Search

We present an electrostatic Particle-In-Cell (PIC) code on multi-block, locally structured, curvilinear meshes called Curvilinear PIC (CPIC). Multi-block meshes are essential to capture complex geometries accurately and with good mesh quality, something that would not be possible with single-block structured meshes that are often used in PIC and for which CPIC was initially developed. Despite the structured nature of the individual blocks, multi-block meshes resemble unstructured meshes in a global sense and introduce several new challenges, such as the presence of discontinuities in the mesh properties and coordinate orientation changes across adjacent blocks, and polyjunction points where an arbitrary number of blocks meet. In CPIC, these challenges have been met by an approach that features: (1) a curvilinear formulation of the PIC method: each mesh block is mapped from the physical space, where the mesh is curvilinear and arbitrarily distorted, to the logical space, where the mesh is uniform and Cartesian on the unit cube; (2) a mimetic discretization of Poisson's equation suitable for multi-block meshes; and (3) a hybrid (logical-space position/physical-space velocity), asynchronous particle mover that mitigates the performance degradation created by the necessity to track particles as they move across blocks. The numerical accuracy of CPIC was verified using two standard plasma–material interaction tests, which demonstrate good agreement with the corresponding analytic solutions. Compared to PIC codes on unstructured meshes, which have also been used for their flexibility in handling complex geometries but whose performance suffers from issues associated with data locality and indirect data access patterns, PIC codes on multi-block structured meshes may offer the best compromise for capturing complex geometries while also maintaining solution accuracy and computational efficiency.

A spectral method for the numerical solution of the multi-dimensional Vlasov-Maxwell equations is presented. The plasma distribution function is expanded in Fourier (for the spatial part) and Hermite (for the velocity part) basis functions, leading to a truncated system of ordinary differential equations for the expansion coefficients (moments) that is discretized with an implicit, second order accurate Crank-Nicolson time discretization. The discrete non-linear system is solved with a preconditioned Jacobian-Free Newton-Krylov method.It is shown analytically that the Fourier-Hermite method features exact conservation laws for total mass, momentum and energy in discrete form. Standard tests involving plasma waves and the whistler instability confirm the validity of the conservation laws numerically. The whistler instability test also shows that we can step over the fastest time scale in the system without incurring in numerical instabilities. Some preconditioning strategies are presented, showing that the number of linear iterations of the Krylov solver can be drastically reduced and a significant gain in performance can be obtained.

Three movers for the orbit integration of charged particles in a given electromagnetic field are analyzed and compared in cylindrical geometry. The classic Boris mover, which is of leap-frog type with position and velocities staggered by half time step, is connected to a second order Strang operator splitting integrator. In general the Boris mover is about 20% faster than the Strang splitting mover without sacrificing much in terms of accuracy. Furthermore, the Boris mover is second order accurate only for a very specific choice of the initial half step needed by the algorithm to get started. Unlike the case in Cartesian geometry, where any initial half step which is at least first order accurate does not compromise the second order accuracy of the method, in cylindrical geometry any attempt to use a more accurate initial half step does in fact decrease the accuracy of the scheme to first order. Through the connection with the Strang operator splitting integrator, this counter-intuitive behavior is explained by the fact that the error in the half step velocities of the Boris mover is proportional to the time step of the simulation.For the case of a uniform and static magnetic field, we discuss the leap-frog cyclotronic mover, cylindrical analogue of the cyclotronic mover of Ref. L. Patacchini and I. Hutchinson, J. Comput. Phys. 228 (7), 2009], where the step involving acceleration due to inertial forces is combined with the acceleration due to the magnetic part of the Lorentz force. The advantage of a cyclotronic mover is that the gyration of a charged particle in a magnetic field is treated analytically and therefore only the dynamics associated with the electric field needs to be resolved.

Mesh-motion (r-refinement) grid adaptivity schemes are attractive due to their potential to minimize the numerical error for a prescribed number of degrees of freedom. However, a key roadblock to a widespread deployment of this class of techniques has been the formulation of robust, reliable mesh-motion governing principles, which (1) guarantee a solution in multiple dimensions (2D and 3D), (2) avoid grid tangling (or folding of the mesh, whereby edges of a grid cell cross somewhere in the domain), and (3) can be solved effectively and efficiently. In this study, we formulate such a mesh-motion governing principle, based on volume equidistribution via Monge-Kantorovich optimization (MK). In earlier publications [1,2], the advantages of this approach with regard to these points have been demonstrated for the time-independent case. In this study, we demonstrate that Monge-Kantorovich equidistribution can in fact be used effectively in a time-stepping context, and delivers an elegant solution to the otherwise pervasive problem of grid tangling in mesh-motion approaches, without resorting to ad hoc time-dependent terms (as in moving-mesh PDEs, or MMPDEs [3,4]). We explore two distinct r-refinement implementations of MK: the direct method, where the current mesh relates to an initial, unchanging mesh, and the sequential method, where the current mesh is related to the previous one in time. We demonstrate that the direct approach is superior with regard to mesh distortion and robustness. The properties of the approach are illustrated with a hyperbolic PDE, the advection of a passive scalar, in 2D and 3D. Velocity flow fields with and without flow shear are considered. Three-dimensional grid, time-step, and nonlinear tolerance convergence studies are presented which demonstrate the optimality of the approach.

A new cell-area equidistribution method for two-dimensional grid adaptation, based on Monge-Kantorovich optimization (or Monge-Kantorovich optimal transport), is presented. The method is based on a rigorous variational principle, in which the L"2 norm of the grid displacement is minimized, constrained locally to produce a prescribed positive-definite cell volume distribution. The procedure involves solving the Monge-Ampere equation: A single, nonlinear, elliptic scalar equation with no free parameters, and with proved existence and uniqueness theorems. We show that, for sufficiently small grid displacement, this method also minimizes the mean grid-cell distortion, measured by the trace of the metric tensor. We solve the Monge-Ampere equation numerically with a Jacobian-Free Newton-Krylov method. The ellipticity property of the Monge-Ampere equation allows multigrid preconditioning techniques to be used effectively, delivering a scalable algorithm under grid refinement. Several challenging test cases demonstrate that this method produces optimal grids in which the constraint is satisfied numerically to truncation error. We also compare this method to the well known deformation method [G. Liao, D. Anderson, Appl. Anal. 44 (1992) 285]. We show that the new method achieves the desired equidistributed grid using comparable computational time, but with considerably better grid quality than the deformation method.