Unwanted jumps of the strain-rate

58 views
Skip to first unread message

Martin Lesueur

unread,
Jun 13, 2017, 12:11:36 AM6/13/17
to moose-users
Hi,
I am running an extended version of the 1D spherical hollow sphere test of MOOSE where I add a constant loading rate and a J2 plasticity.
But my strain-rate has jumps that are not supposed to appear.
The jumps start at the end of elasticity (but before plasticity) and exist only at small timestep size.
I tried to debug it without success. I thought it was a problem of tolerances but I reduced all of those and nothing changed.
I also have noticed that it does not happen if I remove the plasticity material.
Anyone can help me with that ?
Cheers,

Martin
Screen Shot 2017-06-13 at 1.52.36 pm.png
hollow-sphere_spherical1D.i

Cody Permann

unread,
Jun 14, 2017, 10:46:34 AM6/14/17
to moose-users
Pinging the list for help here? Can anyone familiar with the tensor mechanics and/or J2 plasticity modeling assist?

--
You received this message because you are subscribed to the Google Groups "moose-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to moose-users...@googlegroups.com.
Visit this group at https://groups.google.com/group/moose-users.
To view this discussion on the web visit https://groups.google.com/d/msgid/moose-users/9ceb2622-4963-475e-9e83-1a1e34eca5bc%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Jason D. Hales

unread,
Jun 14, 2017, 11:07:13 AM6/14/17
to moose...@googlegroups.com
Specifically, the input file uses ComputeMultiPlasticityStress.

On Wed, Jun 14, 2017 at 8:46 AM, Cody Permann <codyp...@gmail.com> wrote:
Pinging the list for help here? Can anyone familiar with the tensor mechanics and/or J2 plasticity modeling assist?
On Mon, Jun 12, 2017 at 10:11 PM Martin Lesueur <martin....@ecl2015.ec-lyon.fr> wrote:
Hi,
I am running an extended version of the 1D spherical hollow sphere test of MOOSE where I add a constant loading rate and a J2 plasticity.
But my strain-rate has jumps that are not supposed to appear.
The jumps start at the end of elasticity (but before plasticity) and exist only at small timestep size.
I tried to debug it without success. I thought it was a problem of tolerances but I reduced all of those and nothing changed.
I also have noticed that it does not happen if I remove the plasticity material.
Anyone can help me with that ?
Cheers,

Martin

--
You received this message because you are subscribed to the Google Groups "moose-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to moose-users+unsubscribe@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "moose-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to moose-users+unsubscribe@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.



--
Jason D. Hales, PhD
Department Manager
Fuels Modeling & Simulation
Idaho National Laboratory

Andrew....@csiro.au

unread,
Jun 14, 2017, 5:48:51 PM6/14/17
to moose...@googlegroups.com
​I think i'd output all the plastic related Material properties from ComputeMultiPlasticityStress: plastic_strain; plastic_internal_parameter; plastic_yield_function; plastic_NR_iterations. Then you can more easily inspect where and why you're getting plastic failure (assuming you are). It may be, for instance, that you get failure at just a single quadpoint because of the non-infinite resolution of the finite-element mesh, and that gives you the "jumps".

Does this make sense? If not i can be more specific.

a


From: moose...@googlegroups.com <moose...@googlegroups.com> on behalf of Jason D. Hales <jason...@inl.gov>
Sent: Thursday, 15 June 2017 1:07 AM
To: moose...@googlegroups.com
Subject: Re: Unwanted jumps of the strain-rate
 

Specifically, the input file uses ComputeMultiPlasticityStress.


On Wed, Jun 14, 2017 at 8:46 AM, Cody Permann <codyp...@gmail.com> wrote:

Pinging the list for help here? Can anyone familiar with the tensor mechanics and/or J2 plasticity modeling assist?


On Mon, Jun 12, 2017 at 10:11 PM Martin Lesueur <martin....@ecl2015.ec-lyon.fr> wrote:

Hi,
I am running an extended version of the 1D spherical hollow sphere test of MOOSE where I add a constant loading rate and a J2 plasticity.
But my strain-rate has jumps that are not supposed to appear.

The jumps start at the end of elasticity (but before plasticity) and exist only at small timestep size.
I tried to debug it without success. I thought it was a problem of tolerances but I reduced all of those and nothing changed.
I also have noticed that it does not happen if I remove the plasticity material.
Anyone can help me with that ?
Cheers,


Martin
--
You received this message because you are subscribed to the Google Groups "moose-users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to moose-users...@googlegroups.com.


Visit this group at https://groups.google.com/group/moose-users.
To view this discussion on the web visit https://groups.google.com/d/msgid/moose-users/9ceb2622-4963-475e-9e83-1a1e34eca5bc%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "moose-users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to moose-users...@googlegroups.com.


For more options, visit https://groups.google.com/d/optout.

--


Jason D. Hales, PhD

Department Manager
Fuels Modeling & Simulation
Idaho National Laboratory
(208) 526-2293

--
You received this message because you are subscribed to the Google Groups "moose-users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to moose-users...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/moose-users/CALHgsLKaY9%2B%3D0RQ-UvjBDzHfRq2KeKGZxad2P%2BT-NqT7ahThZA%40mail.gmail.com.

Martin Lesueur

unread,
Jun 19, 2017, 11:06:19 PM6/19/17
to moose-users
Thanks for the advice Andy, I did output those variables, but nothing seemed abnormal.
The problem might come from the fact that I do not impose any BC on the hole.
In comparison, MOOSE problem of the hollow sphere has no displacement on the inside.
Do you think this might cause these fluctuations? I remember that in NS, MOOSE had a necessary BC for a sideset withtout BC, NoBCBC.
Cheers,

Martin

Andrew....@csiro.au

unread,
Jun 19, 2017, 11:20:45 PM6/19/17
to moose...@googlegroups.com
Well that is rather strange. I think if there is zero plasticity at every quadpoint then ComputeMultiPlasticityStress does exactly what ComputeFiniteStrainElasticStress does. Did you try to compare with that? Did you check every quadpoint?

a

Ph: +61 7 3327 4497.  Fax: +61 7 3327 4666
Queensland Centre for Advanced Technologies
PO Box 883, Kenmore, Qld, 4069 
To view this discussion on the web visit https://groups.google.com/d/msgid/moose-users/0efc40e3-5c90-4515-96c8-e79cf47a08de%40googlegroups.com.

Martin Lesueur

unread,
Jun 20, 2017, 1:18:29 AM6/20/17
to moose-users
To answer on your previous post, the jumps come from the node of the inside hole.
After refinement of this boundary, I get unchanged jumps of specific pattern, see file attached in the first post.
This is why I don't think there is a problem with "non-infinite resolution of the finite-element mesh" because I obtain mesh convergence.

Also I don't get these jumps using ComputeFiniteStrainElasticStress.

Andrew....@csiro.au

unread,
Jun 20, 2017, 1:28:15 AM6/20/17
to moose...@googlegroups.com
Blast, this is looking like a bug in ComputeMultiPlasticityStress. Probably something quite simple, but that is a beast of a piece of code so it'll take a while to track down. Are you able to send me your input file + mesh?

My only other suggestion is to try plastic_models='' in your input file, which you've probably already tried. That way you're using the entire machinery of ComputeMultiPlasticityStress, but you're guaranteed to have no plastic yield at any quadpoint. If you still get a "jump" then it's a bug, if you don't get a "jump" then probably there is some plasticity somewhere but you failed to see it.
To view this discussion on the web visit https://groups.google.com/d/msgid/moose-users/120bde51-ab84-4414-9a4b-933e3f286ccc%40googlegroups.com.

Martin Lesueur

unread,
Jun 27, 2017, 12:59:22 AM6/27/17
to moose-users
Hi Andy,
The jumps do not appear using as you say plastic_models=''. To rectify my first post, I am pretty convinced that the error comes from a plasticity computation. Some help to understand the problem would really be appreciated Andy. I am currently also comparing with Redback which seems to be giving different results.

The file exposing the jumps is linked to my first post and it is using a generated mesh.
Cheers,

Martin

Andrew....@csiro.au

unread,
Jun 27, 2017, 3:52:39 AM6/27/17
to moose...@googlegroups.com

OK, thanks for doing lots of diagnostic work, Martin. The evidence suggests to me that the machinery of ComputeMultiPlasticityStress is fine, but there is some craziness going on with the individual plasticity model you are using, which makes my work of identifying the error much easier. I'm on holiday at the moment, but will try to get to this ASAP as lots of people are using plasticity and we don't want it to be buggy! Give me about a week,

a


Martin

a

a

Martin

To view this discussion on the web visit https://groups.google.com/d/msgid/moose-users/e0f7ea82-8a3c-4a8f-9bd5-12b2cc902267%40googlegroups.com.

Andrew....@csiro.au

unread,
Jun 27, 2017, 4:00:28 AM6/27/17
to moose...@googlegroups.com
Hey Martin, do you think you could attach (a link to) your input file and the quad mesh (not tri) that you used, at:

https://github.com/idaholab/moose/issues/9391

Please also include a description (can be inside your input file) of which times you see the unwanted jumps of strain rate, as i might accidentally try to debug the wrong thing!

a

________________________________________
From: Wilkins, Andy (Energy, Pullenvale)
Sent: Tuesday, 27 June 2017 5:52 PM

Reply all
Reply to author
Forward
0 new messages