A important question

[snowden]

Dr Jorgen:

Hi, Dr Jorgen.
It has been half one year since I started to learn ABAQUS umat. In fact, I am more confused with ABQUS-umat algorithm.

It is known Abaqus Standard uses implicit algorithm but Explicit uses explicit algorithm.

(1) If that means only implicit stress update algorithm can be used in
Abaqus Standard-umat, and explicit update algorithm in Explicit-
VUMAT?

Recently, I finished writing a UMAT(elastic-viscoplastic constitutive model) using explicit update algorithm. To my surprise, It can also runs.

So I want to know what is the matter? If it is necessary that the implicit algorithm should be consistant with Standard and explicit algorithm with Explicit? If it is true, why?

thank you very much!

[Jorgen]

Hello Snowden,

When you say implicit and explicit stress update algorithms I assume that you are referring to a constitutite equation that is written in rate form and you are attempting to take a time increment using implicit and explicit ODE steppers. Is this what you had in mind?

You can use an explicit or implicit ODE solution approach for both UMATs and VUMATs. The only real difference between an UMAT and a VUMAT is that the VUMAT does not require you to calculate the Jacobian.

One of the benefits, however, of an implicit UMAT is that is can be used to take longer time increments since the solution is always in equilibrium. To acheive large time-increments using a UMAT you might have to use a more sophisticated (higher order) integration scheme.

Jorgen

[snowden]

Dr Jorgen

Thank you very much!

That's what I want.

Recently, I wrote a UMAT (elastic-viscoplastic model) using explicit alrorithm. That means stress tensor is updated directly. No local newton iteration is used in it. So I turn to you!

By the way, Is local newton iteration also the difference between Explicit and Implicit algorithm?? :roll: :roll:

Thank you very much!

[Jorgen]

There are two places Newton iterations can be used:
:arrow: The FE program uses Newton iterations when solving implicit simulations. You do not need to care too much about this when writing your UMAT/VUMAT.
:arrow: The UMAT/VUMAT subroutine can internally use a newton iteration if the usermaterial model is formulated as an implicit equation.

Jorgen

[snowden]

I see.

Thank you very very much! I am appreciated that you are always glad to answer my questions.

Dr Jorgen, did you write umats considering the coupling with pore water pressure? I mean as for geotechnical problem effective stress analysis is always related to this water-soil coupling problem.

As far as I know, I have seen reports involving this issue. I think it may be very difficult.