Abstract.
We discuss the use of linear multistep methods for the solution of conservative (in particular Hamiltonian) problems. Despite the lack of good results concerning their behaviour, linear multistep methods have been satisfactorily used by researchers in applicative fields. We try to elucidate the requirements that a linear multistep method should fulfill in order to be suitable for the integration of such problems, relating our analysis to their use as boundary value methods. We collect a number of results obtained for the linear autonomous case and also consider some numerical tests on the Kepler problem.
Similar content being viewed by others
Author information
Authors and Affiliations
Corresponding author
Additional information
Work supported by GNCS and MIUR.
Rights and permissions
About this article
Cite this article
Iavernaro, F., Mazzia, F. & Trigiante, D. Multistep Methods for Conservative Problems. MedJM 2, 53–69 (2005). https://doi.org/10.1007/s00009-005-0030-1
Received:
Accepted:
Issue Date:
DOI: https://doi.org/10.1007/s00009-005-0030-1