Kernel aggregated fast multipole method

Many different simulation methods for Stokes flow problems involve a common computationally intense task -- the summation of kernel function over $O(N^2)$ pairs points. One popular technique is Kernel Independent Fast Multipole Method (KIFMM), which constructs spatial adaptive octree all points and places small number equivalent multipole local around each box, completes sum with $O(N)$ cost, using these Simpler kernels can be used between to improve efficiency KIFMM. Here we present further extensions applications this idea, enable efficient summations flexible boundary conditions various kernels. We call our method Aggregated (KAFMM), because it uses functions at stages traversal. have implemented as an open-source software library STKFMM based on high performance PVFMM, support Laplace kernels, Stokeslet, regularized Rotne-Prager-Yamakawa (RPY) tensor, double-layer traction operators. Open periodic are supported no-slip wall condition Stokeslet RPY tensor. The package designed ready-to-use well being readily extensible additional