Abstract:The time-dependent Gross-Pitaevksii equation (GPE) is a nonlinear Schrödinger equation which is used in quantum physics to model the dynamics of Bose-Einstein condensates. In this work we consider numerical approximations of the GPE based on a multiscale approach known as the localized orthogonal decomposition. Combined with an energy preserving time integrator one derives a method which is of high order in space and time under mild regularity assumptions. In previous work, the method has been shown to be numerically very efficient compared to first order Lagrange FEM. In this paper, we further investigate the performance of the method and compare it with higher order Lagrange FEM. For rough problems we observe that the novel method performs very efficient and retains its high order, while the classical methods can only compete well for smooth problems.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to compare the performance of the Localized Orthogonal Decomposition (LOD) method with that of the traditional Finite Element Method (FEM) when solving the time - dependent Gross - Pitaevskii equation (GPE). Specifically, the paper focuses on how these numerical methods perform under different smoothness conditions. The GPE is a nonlinear Schrödinger equation used to describe the dynamic behavior of Bose - Einstein Condensates (BECs) in quantum physics.
### Main contributions of the paper:
1. **Method comparison**: Through numerical experiments, the paper compares the performance of the LOD - based spatial discretization method with that of the traditional spatial discretization method based on high - order Lagrangian FEM. In particular, the paper explores whether the LOD method can maintain its high - order convergence when dealing with rough problems (i.e., low - smoothness problems), while the traditional method may perform poorly due to lack of smoothness.
2. **High - order time integrator**: The paper combines an energy - conserving high - order time integrator to achieve highly accurate time discretization of the GPE. This makes the method have high - order accuracy in both time and space.
3. **Numerical results**: Through numerical experiments, the paper verifies that when dealing with smooth problems, the traditional method can compete with the LOD method; but when dealing with rough problems, the LOD method shows obvious advantages due to its robustness to low - smoothness problems.
### Mathematical model and equations:
The GPE has the following form:
\[ i\partial_t u = -\Delta u + Vu + \beta |u|^2 u \]
where \( u = u(x, t) \) is a complex - valued function, \( V = V(x) \) is a real - valued positive function representing potential energy, and \( \beta \) is a constant. The boundary condition is the homogeneous Dirichlet boundary condition, that is, \( u(\cdot, t) = 0 \) on \( \partial D \), and the initial condition is \( u(\cdot, 0) = u_0 \).
### Conclusion:
The paper shows through numerical experiments that when dealing with smooth problems, the traditional high - order FEM method can compete with the LOD method; but when dealing with rough problems, the LOD method can maintain its high - order convergence, while the traditional method performs poorly due to lack of smoothness. Therefore, for GPE simulations that need to deal with low - smoothness problems, the LOD method is a more efficient choice.