not working fater upgrading
Hi!
I was runnind some code under sfepy2014.3 version and today I've upgraded to 2015.3 and now my simulations only do 1 iteration and stop when it should do 100 hundred. Why?
Hi Patricia,
On 07/31/2015 05:10 PM, Patricia Garcia Cañadilla wrote:
Hi!
I was runnind some code under sfepy2014.3 version and today I've upgraded to 2015.3 and now my simulations only do 1 iteration and stop when it should do 100 hundred. Why?
You mean 2015.2, right? Could you send an example to this list? Your description does not provide any hints on what exactly does not work.
r.
Yes, version 2015.2
Sorry, what is happening now is that is not working the time_stepper solver, so, only solve the first iteration and stops when before I could do the 100 iterations. I think that maybe you change something in the time solvers in this last version, isn't?
This is how my simulation end:
sfepy: reading mesh [line2, tri3, quad4, tetra4, hexa8] (/Users/Patricia/Documents/sfepy-2014.3/meshes/2D/realcell_mesh_22.mesh)...
sfepy: ...done in 0.06 s
sfepy: reading mesh [line2, tri3, quad4, tetra4, hexa8] (/Users/Patricia/Documents/sfepy-2014.3/meshes/2D/realcell_mesh_22.mesh)...
sfepy: ...done in 0.05 s
sfepy: reading mesh [line2, tri3, quad4, tetra4, hexa8] (/Users/Patricia/Documents/sfepy-2014.3/meshes/2D/realcell_mesh_22.mesh)...
sfepy: ...done in 0.05 s
sfepy: left over: ['help', 'Conditions', 'row', 'ifile', 'verbose', '_filename', 'ScipyDirect', 'min_x', 'lame_from_youngpoisson', 'Mesh', 'get_elastic_pars', 'usage', 'main', '__doc__', 'OptionParser', '__file__', 'vf_fibres', 'fix_point', 'SimpleTimeSteppingSolver', 'Problem', 'b', 'd', 'strain', 'Newton', 'Integral', 'posx_ca', 'stress_strain', 'IndexedStruct', 'x', 'eps2', 'domain', 'build_activation', 'Field', 'FEDomain', 'sfepy', 'vf_matrix', 'get_pars_fibres', 'Function', 'data_dir', 'nm', 'Material', '__package__', 'lreal', 'reader', 'csv', 'Equations', 'max_x', 'FieldVariable', 'Term', 'E', 'linear_tension', '__builtins__', 'Equation', 'sys', 'mesh', 'posx', '__name__', 'a', 'c', 'limits', 'eps', 'EssentialBC']
sfepy: reading mesh [line2, tri3, quad4, tetra4, hexa8] (/Users/Patricia/Documents/sfepy-2014.3/meshes/2D/realcell_mesh_22.mesh)...
sfepy: ...done in 0.05 s
sfepy: creating regions...
sfepy: Omega
sfepy: Gamma
sfepy: Fix1
sfepy: Fix2
sfepy: Fix
sfepy: ...done in 0.01 s
sfepy: using solvers:
ts: ts
nls: newton
ls: lssfepy: equation "eq":
sfepy: dw_tl_he_neohook.1.Omega( solid.mu, v, u )
+ dw_tl_he_mooney_rivlin.1.Omega( solid.kappa, v, u )
+ dw_tl_bulk_penalty.1.Omega( solid.K, v, u )
+ dw_tl_fib_a.i1.Omega( f1.fmax, f1.eps_opt, f1.s, f1.fdir, f1.act, v, u )
= - dw_surface_dot.2.Gamma(load.val,v,u)sfepy: updating variables...
sfepy: ...done
sfepy: setting up dof connectivities...
sfepy: ...done in 0.00 s
sfepy: matrix shape: (7888, 7888)
sfepy: assembling matrix graph...
sfepy: ...done in 0.01 s
sfepy: matrix structural nonzeros: 107632 (1.73e-03% fill)
sfepy: updating materials...
sfepy: solid
sfepy: load
sfepy: f1
sfepy: ...done in 0.05 s
sfepy: nls: iter: 0, residual: 0.000000e+00 (rel: 0.000000e+00)
I'be tried to move again to version 2014.3 and I couldn't since I got compilation errors in cmesh modul :(
El viernes, 31 de julio de 2015, 18:57:11 (UTC+2), Robert Cimrman escribió:
Hi Patricia,
On 07/31/2015 05:10 PM, Patricia Garcia Cañadilla wrote:
Hi!
I was runnind some code under sfepy2014.3 version and today I've upgraded to 2015.3 and now my simulations only do 1 iteration and stop when it should do 100 hundred. Why?
You mean 2015.2, right? Could you send an example to this list? Your description does not provide any hints on what exactly does not work.
r.
On 07/31/2015 07:08 PM, Patricia Garcia Cañadilla wrote:
Yes, version 2015.2
Sorry, what is happening now is that is not working the time_stepper solver, so, only solve the first iteration and stops when before I could do the 100 iterations. I think that maybe you change something in the time solvers in this last version, isn't?
Some things were changed so that the time stepping solvers can be used interactively.
This is how my simulation end:
sfepy: reading mesh [line2, tri3, quad4, tetra4, hexa8] (/Users/Patricia/Documents/sfepy-2014.3/meshes/2D/realcell_mesh_22.mesh)... sfepy: ...done in 0.05 s
sfepy: nls: iter: 0, residual: 0.000000e+00 (rel: 0.000000e+00)
So the initial time is in equilibrium?
I'be tried to move again to version 2014.3 and I couldn't since I got compilation errors in cmesh modul :(
That should work - try removing the "build" directory and running "python setup.py clean --all"
r.
El viernes, 31 de julio de 2015, 18:57:11 (UTC+2), Robert Cimrman escribió:
Hi Patricia,
On 07/31/2015 05:10 PM, Patricia Garcia Cañadilla wrote:
Hi!
I was runnind some code under sfepy2014.3 version and today I've upgraded to 2015.3 and now my simulations only do 1 iteration and stop when it should do 100 hundred. Why?
You mean 2015.2, right? Could you send an example to this list? Your description does not provide any hints on what exactly does not work.
r.
El viernes, 31 de julio de 2015, 20:05:49 (UTC+2), Robert Cimrman escribió:
On 07/31/2015 07:08 PM, Patricia Garcia Cañadilla wrote:
Yes, version 2015.2
Sorry, what is happening now is that is not working the time_stepper solver, so, only solve the first iteration and stops when before I could do the 100 iterations. I think that maybe you change something in the time solvers in this last version, isn't?
Some things were changed so that the time stepping solvers can be used interactively.
This is how my simulation end:
sfepy: reading mesh [line2, tri3, quad4, tetra4, hexa8]
(/Users/Patricia/Documents/sfepy-2014.3/meshes/2D/realcell_mesh_22.mesh)...
sfepy: ...done in 0.05 s
sfepy: nls: iter: 0, residual: 0.000000e+00 (rel: 0.000000e+00)
So the initial time is in equilibrium?
Yes, the initial state is in equilibrium.
I'be tried to move again to version 2014.3 and I couldn't since I got compilation errors in cmesh modul :(
That should work - try removing the "build" directory and running "python setup.py clean --all"
I did it, but when I build again the version 2014.3 give me errors of cpython compiler in cmesh module...
r.
El viernes, 31 de julio de 2015, 18:57:11 (UTC+2), Robert Cimrman escribió:
Hi Patricia,
On 07/31/2015 05:10 PM, Patricia Garcia Cañadilla wrote:
Hi!
I was runnind some code under sfepy2014.3 version and today I've upgraded to 2015.3 and now my simulations only do 1 iteration and stop when it should do 100 hundred. Why?
You mean 2015.2, right? Could you send an example to this list? Your description does not provide any hints on what exactly does not work.
r.
In my main code I had this:
problem = Problem.from_conf(conf, init_equations=False)
problem.setup_default_output()
problem.set_equations({'eq' : conf.equations['balance']})
problem.time_update(ebcs=conf.ebcs)
state_el = problem.solve()
time_solver = problem.get_time_solver()
time_solver(save_results=True,post_process_hook=stress_strain)
ts = problem.get_timestepper()
Ans these were my solvers:
# Solvers etc. solver_0 = { 'name' : 'ls', 'kind' : 'ls.scipy_direct', }
solver_1 = { 'name' : 'newton', 'kind' : 'nls.newton',
'i_max' : 7,
'eps_a' : 1e-10,
'eps_r' : 1.0,
'macheps' : 1e-16,
'lin_red' : 1e-2, # Linear system error < (eps_a * lin_red).
'ls_red' : 0.1,
'ls_red_warp': 0.001,
'ls_on' : 1.1,
'ls_min' : 1e-5,
'check' : 0,
'delta' : 1e-6,
'problem' : 'nonlinear', # 'nonlinear' or 'linear' (ignore i_max)}
solver_2 = { 'name' : 'ts', 'kind' : 'ts.simple',
't0' : 0,
't1' : 1,
'dt' : None,
'n_step' : 101, # has precedence over dt!}
El viernes, 31 de julio de 2015, 18:57:11 (UTC+2), Robert Cimrman escribió:
Hi Patricia,
On 07/31/2015 05:10 PM, Patricia Garcia Cañadilla wrote:
Hi!
I was runnind some code under sfepy2014.3 version and today I've upgraded to 2015.3 and now my simulations only do 1 iteration and stop when it should do 100 hundred. Why?
You mean 2015.2, right? Could you send an example to this list? Your description does not provide any hints on what exactly does not work.
r.
participants (2)
-
Patricia Garcia Cañadilla -
Robert Cimrman