Is there any particular reason why you haven't implemented something akin to Broyden's method that updates a non-symmetric Jacobian instead of a symmetric Hessian? This seems the logical thing to do since the forces are non-conservative. Is there a problem with this that I'm overlooking?
Craig
Using the Jacobian instead of the Hessian for optimization
Moderator: moderators
Re: Using the Jacobian instead of the Hessian for optimization
It sounds like you know more about these methods than we do. This sounds like a good suggestion.