SfePy December 2015

sfepy@python.org

6 participants
14 discussions

by Robert Cimrman

In examples/large_deformation/hyperelastic.py a rotation by displacements is applied. By using a similar function the vectors defining the force couples could be defined for dw_surface_ltr (IMHO). Does it make sense? r. ----- Reply message ----- From: "Andre Smit" <freev...(a)gmail.com> To: <sfepy...(a)googlegroups.com> Subject: Torque Date: Sat, Dec 18, 2010 05:10 What is the best way to apply a torque load to a model? -- 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...(a)googlegroups.com. To unsubscribe from this group, send email to sfepy-devel...(a)googlegroups.com. For more options, visit this group at http://groups.google.com/group/sfepy-devel?hl=en.

2 years, 2 months

new interactive example with time stepping

by Robert Cimrman

Hi, I have just updated the time stepping solvers in sfepy for interactive use, as demonstrated in the new example [1]. For basic use, ignore the probing code - the time stepper can be used as simply as: tss = SimpleTimeSteppingSolver({'t0' : 0.0, 't1' : 100.0, 'n_step' : 11}, problem=problem) tss.init_time() for step, time, state in tss(): pass r. [1] http://sfepy.org/doc-devel/examples/diffusion/time_poisson_interactive.html

3 years, 2 months

postprocessing

by Nina Halabi

Dear Robert I could overcome my previous problems of not having root privileges, so it all worked up till the command: ./postproc.py cylinder.vtk This echoed the below message: Exception In /home/gmh/.local/lib/python2.7/site-packages/mayavi-4.4.4.dev0-py2.7-linux-x86_64.egg/mayavi/sources/vtk_xml_file_reader.py:86 AttributeError: 'AlgorithmOutput' object has no attribute 'point_data' (in get_all_attributes) Traceback (most recent call last): File "/home/gmh/.local/lib/python2.7/site-packages/traits-4.6.0.dev354-py2.7-linux-x86_64.egg/traits/trait_notifiers.py", line 340, in __call__ self.handler( *args ) File "/home/gmh/sfepy/sfepy/postprocess/dataset_manager.py", line 177, in _dataset_changed self._assign_attribute.input = value File "/home/gmh/.local/lib/python2.7/site-packages/traits-4.6.0.dev354-py2.7-linux-x86_64.egg/traits/trait_handlers.py", line 104, in _read_only name, class_of( object ) ) TraitError: The 'input' trait of an AssignAttribute instance is 'read only'. sfepy: point scalars t at [-0.05 -0.02 0. ] sfepy: range: -2.00e+00 2.00e+00 l2 norm range: 1.06e-02 2.00e+00 No handlers could be found for logger "mayavi.core.common" I'm using Python 2.7.8 if that's relevant to the current issue. Thanks again Cheers

4 years

1D line elements in SfePy

by Nimish

I am currrently looking for FEM packages to help me solve a system of beams and columns, basically a collection of 1D bernoulli/timoshenko line elements. I started reading SfePy docs and i am getting the idea that doing the above is not really possible here, am i right? Are only 2D area elements permitted in SfePy? Or is there any direct support for solving 1D line elements too.. Cheers Nimish

4 years, 4 months

Non linear Poisson Equation

by Francesca Camorrino

Hello! I'm trying to solve Poisson equation with Sfepy in a very simple domain, but I found some problem: 1) I need triangular mesh on a rectangular/square domain 2) I have null von neumann conditions ( and I omitted them in boundary condition) like in the discussion https://groups.google.com/forum/#!searchin/sfepy-devel/null$20von$20neumann… 3)I know that this problem have many convergence problem, in one of this case is the simulation successful? Or I find some error? I don't know because the simulation is successful but in the postprocessing the result isn't correct I attach the problem description file too. If you want I would like some advice about my problem Thank you very much -Francesca

4 years, 5 months

cannot set toolkit to wx because it has already been set to qt4

by silesa...＠hotmail.com

I have made an inline instalation of sfepy-2013.3 on fedora22 , I run the tests and I have only 3 failed, they maybe related to the parallel option since I don't have instaled that option. I have python 2.7.10, mayavi 4.4.3, VTK 6.1.0, Qt 4.8.6 Either on the python or ipython window when I try to import: *from sfepy.postprocess.viewer import Viewe*r I get the following message, *cannot set toolkit to wx because it has already been set to qt4* and of course any postprocessing will fail

4 years, 5 months

failed test

by Nina Halabi

Hello Kindly , I'm trying to install SfePy on a redhat linux workstation. I was running the python tests and test_linear_solvers.py failed. the command: python run_tests.py --debug tests/test_linear_solvers.py gave the below. Please note that I got the message "mpi4py (optional) is available in version 1.3, but >= 1.3.1 is needed; you may get run-time errors" during compilation. Cheers, gmh@calx039$ python run_tests.py --debug tests/test_linear_solvers.py <<< directory: tests, test files: 1 <<< tests/test_linear_solvers.py sfepy: left over: ['verbose', '__builtins__', '__doc__', '__name__', 'TestCommon', 'data_dir', '_filename', '__package__', 'output_name', '__file__'] sfepy: reading mesh [line2, tri3, quad4, tetra4, hexa8] (/home/gmh/Desktop/sfepy/meshes/3d/special/cube_cylinder.mesh)... sfepy: ...done in 0.14 s sfepy: creating regions... sfepy: Gamma_Right sfepy: Omega sfepy: Gamma_Left sfepy: ...done in 0.18 s sfepy: equation "Temperature": sfepy: dw_laplace.i.Omega(coef.val, s, t) = 0 sfepy: using solvers: ts: no ts nls: newton ls: d01 sfepy: updating variables... sfepy: ...done sfepy: setting up dof connectivities... sfepy: ...done in 0.00 s sfepy: matrix shape: (9511, 9511) sfepy: assembling matrix graph... sfepy: ...done in 0.04 s sfepy: matrix structural nonzeros: 132239 (1.46e-03% fill) >>> test instance prepared (1 test(s)) ... d00 ls.scipy_direct umfpack ... matrix size: (9511, 9511) ... nnz: 132239 sfepy: updating materials... sfepy: coef sfepy: ...done in 0.01 s sfepy: nls: iter: 0, residual: 3.587850e+00 (rel: 1.000000e+00) sfepy: rezidual: 0.01 [s] sfepy: solve: 1.49 [s] sfepy: matrix: 0.02 [s] sfepy: nls: iter: 1, residual: 4.898229e-15 (rel: 1.365227e-15) ... rezidual: 0.01 [s] ... solve: 1.49 [s] ... matrix: 0.02 [s] ... condition: 0, err0: 3.588e+00, err: 4.898e-15 ... d01 ls.scipy_direct superlu ... matrix size: (9511, 9511) ... nnz: 132239 sfepy: updating materials... sfepy: coef sfepy: ...done in 0.00 s sfepy: nls: iter: 0, residual: 3.587850e+00 (rel: 1.000000e+00) sfepy: rezidual: 0.01 [s] sfepy: solve: 1.34 [s] sfepy: matrix: 0.02 [s] sfepy: nls: iter: 1, residual: 1.554473e-14 (rel: 4.332602e-15) ... rezidual: 0.01 [s] ... solve: 1.34 [s] ... matrix: 0.02 [s] ... condition: 0, err0: 3.588e+00, err: 1.554e-14 ... i00 ls.pyamg ruge_stuben_solver ... matrix size: (9511, 9511) ... nnz: 132239 ... solver failed: ... cannot import pyamg! ... i01 ls.pyamg smoothed_aggregation_solver ... matrix size: (9511, 9511) ... nnz: 132239 ... solver failed: ... cannot import pyamg! ... i10 ls.petsc cg icc ... matrix size: (9511, 9511) ... nnz: 132239 sfepy: updating materials... sfepy: coef sfepy: ...done in 0.00 s sfepy: nls: iter: 0, residual: 3.587850e+00 (rel: 1.000000e+00) sfepy: rezidual: 0.01 [s] sfepy: solve: 0.09 [s] sfepy: matrix: 0.02 [s] sfepy: nls: iter: 1, residual: 2.427756e-12 (rel: 6.766603e-13) ... rezidual: 0.00 [s] ... solve: 0.09 [s] ... matrix: 0.02 [s] ... condition: 0, err0: 3.588e+00, err: 2.428e-12 ... i20 ls.scipy_iterative cg ... matrix size: (9511, 9511) ... nnz: 132239 sfepy: updating materials... sfepy: coef sfepy: ...done in 0.00 s sfepy: nls: iter: 0, residual: 3.587850e+00 (rel: 1.000000e+00) sfepy: cg convergence: 0 (successful exit) sfepy: rezidual: 0.02 [s] sfepy: solve: 0.05 [s] sfepy: matrix: 0.02 [s] sfepy: nls: iter: 1, residual: 3.287606e-12 (rel: 9.163164e-13) ... rezidual: 0.00 [s] ... solve: 0.05 [s] ... matrix: 0.02 [s] ... condition: 0, err0: 3.588e+00, err: 3.288e-12 ... i21 ls.scipy_iterative bicgstab ... matrix size: (9511, 9511) ... nnz: 132239 sfepy: updating materials... sfepy: coef sfepy: ...done in 0.00 s sfepy: nls: iter: 0, residual: 3.587850e+00 (rel: 1.000000e+00) sfepy: bicgstab convergence: 0 (successful exit) sfepy: rezidual: 0.01 [s] sfepy: solve: 0.07 [s] sfepy: matrix: 0.02 [s] sfepy: nls: iter: 1, residual: 2.835370e-12 (rel: 7.902699e-13) ... rezidual: 0.01 [s] ... solve: 0.07 [s] ... matrix: 0.02 [s] ... condition: 0, err0: 3.588e+00, err: 2.835e-12 ... i22 ls.scipy_iterative qmr ... matrix size: (9511, 9511) ... nnz: 132239 sfepy: updating materials... sfepy: coef sfepy: ...done in 0.00 s sfepy: nls: iter: 0, residual: 3.587850e+00 (rel: 1.000000e+00) sfepy: qmr convergence: 0 (successful exit) sfepy: rezidual: 0.01 [s] sfepy: solve: 0.11 [s] sfepy: matrix: 0.02 [s] sfepy: nls: iter: 1, residual: 3.549567e-12 (rel: 9.893297e-13) ... rezidual: 0.01 [s] ... solve: 0.11 [s] ... matrix: 0.02 [s] ... condition: 0, err0: 3.588e+00, err: 3.550e-12 ... solution times (rezidual norms): ... 0.05 [s] (3.288e-12) : i20 ls.scipy_iterative cg ... 0.07 [s] (2.835e-12) : i21 ls.scipy_iterative bicgstab ... 0.09 [s] (2.428e-12) : i10 ls.petsc cg icc ... 0.11 [s] (3.550e-12) : i22 ls.scipy_iterative qmr ... 1.34 [s] (1.554e-14) : d01 ls.scipy_direct superlu ... 1.49 [s] (4.898e-15) : d00 ls.scipy_direct umfpack ... 10000000000.00 [s] (1.000e+10) : i00 ls.pyamg ruge_stuben_solver ... 10000000000.00 [s] (1.000e+10) : i01 ls.pyamg smoothed_aggregation_solver +++ test_solvers: ok >>> all passed in 5.02 s 1 test file(s) executed in 5.02 s, 0 failure(s) of 1 test(s) WARNING! There are options you set that were not used! WARNING! could be spelling mistake, etc! Option left: name:-sub_pc_type (no value) gmh@calx039$

4 years, 5 months

Explicit FEM analysis

by nayan...＠gmail.com

Dear Robert, Which linear and nonlinear solver should I use in case of Explicit FEM analysis? Regards, Nayan

4 years, 5 months

Re: [sfepy-devel] Non linear Poisson Equation

by Robert Cimrman

Hi Francesca, On 12/10/2015 11:41 AM, Francesca Camorrino wrote: > Hello! > I'm trying to solve Poisson equation with Sfepy in a very simple domain, > but I found some problem: > 1) I need triangular mesh on a rectangular/square domain For this, you can try gmsh. Why do you need triangles, by the way? > 2) I have null von neumann conditions ( and I omitted them in boundary > condition) like in the > discussion https://groups.google.com/forum/#!searchin/sfepy-devel/null$20von$20neumann… OK > 3)I know that this problem have many convergence problem, in one of this > case is the simulation successful? Or I find some error? You fix the potential in some, so it should work fine. But your definition for region 'p' selects nothing for the mesh you generate - try 'p' : ('vertices in (x <= 0.99951 )', 'vertex'), instead. Always verify your regions using --save-regions-as-groups option: ./simple.py prob_desc_file.py --save-regions-as-groups - then check the 0_regions.vtk file using postproc.py. > I don't know because the simulation is successful but in the postprocessing > the result isn't correct > I attach the problem description file too. If you want I would like some > advice about my problem When I run your code, it does not do anything, because there is no load - maybe some right-hand side term is missing? Or provide a Dirichlet BC also in the region 'n' (again, fix the region definition). r. PS: I missed your message, so my reply is late.

4 years, 5 months

Navier Stokes problem with arbitrary inlet/ outlet plane

by Jacky Ko

The example http://sfepy.org/doc-devel/examples/navier_stokes/navier_stokes.html clearly illustrated a NS problem with inlet velocity normal to y axis. Is it possible to set the inlet with arbitrary plane that is not normal to any axis? e.g. inlet normal velocity = (u1, u2, u3) I see that in this example the inlet and outlet boundaries are defined by the cinc functions: cinc_name = 'cinc_' + op.splitext(op.basename(filename_mesh))[0]cinc = getattr(utils, cinc_name) functions = { 'cinc0' : (lambda coors, domain=None: cinc(coors, 0),), 'cinc1' : (lambda coors, domain=None: cinc(coors, 1),),} how would this utility work with my arbitrary inlet/outlet boundaries?

4 years, 5 months

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

SfePy December 2015