TU Zhanchun. Brownian motion with high-derivative dynamics[J]. Journal of Beijing Normal University(Natural Science), 2023, 59(6): 914-919. DOI: 10.12202/j.0476-0301.2023187
Citation: TU Zhanchun. Brownian motion with high-derivative dynamics[J]. Journal of Beijing Normal University(Natural Science), 2023, 59(6): 914-919. DOI: 10.12202/j.0476-0301.2023187

Brownian motion with high-derivative dynamics

  • Brownian motion with higher-derivative dynamics is investigated in this work. As a model, we consider a particle coupling with a heat bath consisting of harmonic oscillators. Assume that motion of particle without bath is determined by a Lagrangian L = L\left(t,x,x_1,\cdots ,x_N\right) where x_n (n = 1,2,\cdots , N) is the n-th order derivative of x with respect to time t. After integrating variables of bath, we derived a generalized Langevin equation for Brownian motion as follows:
                  \displaystyle\sum _n = 0^N\left(-\dfrac\mathrmd\mathrmdt\right)^n\dfrac\partial L\partial x_n-\mu x_1+\xi \left(t\right) = 0 ,
      where \mu represents effective constant of viscosity and \xi \left(t\right) is Gaussian noise. Note that we set x_0 = x in the above equation.
    Define p_N-1 = \dfrac\partial L\left(t,x,x_1,\cdots ,x_N\right)\partial x_N . From this equation, we can solve x_N and express it as a function x_N = \varphi (t,x,x_1,\cdots , p_N-1) . The Fokker-Planck equation corresponding to generalized Langevin equation is derived, which may be expressed as
             \dfrac\partial \rho \partial t = -\displaystyle\sum _n = 0^N-1\left\\dfrac\partial \partial x_n\left(x_n+1\rho \right)+\dfrac\partial \partial p_n\left\left(\dfrac\partial L\partial x_n-p_n-1\right)\rho \right\right\+\mu k_\mathrmBT\dfrac\partial ^2\rho \partial p_0^2 ,
                  where \rho = \rho (t,x,x_1,\cdots,x_N-1,p_0,p_1,\cdots,p_N-1) is the distribution function in phase space. T is temperature of the bath. Note that we set p_-1 = \mu x_1 and replace x_N with a function of t,x,x_1,\cdots ,p_N-1 in the above equation.
    As an example, we consider Pais-Uhlenbeck oscillator whose Lagrangian is
                L = \dfracY2x_2^2-\left(\omega _1^2+\omega _2^2\right)x_1^2+\omega _1^2\omega _2^2x^2,
                       where Y is a constant, and frequencies \omega _1,\omega _2 are independent of time. The corresponding Langevin equation and Fokker-Planck equation are
               Y\left\dfrac\mathrmd^4x\mathrmdt^4+\left(\omega _1^2+\omega _2^2\right)\dfrac\mathrmd^2x\mathrmdt^2+\omega _1^2\omega _2^2x\right-\mu \dfrac\mathrmdx\mathrmdt+\xi \left(t\right) = 0,
                    and
           \dfrac\partial \rho \partial t = -x_1\dfrac\partial \rho \partial x-\dfracp_1Y\dfrac\partial \rho \partial x_1-\left(Y\omega _1^2\omega _2^2x-\mu x_1\right)\dfrac\partial \rho \partial p_0+\leftY\left(\omega _1^2+\omega _2^2\right)x_1+p_0\right\dfrac\partial \rho \partial p_1+\mu k_\mathrmBT\dfrac\partial ^2\rho \partial p_0^2 ,
    respectively.
  • loading

Catalog

    Turn off MathJax
    Article Contents

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return
    Baidu
    map