Author: Manikonda, S.L.
Paper Title Page
MOPWO074 A Novel Differential Algebraic Adaptive Fast Multipole Method 1055
  • S. Abeyratne, B. Erdelyi
    Northern Illinois University, DeKalb, Illinois, USA
  • B. Erdelyi, S.L. Manikonda
    ANL, Argonne, USA
  The direct pairwise calculation of the potential/electric field created by a very large number of particles is computationally impracticable since it requires long run time and a large amount of memory. The Fast Multipole Method (FMM) is a fast algorithm which scales linearly with the number of particles and it enables highly accurate evaluation of the potentials and fields among the large number of particles using less memory compared to the direct evaluation. The FMM has two main forms, non-adaptive and adaptive. The former is suitable for uniform distributions while the latter is more efficient for non-uniform distributions typically encountered in beam physics. This paper presents an implementation of a novel 3D adaptive FMM algorithm and some results obtained from simulations performed with non-uniform particle distributions.