KrylovNewtonAlgorithm

KrylovNewtonAlgorithm

KrylovNewtonAlgorithm(tang_iter: str = 'current', tang_incr: str = 'current', max_dim: int = 3)

Bases: Algorithm

Krylov-Newton accelerated nonlinear solution algorithm.

KrylovNewtonAlgorithm uses a modified Newton-Raphson method combined with Krylov subspace acceleration to speed up iteration convergence.

Tcl form

algorithm KrylovNewton [-iterate <type>] [-increment <type>] [-maxDim <dim>]

Example

from femora.core.model import Model

model = Model()
algo = model.analysis.algorithm.krylovnewton(max_dim=5)

Create a Krylov-Newton solution algorithm.

Parameters:

Name	Type	Description	Default
tang_iter	str	Tangent to iterate on. Must be 'current', 'initial', or 'noTangent'.	'current'
tang_incr	str	Tangent to increment on. Must be 'current', 'initial', or 'noTangent'.	'current'
max_dim	int	Max number of subspace iterations before reforming the tangent.	3

Raises:

Type	Description
ValueError	If tang_iter or tang_incr is not a valid option.