Author Topic: Accepting Inputs from UMATin file  (Read 4322 times)

AvinRao

  • Jr. Member
  • **
  • Posts: 10
Accepting Inputs from UMATin file
« on: July 27, 2020, 06:49:14 AM »
Dear FEAP community,

Is it ok if we take inputs through UMATin file for a user element without calling or invoking the corresponding UMATln file ?

Thanks

Prof. S. Govindjee

  • Administrator
  • FEAP Guru
  • *****
  • Posts: 1164
Re: Accepting Inputs from UMATin file
« Reply #1 on: July 27, 2020, 02:45:37 PM »
Yes that is fine.

shash_agrawal

  • New Member
  • *
  • Posts: 2
Re: Accepting Inputs from UMATin file
« Reply #2 on: July 28, 2020, 05:21:36 PM »
Hello,

I want to know about what is the primary intended purpose of uMatLn function for user elements? For a user defined element, can it be used for post processing or in element stiffness matrix development, or both?

How to control the input arguments eps/deformation gradient (F/ detF)? I mean how to take control over calling uMatLn & using its outputs dd and sigma or is it being done automatically somewhere?

Prof. S. Govindjee

  • Administrator
  • FEAP Guru
  • *****
  • Posts: 1164
Re: Accepting Inputs from UMATin file
« Reply #3 on: July 29, 2020, 11:25:48 AM »
If you are programing a user element, then there is no reason to call a umatiXX.f subroutine.

In a user element you are in charge of all material inputs under isw.eq.1.  FEAP will only do what
you tell it to do.  In particular where your element is called with isw.eq.1 you are expected to
read any data you need from the material cards.  When you are done reading the data, your user
element should exit (which hands control back over to FEAP).

See for example: http://projects.ce.berkeley.edu/feap/elmt11.f

The same follows for umatlXX.f.  In your user elements you need to evaluate your user model
yourself; there is not need to call umatlXX.f.
« Last Edit: July 29, 2020, 11:28:01 AM by Prof. S. Govindjee »

shash_agrawal

  • New Member
  • *
  • Posts: 2
Re: Accepting Inputs from UMATin file
« Reply #4 on: July 30, 2020, 01:57:22 AM »
Thanks a lot, Prof Govindjee for your kind response.