Newton and ModifiedNewton methods

Forum for OpenSees users to post questions, comments, etc. on the use of the OpenSees interpreter, OpenSees.exe

Moderators: silvia, selimgunay, Moderators

Post Reply
ftong
Posts: 10
Joined: Tue Nov 06, 2012 10:05 am
Location: University of Toronto

Newton and ModifiedNewton methods

Post by ftong »

Hi everyone,

Can someone help clarify about the options applied to the iterative algorithms of Newton and ModifiedNewton?

Newton -initial
Is it the initial tangent stiffness (associated with the undeformed structure) that will be used throughout all the integration time steps and for each iterative cycle within one time step?

Newton -initialThenCurrent
Does this mean that within the first integration time step, the initial tangent stiffness will be used, while in the rest time steps, the tangent stiffness matrix will be updated at the beginning of the step?
For this setup of Newton method, will the stiffness be updated for every single iterative cycle of an integration time step (regardless it is the first step or the rest ones)?

Which option does it take effect by default if I set "algorithm Newton"?

ModifiedNewton -initial
Is it the initial tangent stiffness (associated with the undeformed structure) that will be used throughout all the integration time steps and for each iterative cycle within one time step?

ModifiedNewton
If no "-initial" is added, it is the classic ModifiedNewton method, isn't it? That's to say the tangent stiffness will only be updated at the onset of each integration time step.

Thanks.
Felix
fmk
Site Admin
Posts: 5884
Joined: Fri Jun 11, 2004 2:33 pm
Location: UC Berkeley
Contact:

Re: Newton and ModifiedNewton methods

Post by fmk »

YES ro Newton -initial

Newton -initialThenCurrent, for each dT the first trial step will be computed using intial, all siubsequent the current

Newton just uses the current tangent, so neither of above

ModifiedNewton -initial will be the same as Newton -initial

ModifiedNewton without anything is just regular ModifiedNewton as you have stated.
ftong
Posts: 10
Joined: Tue Nov 06, 2012 10:05 am
Location: University of Toronto

Re: Newton and ModifiedNewton methods

Post by ftong »

Thanks Frank.
Post Reply