Re: Revisit viscoelastic materials
Hi Andre,
the time history terms *th were neglected during the easier_terms and cython updates. I have planned to update them next week or so as I need them working again. So wait please a few days, I will try to fix them asap.
r.
----- Reply message ----- From: "freevryheid" <freev...@gmail.com> To: <sfepy...@googlegroups.com> Subject: Revisit viscoelastic materials Date: Sun, Jan 29, 2012 13:44 No examples or tests for the dw_lin_elastic_th or dw_lin_elastic_eth terms. Attached problem fails with the following error. Any suggestions to point me in the right direction.
Andre
TypeError Traceback (most recent call last) /home/grassy/sfepy/simple.py in <module>() 123 124 if __name__ == '__main__': --> 125 main()
/home/grassy/sfepy/simple.py in main() 120 parametric_hook = conf.get_function(opts.parametric_hook) 121 app.parametrize( parametric_hook ) --> 122 app() 123 124 if __name__ == '__main__':
/home/grassy/sfepy/sfepy/applications/application.pyc in call_basic(self, **kwargs) 27 28 def call_basic( self, **kwargs ): ---> 29 return self.call( **kwargs ) 30 31 def call_parametrized( self, **kwargs ):
/home/grassy/sfepy/sfepy/applications/simple_app.pyc in call(self) 115 post_process_hook=self.post_process_hook, 116 post_process_hook_final=self.post_process_hook_final, --> 117 pre_process_hook=self.pre_process_hook) 118 119 return out
/home/grassy/sfepy/sfepy/solvers/generic.pyc in solve_direct(conf, options, problem, step_hook, post_process_hook, post_process_hook_final, pre_process_hook, nls_status) 214 step_hook=step_hook, 215 post_process_hook=post_process_hook, --> 216 nls_status=nls_status) 217 else: 218 ##
/home/grassy/sfepy/sfepy/solvers/generic.pyc in solve_evolutionary_op(problem, save_results, return_history, step_hook, post_process_hook, nls_status) 134 135 ii = 0 --> 136 for ts, state in time_solver( state0 ): 137 138 if step_hook is not None:
/home/grassy/sfepy/sfepy/solvers/ts.pyc in __call__(self, state0, conf, step_fun, step_args) 189 output( self.format % (time, step + 1, self.ts.n_step) ) 190 --> 191 state = step_fun( self.ts, state0, *step_args ) 192 state0 = state.copy(deep=True) 193 yield self.ts, state
/home/grassy/sfepy/sfepy/solvers/generic.pyc in time_step_function(ts, state0, problem, nls_status) 80 ev = problem.get_evaluator() 81 try: ---> 82 vec_r = ev.eval_residual(state(), is_full=True) 83 except ValueError: 84 output('initial residual evaluation failed, giving up...')
/home/grassy/sfepy/sfepy/fem/evaluate.pyc in eval_residual(self, vec, is_full) 50 try: 51 pb = self.problem ---> 52 vec_r = pb.equations.eval_residuals(vec) 53 54 except StopIteration, exc:
/home/grassy/sfepy/sfepy/fem/equations.pyc in eval_residuals(self, state, by_blocks, names) 560 out = self.create_stripped_state_vector() 561 --> 562 self.evaluate(mode='weak', dw_mode='vector', asm_obj=out) 563 564 return out
/home/grassy/sfepy/sfepy/fem/equations.pyc in evaluate(self, mode, dw_mode, term_mode, asm_obj) 503 for eq in self: 504 eout = eq.evaluate(mode=mode, dw_mode=dw_mode, term_mode=term_mode, --> 505 asm_obj=asm_obj) 506 if mode != 'weak': 507 out[eq.name] = eout
/home/grassy/sfepy/sfepy/fem/equations.pyc in evaluate(self, mode, dw_mode, term_mode, asm_obj) 730 term_mode=term_mode, 731 standalone=False, --> 732 ret_status=True) 733 term.assemble_to(asm_obj, val, iels, mode=dw_mode) 734
/home/grassy/sfepy/sfepy/terms/terms.pyc in evaluate(self, mode, diff_var, standalone, ret_status, **kwargs) 1501 1502 _args = tuple(args) + (mode, term_mode, diff_var) -> 1503 fargs = self.get_fargs(*_args, **kwargs) 1504 1505 n_elr, n_qpr, dim, n_enr, n_cr = self.get_data_shape(varr)
TypeError: get_fargs() takes at most 3 arguments (8 given)
-- You received this message because you are subscribed to the Google Groups "sfepy-devel" group. To view this discussion on the web visit https://groups.google.com/d/msg/sfepy-devel/-/eNJ3ZUbDbuYJ. To post to this group, send email to sfepy...@googlegroups.com. To unsubscribe from this group, send email to sfepy-devel...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/sfepy-devel?hl=en.
On 01/29/2012 02:14 PM, Robert Cimrman wrote:
Hi Andre,
the time history terms *th were neglected during the easier_terms and cython updates. I have planned to update them next week or so as I need them working again. So wait please a few days, I will try to fix them asap.
r.
re-hi!
Check out [1] to see how to work with the fading memory terms. It's a freshly updated code, so there might be bugs lurking, although the example seems to work fine. I have updated only the linear elastic terms so far, the other fading memory terms will follow soon - it should be easy, as the "infrastructure" is there.
Cheers, r.
[1] http://docs.sfepy.org/doc-devel/examples/linear_elasticity/linear_viscoelast...
Thanks! On Feb 1, 2012 11:47 AM, "Robert Cimrman" <cimr...@ntc.zcu.cz> wrote:
On 01/29/2012 02:14 PM, Robert Cimrman wrote:
Hi Andre,
the time history terms *th were neglected during the easier_terms and
updates. I have planned to update them next week or so as I need them working again. So wait please a few days, I will try to fix them asap.
r.
re-hi!
Check out [1] to see how to work with the fading memory terms. It's a freshly updated code, so there might be bugs lurking, although the example seems to work fine. I have updated only the linear elastic terms so far,
cython the other fading memory terms will follow soon - it should be easy, as the "infrastructure" is there.
Cheers, r.
[1]
http://docs.sfepy.org/doc-devel/examples/linear_elasticity/linear_viscoelast...
-- You received this message because you are subscribed to the Google Groups
"sfepy-devel" group.
To post to this group, send email to sfepy...@googlegroups.com. To unsubscribe from this group, send email to sfepy-devel...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/sfepy-devel?hl=en.
The stress/strain response of these materials are non-linear with time. I suggest extending the example to show these responses with plots of stress vs. time and strain vs. time. Can the de_cauchy_strain/stress terms be used for this?
Andre
On Wed, Feb 1, 2012 at 12:32 PM, Andre Smit <freev...@gmail.com> wrote:
Thanks!
On Feb 1, 2012 11:47 AM, "Robert Cimrman" <cimr...@ntc.zcu.cz> wrote:
On 01/29/2012 02:14 PM, Robert Cimrman wrote:
Hi Andre,
the time history terms *th were neglected during the easier_terms and
updates. I have planned to update them next week or so as I need them working again. So wait please a few days, I will try to fix them asap.
r.
re-hi!
Check out [1] to see how to work with the fading memory terms. It's a freshly updated code, so there might be bugs lurking, although the example seems to work fine. I have updated only the linear elastic terms so far,
cython the other fading memory terms will follow soon - it should be easy, as the "infrastructure" is there.
Cheers, r.
[1]
http://docs.sfepy.org/doc-devel/examples/linear_elasticity/linear_viscoelast...
-- You received this message because you are subscribed to the Google
Groups "sfepy-devel" group.
To post to this group, send email to sfepy...@googlegroups.com. To unsubscribe from this group, send email to sfepy-devel...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/sfepy-devel?hl=en.
-- Andre
Yes, that would be nice. The strain-time plot can be readily done, but for the stress-time we would need to add a term analogous to de_cauchy_stress for the convolution kernel part. Create a new issue, please, on the github site.
r.
On 02/17/2012 04:37 PM, Andre Smit wrote:
The stress/strain response of these materials are non-linear with time. I suggest extending the example to show these responses with plots of stress vs. time and strain vs. time. Can the de_cauchy_strain/stress terms be used for this?
Andre
On Wed, Feb 1, 2012 at 12:32 PM, Andre Smit<freev...@gmail.com> wrote:
Thanks!
On Feb 1, 2012 11:47 AM, "Robert Cimrman"<cimr...@ntc.zcu.cz> wrote:
On 01/29/2012 02:14 PM, Robert Cimrman wrote:
Hi Andre,
the time history terms *th were neglected during the easier_terms and
updates. I have planned to update them next week or so as I need them working again. So wait please a few days, I will try to fix them asap.
r.
re-hi!
Check out [1] to see how to work with the fading memory terms. It's a freshly updated code, so there might be bugs lurking, although the example seems to work fine. I have updated only the linear elastic terms so far,
cython the other fading memory terms will follow soon - it should be easy, as the "infrastructure" is there.
Cheers, r.
[1]
http://docs.sfepy.org/doc-devel/examples/linear_elasticity/linear_viscoelast...
Even better would be to remove the load after 10 steps, say, to see delayed stress/strain recovery.
On Fri, Feb 17, 2012 at 9:37 AM, Andre Smit <freev...@gmail.com> wrote:
The stress/strain response of these materials are non-linear with time. I suggest extending the example to show these responses with plots of stress vs. time and strain vs. time. Can the de_cauchy_strain/stress terms be used for this?
Andre
On Wed, Feb 1, 2012 at 12:32 PM, Andre Smit <freev...@gmail.com> wrote:
Thanks!
On Feb 1, 2012 11:47 AM, "Robert Cimrman" <cimr...@ntc.zcu.cz> wrote:
On 01/29/2012 02:14 PM, Robert Cimrman wrote:
Hi Andre,
the time history terms *th were neglected during the easier_terms and
updates. I have planned to update them next week or so as I need them working again. So wait please a few days, I will try to fix them asap.
r.
re-hi!
Check out [1] to see how to work with the fading memory terms. It's a freshly updated code, so there might be bugs lurking, although the example seems to work fine. I have updated only the linear elastic terms so far,
cython the other fading memory terms will follow soon - it should be easy, as the "infrastructure" is there.
Cheers, r.
[1]
http://docs.sfepy.org/doc-devel/examples/linear_elasticity/linear_viscoelast...
-- You received this message because you are subscribed to the Google
Groups "sfepy-devel" group.
To post to this group, send email to sfepy...@googlegroups.com. To unsubscribe from this group, send email to sfepy-devel...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/sfepy-devel?hl=en.
-- Andre
-- Andre
Hi Andre,
the new terms for evaluating fading memory stresses are implemented now, check [1]. It was a little bit tricky for the 'eth' version of terms because to compute stress one needs to know whole history of loading - there is no cut-off. Therefore 'preserve_caches' argument has been introduced to ProblemDefinition.evaluate() - it must be set to True to evaluate such terms correctly.
I have also changed the loading according to your suggestion. To get the time history plots, follow the instructions at [1] (or in the example docstring).
r.
[1] http://sfepy.org//doc-devel/examples/linear_elasticity/linear_viscoelastic.h...
On 02/17/2012 06:50 PM, Andre Smit wrote:
Even better would be to remove the load after 10 steps, say, to see delayed stress/strain recovery.
On Fri, Feb 17, 2012 at 9:37 AM, Andre Smit<freev...@gmail.com> wrote:
The stress/strain response of these materials are non-linear with time. I suggest extending the example to show these responses with plots of stress vs. time and strain vs. time. Can the de_cauchy_strain/stress terms be used for this?
Andre
On Wed, Feb 1, 2012 at 12:32 PM, Andre Smit<freev...@gmail.com> wrote:
Brilliant!
I'm closing issues #172 and #114. When I get some time I'll work in a tutorial on these terms. A comparison with Abaqus that uses Prony series for viscoelastic definition may be interesting.
a
On Tue, Feb 21, 2012 at 7:52 AM, Robert Cimrman <cimr...@ntc.zcu.cz> wrote:
Hi Andre,
the new terms for evaluating fading memory stresses are implemented now, check [1]. It was a little bit tricky for the 'eth' version of terms because to compute stress one needs to know whole history of loading - there is no cut-off. Therefore 'preserve_caches' argument has been introduced to ProblemDefinition.evaluate() - it must be set to True to evaluate such terms correctly.
I have also changed the loading according to your suggestion. To get the time history plots, follow the instructions at [1] (or in the example docstring).
r.
[1] http://sfepy.org//doc-devel/**examples/linear_elasticity/** linear_viscoelastic.html<http://sfepy.org//doc-devel/examples/linear_elasticity/linear_viscoelastic.h...>
On 02/17/2012 06:50 PM, Andre Smit wrote:
Even better would be to remove the load after 10 steps, say, to see delayed stress/strain recovery.
On Fri, Feb 17, 2012 at 9:37 AM, Andre Smit<freev...@gmail.com> wrote:
The stress/strain response of these materials are non-linear with time. I
suggest extending the example to show these responses with plots of stress vs. time and strain vs. time. Can the de_cauchy_strain/stress terms be used for this?
Andre
On Wed, Feb 1, 2012 at 12:32 PM, Andre Smit<freev...@gmail.com> wrote:
-- You received this message because you are subscribed to the Google Groups "sfepy-devel" group. To post to this group, send email to sfepy...@googlegroups.com. To unsubscribe from this group, send email to sfepy-devel+unsubscribe@** googlegroups.com <sfepy-devel%...@googlegroups.com>. For more options, visit this group at http://groups.google.com/** group/sfepy-devel?hl=en <http://groups.google.com/group/sfepy-devel?hl=en> .
-- Andre
Quick googling told me, that Prony series is a sum of exponential terms with different decays, right? As such, a single term in the series IMHO directly corresponds to the ETH version of the fading memory term. One could simulate the whole series by adding several ETH terms. The TH terms are more general, as the fading memory kernel can by any (even weird) function.
Looking forward to the tutorial.
r.
On 02/21/2012 03:41 PM, Andre Smit wrote:
Brilliant!
I'm closing issues #172 and #114. When I get some time I'll work in a tutorial on these terms. A comparison with Abaqus that uses Prony series for viscoelastic definition may be interesting.
a
On Tue, Feb 21, 2012 at 7:52 AM, Robert Cimrman<cimr...@ntc.zcu.cz> wrote:
Hi Andre,
the new terms for evaluating fading memory stresses are implemented now, check [1]. It was a little bit tricky for the 'eth' version of terms because to compute stress one needs to know whole history of loading - there is no cut-off. Therefore 'preserve_caches' argument has been introduced to ProblemDefinition.evaluate() - it must be set to True to evaluate such terms correctly.
I have also changed the loading according to your suggestion. To get the time history plots, follow the instructions at [1] (or in the example docstring).
r.
[1] http://sfepy.org//doc-devel/**examples/linear_elasticity/** linear_viscoelastic.html<http://sfepy.org//doc-devel/examples/linear_elasticity/linear_viscoelastic.h...>
On 02/17/2012 06:50 PM, Andre Smit wrote:
Even better would be to remove the load after 10 steps, say, to see delayed stress/strain recovery.
On Fri, Feb 17, 2012 at 9:37 AM, Andre Smit<freev...@gmail.com> wrote:
The stress/strain response of these materials are non-linear with time. I
suggest extending the example to show these responses with plots of stress vs. time and strain vs. time. Can the de_cauchy_strain/stress terms be used for this?
Andre
Yes - attached paper shows for example how to convert Burger's model into Prony series.
On Tue, Feb 21, 2012 at 9:08 AM, Robert Cimrman <cimr...@ntc.zcu.cz> wrote:
Quick googling told me, that Prony series is a sum of exponential terms with different decays, right? As such, a single term in the series IMHO directly corresponds to the ETH version of the fading memory term. One could simulate the whole series by adding several ETH terms. The TH terms are more general, as the fading memory kernel can by any (even weird) function.
Looking forward to the tutorial.
r.
On 02/21/2012 03:41 PM, Andre Smit wrote:
Brilliant!
I'm closing issues #172 and #114. When I get some time I'll work in a tutorial on these terms. A comparison with Abaqus that uses Prony series for viscoelastic definition may be interesting.
a
On Tue, Feb 21, 2012 at 7:52 AM, Robert Cimrman<cimr...@ntc.zcu.cz> wrote:
Hi Andre,
the new terms for evaluating fading memory stresses are implemented now, check [1]. It was a little bit tricky for the 'eth' version of terms because to compute stress one needs to know whole history of loading - there is no cut-off. Therefore 'preserve_caches' argument has been introduced to ProblemDefinition.evaluate() - it must be set to True to evaluate such terms correctly.
I have also changed the loading according to your suggestion. To get the time history plots, follow the instructions at [1] (or in the example docstring).
r.
[1] http://sfepy.org//doc-devel/****examples/linear_elasticity/**<http://sfepy.org//doc-devel/**examples/linear_elasticity/**> linear_viscoelastic.html<http:**//sfepy.org//doc-devel/** examples/linear_elasticity/**linear_viscoelastic.html<http://sfepy.org//doc-devel/examples/linear_elasticity/linear_viscoelastic.h...>
On 02/17/2012 06:50 PM, Andre Smit wrote:
Even better would be to remove the load after 10 steps, say, to see
delayed stress/strain recovery.
On Fri, Feb 17, 2012 at 9:37 AM, Andre Smit<freev...@gmail.com> wrote:
The stress/strain response of these materials are non-linear with time. I
suggest extending the example to show these responses with plots of stress vs. time and strain vs. time. Can the de_cauchy_strain/stress terms be used for this?
Andre
-- You received this message because you are subscribed to the Google Groups "sfepy-devel" group. To post to this group, send email to sfepy...@googlegroups.com. To unsubscribe from this group, send email to sfepy-devel+unsubscribe@** googlegroups.com <sfepy-devel%...@googlegroups.com>. For more options, visit this group at http://groups.google.com/** group/sfepy-devel?hl=en <http://groups.google.com/group/sfepy-devel?hl=en> .
-- Andre
participants (2)
-
Andre Smit -
Robert Cimrman