic thermodynamics of driven systems. Here we use such a device to probe the state of an assembly of macroscopic self propelled rod-like robots, a system that is out of equilibrium and very sensitive to the surrounding boundaries. When the persistent motion of the active particles plays an important role or when the particle assembly becomes dense, the probe dynamics shows non Markovian behaviour, i.e. memory effects, and a Gaussian but exponentially correlated noise arises. The correlation time of this active noise increases with the number of particles and undergoes a large jump when a transition to cluster formation occurs. The system allows direct tests of fluctuation theorems for second order Langevin equations with correlated noise. Effective temperatures are then obtained for a wide range of particle densities including the range where clusters dominate the dynamics.