Skip to forum content
OOFEM forum
OOFEM - Object Oriented Finite Element Solver
You are not logged in. Please login or register.
Active topics Unanswered topics
OOFEM news
We are happy to announce
OOFEM training courses. See
this page for more information and registration.
New forum registrations: Due to the large amount of spam, the new registrations have been disabled.
To get a forum account, please send a request to oofem(at)fsv.cvut.cz
Search options
Hi everybody, I want to conduct an eigenvalue analysis over a thin plate consisted of shell elements. I want to do this in two steps, first initiate a linear analysis then change some boundary conditions and calculate the eigenvalues of this deformed state? How can this be done in OOFEM? I tried to use eigenvalue dynamics but it seems improbable to perturb the structure with initial loading and then change the boundary conditions by using just the eigenvalue solver. You can find the link to my desired model below.
Best Regards
link:https://classes.engineering.wustl.edu/2009/spring/mase5513/abaqus/docs/v6.6/books/bmk/default.htm?startat=ch01s02ach16.html
Hi everyone!! I was glancing through the manual. I saw that creep formulation is only available for concrete models. Now I have very little info regarding concrete model, yet I am interested in modelling creep for metals and other solid material models. My question is this: Is OOFEM available for all kinds of creep analysis or only for concrete, If so can any kind of material creep be modelled using your concrete creep model?
Best Regards,
deniz
Hello again,
I have tried the python bindings and I was able to successfully execute all the test cases except for the fifth one which is the case I needed to use the most. When I run test5, I am getting the following error while all the other tests run properly. What may be the reason behind this?
Solving problem
OctreeLocalizer: init
Spatial localizer init done
Computing initial guess
StaticStructural :: solveYourselfAt - Solving step 1, metastep 1, (neq = 3)
NRSolver: Iteration ForceError
Dear all, I am trying to run a big model of which has a very good chance of crashing at some point and needs to be restarted from a certain time point. It results the following output when I run the restart command "-r " as used in the “context01.sh” shell script. I could not manage to restart the solution. It generates binary osf files and empty .out files when I run it. I asked a similar question before and I was told that the restart function was bugged. I was wondering if anyone resolved this issue or do you experience a similar kind of bug in the restart function?
“oofem -f context01.in.0 -r 1
“Restoring context for time step 1.0 /Segmentation fault”
dear borek sorry for the delay i am adding the input below
best regards
Dear nitramkaroh, I have been trying to model this problem as you suggested. I am using arc length implementation of "Calmls" with customized parameters taken from the benchmark itself yet I recieve an error as "discriminant is negative, solution failed". I used NLBeam with Linear Elastic material. Yet I cannot resolve the problem. Either NLBeam is not geometrically non-linear or I cannot initiate initial buckling with perturbation load. Can you forward me in the direction for resolving this problem? I am using a custom code for this analysis therefore I cannot add the whole mesh and all nodes but I am adding significant parts and the error I recieve. Thanks in advence
Best regards,
nlb.out
beam
NonLinearStatic nsteps 4 solverType "calm" rtolf 1e-4 stepLength 0.001 minsteplength 0.0001 nmodules 1
vtkxml tstep_all domain_all vars 1 1 primvars 1 1 stype 0
domain 3dShell
OutputManager tstep_all dofman_all element_all
ndofman 42 nelem 40 ncrosssect 1 nmat 1 nbc 6 nic 0 nltf 3 nset 2
node 1 coords 3 -169.0 0.0 0.0 bc 6 0 1 1 1 1 6
node 2 coords 3 -160.51471862576142 8.485281374238571 0.0 bc 6 1 0 0 0 1 1
node 3 coords 3 -152.02943725152286 16.970562748477143 0.0 bc 6 1 0 0 0 1 1
node 4 coords 3 -143.54415587728428 25.45584412271571 0.0 bc 6 1 0 0 0 1 1
node 5 coords 3 -135.05887450304573 33.941125496954285 0.0 bc 6 1 0 0 0 1 1
node 6 coords 3 -126.57359312880715 42.42640687119285 0.0 bc 6 1 0 0 0 1 1
....
node 42 coords 3 0.00 0.0 0.0
LIBeam3dNL 1 nodes 2 1 2 refnode 42 nlgeo 1
LIBeam3dNL 2 nodes 2 2 3 refnode 42 nlgeo 1
LIBeam3dNL 3 nodes 2 3 4 refnode 42 nlgeo 1
LIBeam3dNL 4 nodes 2 4 5 refnode 42 nlgeo 1
....
LIBeam3dNL 40 nodes 2 40 41 refnode 42 nlgeo 1
SimpleCS 1 area 18 Iy 1350 Iz 0.54 Ik 0 beamShearCoeff 1 material 1 set 1
IsoLE 1 d 1.0 E 71240 n 0.31 tAlpha 1.2e-5
BoundaryCondition 1 loadTimeFunction 1 prescribedvalue 0.0
BoundaryCondition 2 loadTimeFunction 1 dofs 1 1 values 1 0
BoundaryCondition 3 loadTimeFunction 1 dofs 1 1 values 1 0
NodalLoad 4 loadTimeFunction 3 dofs 2 1 2 components 2 0.0 -0.18 set 2
BoundaryCondition 5 loadTimeFunction 2 prescribedvalue 1.0
BoundaryCondition 6 loadTimeFunction 2 prescribedvalue -1.0
PiecewiseLinFunction 1 npoints 4 t 4 0 1 2 3 f(t) 4 1.0 1.0 1.0 1.0
PiecewiseLinFunction 2 npoints 4 t 4 0 1 2 3 f(t) 4 0 0.85 1.12 6.28
PiecewiseLinFunction 3 npoints 4 t 4 0 1 2 3 f(t) 4 1.0 0 0 0 (initiation and removal of buckling)
Set 1 elementranges {(1 40)}
set 2 noderanges {(1 41)}
error:computational time -1.000000 is out of given time 0.000000, using closest value
_______________________________________________________
CALMLS: 1 8.895990e-04 2.119532e-07 1.484564e-27
CALMLS: Adjusted step length: 1.000000e-03
Equilibrium reached at load level = 0.000890 in 1 iterations
EngngModel info: user time consumed by solution step 1: 0.02s
Solving [step number 2.0, time = 1.000000e+00]
CALMLS: Initial step length: 1.000000e-03
CALMLS: Iteration LoadLevel ForceError DisplError
----------------------------------------------------------------------------
_______________________________________________________
Error: (/home/bp/devel/oofem.git/src/oofemlib/calmls.C:1045)
In CylindricalALM::computeDeltaLambda:
discriminant is negative, solution failed
_______________________________________________________
Dear all,
I have been trying to conduct an EigenValueDynamic analysis using OOFEM. I tried to make the FV52 benchmark from NAFEMS that can be found on this link: https://abaqus-docs.mit.edu/2017/Englis … c-fv52.htm
I have created the model, applied the boundary conditions properly and got some results. However, the results I've been getting are orders of magnitudes of higher than what they are supposed to be.
I have used both the default OOFEM solver and SLEPc to solve this problem. They both end up in roughly the same results. I conducted the analysis with both fully constrained and under-constrained (the original)versions of the benchmark.
In all cases, the results are off by a high margin. Does anyone have an idea on how to tackle this issue?
The solver is called like the following line:
EigenValueDynamic nroot 10 stype 1 rtolv 1.e-4 nmodules 1 niter 10000
Thanks in advance.
Dear all,
I've been trying to implement and solve a thermal strain problem in OOFEM. However, I am not sure which solver and domain to use.
The problem I want to solve is the LE11 benchmark from NAFEMS. The problem can be found on this link: https://abaqus-docs.mit.edu/2017/Englis … c-le11.htm
When I used LinearStatic solver and 3D domain, I had a hard time initializing the temperature loads on the nodes. How can I load the temperatures on the nodes?
If I have to change the solver or the domain, to which ones do I have to switch?
Any help is appreciated.
Dear all , I am trying to model the non linear behavior of a buckled beam. My model requires a pre load for initializing the buckling and then applying a cyclic loading in order to observe the non linear behaviour. In abaqus benchmark, it is done by first applying an initial pressure to the structure and then restarting the process from the buckled mode . I know that restart module is bugged, but when I investigated I came across certain methods for this kind of analysis such as staggered solver. Is there a method in OOFEM that can serve my purpose? I am adding my benchmark below. Thanks in advance
benchmark:https://classes.engineering.wustl.edu/2009/spring/mase5513/abaqus/docs/v6.6/books/bmk/ch04s10anf86.html
Hello Everyone. I have been trying to use SLEPC with OOFEM and compiled the program accordingly. However, when I use SLEPC as the eigen solver and smtype 7, I seem to get this error. Can anyone help on how to resolve this problem?
Assembling stiffness and mass matrices
Solving ...
[0]PETSC ERROR: --------------------- Error Message --------------------------------------------------------------
[0]PETSC ERROR: Shift-and-invert requires a target 'which' (see EPSSetWhichEigenpairs), for instance -st_type sinvert -eps_target 0 -eps_target_magnitude
[0]PETSC ERROR: See https://www.mcs.anl.gov/petsc/documentation/faq.html for trouble shooting.
[0]PETSC ERROR: Petsc Development GIT revision: v3.14.5-729-g5947d95822 GIT Date: 2021-03-08 14:49:56 -0600
[0]PETSC ERROR: /home/deniz/build/debug/oofem on a arch-linux-c-debug named debian by deniz Wed Mar 24 09:14:46 2021
[0]PETSC ERROR: Configure options
[0]PETSC ERROR: #1 EPSCheckCompatibleST() line 139 in /home/deniz/Programs/slepc/src/eps/interface/epssetup.c
[0]PETSC ERROR: #2 EPSSetUp() line 355 in /home/deniz/Programs/slepc/src/eps/interface/epssetup.c
[0]PETSC ERROR: #3 EPSSolve() line 136 in /home/deniz/Programs/slepc/src/eps/interface/epssolve.c
[0]PETSC ERROR: #4 solve() line 154 in /home/deniz/Programs/oofem/src/oofemlib/slepcsolver.C
EngngModel info: user time consumed by solution: 0.00s
ANALYSIS FINISHED
dear nitramkaroh , the input i have written above is in a different computer. I am adding a similar input of which gives a different error yet they have a similar problem. I used Nonlinear statics instead. Since i believe i can use nonlinform 2 to use the displacements of each previous loadstep therefore i used it also but it does not create any answers.
hello nitramkaroh, i am grateful for your replies yet i seem to get stuck with this analysis so far. I now tried to use staticstructural.As:
keyhole.out
Test for SAE Keyhole 8 noded brick element
staticstructural nsteps 67
domain 2dPlaneStress
OutputManager tstep_all dofman_all element_all
ndofman 397 nelem 334 ncrosssect 1 nmat 1 nbc 4 nic 0 nltf 2 nset 5
node 1 coords 3 -1.551126218 1.101126218 0.1875
node 2 coords 3 -1.617417479 1.167417479 0.1875
....
planestress2d 1 nodes 4 21 246 244 245 nlgeo 1
...
planestress2d 334 nodes 4 246 1171 442 1086 nlgeo 1
SimpleCS 1 thick 1.0 width 1.0
IsoLE 1 d 0. E 15.0 n 0.25 talpha 1.0
BoundaryCondition 1 loadTimeFunction 1 prescribedvalue 0.0
NodalLoad 2 loadTimeFunction 4 Components 2 10.0 0.0
NodalLoad 3 loadTimeFunction 4 Components 2 0.0 10.0
ConstantFunction 1 f(t) 1.0
PiecewiseLinFunction 2 npoints 2 t 3 0. 1. 9. f(t) 3 0. 1. 1.
PiecewiseLinFunction 3 npoints 2 t 3 0. 4. 9. f(t) 3 0. 0. 1.
PiecewiseLinFunction 4 npoints 67 t 67 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 .. f(t) 67 10 10.0 -2.04104104104104104 -3.6466466466466466 2.1931931931931932 ....0.12212212212212212
I get the warning solution does not converge I am confused with this warning since from what i understand from the input manual static structual solver does not accept any solution parameters such as rtolv. Do you have any idea i can not take any results? For the non linear solver i find a function called "nlinform" do you have any examples how to use this function ?
As it turns out you indeed have a function for this purpose called nlgeo yet i am not quite sure if it works or not
testMetaSteps.out
playing with steps and metasteps
# rectangular specimen
#
# 4------------------3
# | |
# 1------------------2
#
#NonLinearStatic nsteps 67 nmsteps 1 nmodules 1 nlinform 2
LinearStatic nsteps 67 nmodules 1
#nsteps 67 controlmode 1 rtolv 0.001 MaxIter 2000 refloadmode 1
vtkxml tstep_all domain_all primvars 1 1 vars 2 1 4 stype 2
domain 2dPlaneStress
OutputManager tstep_all dofman_all element_all
ndofman 4 nelem 1 ncrosssect 1 nmat 1 nbc 3 nic 0 nltf 4
Node 1 coords 3 0.0 0.0 0.0 bc 2 1 1
Node 2 coords 3 4.0 0.0 0.0 bc 2 0 1 load 1 2
Node 3 coords 3 4.0 3.0 0.0 bc 2 0 0 load 2 2 3
Node 4 coords 3 0.0 3.0 0.0 bc 2 1 0 load 1 3
PlaneStress2d 1 nodes 4 1 2 3 4 mat 1 NIP 4 crossSect 1 nlgeo 1
SimpleCS 1 thick 1.0 width 1.0
IsoLE 1 d 0. E 15.0 n 0.25 talpha 1.0
BoundaryCondition 1 loadTimeFunction 1 prescribedvalue 0.0
NodalLoad 2 loadTimeFunction 4 Components 2 10.0 0.0
NodalLoad 3 loadTimeFunction 4 Components 2 0.0 10.0
ConstantFunction 1 f(t) 1.0
PiecewiseLinFunction 2 npoints 2 t 3 0. 1. 9. f(t) 3 0. 1. 1.
PiecewiseLinFunction 3 npoints 2 t 3 0. 4. 9. f(t) 3 0. 0. 1.
PiecewiseLinFunction 4 npoints 67 t 67 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 .. f(t) 67 10 10.0 -2.04104104104104104 -3.6466466466466466 2.1931931931931932 -0.7737737737737738 0.0 -0.943943943943944 ....0.12212212212212212
it makes no difference if i include this nlgeo function. It still fails to load its previous state and i get the same result as with nlgeo 0. Is it something wrong with this function? Or is there another way to include large deformations to the model?
thank you for your support nitramkaroh. It is true that i am using linear elastic material. Yet i would like to include finite deformation to my analysis therefore even though i am dealing with linear elastic material. As i mentioned above my problem is to analyse a transmission load. Can you please suggest a way for me to analyze that kind of loading history correctly. I am adding significant parts of my inp files below.
keyhole.out
Test for SAE Keyhole 8 noded brick element
IncrLinearStatic endOfTimeOfInterest 67 prescribedTimes 67 0 1 2 3.....66 nmodules 1
#LinearStatic nsteps 5 nmodules 1
vtkxml tstep_all domain_all primvars 1 1 vars 5 1 2 81 33 124 stype 1
domain 2dPlaneStress
OutputManager tstep_all dofman_all element_all
ndofman 397 nelem 334 ncrosssect 1 nmat 1 nbc 4 nic 0 nltf 2 nset 5
node 1 coords 3 -1.551126218 1.101126218 0.1875
...
planestress2d 334 nodes 4 246 1171 442 1086
simplecs 1 material 1 thick 1 set 1
IsoLE 1 d 7800 E 203e9 n 0.3 tAlpha 0.000012
BoundaryCondition 1 loadTimeFunction 1 dofs 2 1 2 values 2 0.0 0.0 set 4
NodalLoad 2 loadTimeFunction 2 dofs 2 1 2 components 2 1e6 0.0 set 2
NodalLoad 3 loadTimeFunction 2 dofs 2 1 2 components 2 -1e6 0.0 set 3
NodalLoad 4 loadTimeFunction 2 dofs 2 1 2 components 2 4200 0 set 5
ConstantFunction 1 f(t) 1.0
PiecewiseLinFunction 2 npoints 67 t 67 0 1 2 3 4 5 6 7 8 9....66 f(t) 67 0 1.0 -0.0410 ....
nitramkoh thanks for your reply. I know that i can use "PiecewiseLinFunction" for this purpose but i am not sure whether or not this gives correct results for a time dependent analysis. I believe it treats every loading condition on its own. Since when i use this function, for "0" loading at a certain point i get a zero value for everything from displacement to stress and this is incorrect since our geometry must have residual displacements and stresses from the loading in the previous time steps. This got me thinking that "PiecewiseLinFunction" takes into account only its current time step. My purpose is to model a loading in a course of time,such as a car part under the influence of transmission loading, and "PiecewiseLinFunction" seems to "take picture " of a geometry for certain loading conditions in only one time step and then combines these pictures back to back rather than assessing loading history . Is there a way to model this kind of history on OOFEM(Like a geometry under the influence of suspension or transmission loading )?
Hello all, I have been getting myself familiar with oofem for quite some time now. And I would like to use OOFEM for analyzing a time dependent loading case. For this purpose, I want OOFEM to use displacements calculated from the previous time step and apply a different loading condition to this altered system. So basically I want OOFEM to remember the previous state of the system. You can think this as running an OOFEM script “n ” times, n being the number of different loading conditions. I tried the “oofem -f context01.in.0 -c” command on your context01.in script to produce osf files and then used “oofem -f context01.in-r 1 “ to restart it from the 1st step. As used in the “context01.sh” shell script. Yet I could not manage to restart the solution. It generates binary osf files and when I run “oofem -f context01.in.0 -r 1” I get the error “Restoring context for time step 1.0 /Segmentation fault” and it generates empty .out files in return. Is there a problem with my script even though it is taken directly from the test files? Is it normal for .osf files to be in the binary format? Finally, can I use this restart function as a way to analyze time varying loading conditions? I am adding input file, osf and shell script as txt. (original file is located at tests/sm). Finally my desired model is given below.
https://www.efatigue.com/benchmarks/SAE … yhole.html
your shell file:
#
# this test check basic save/restore context functionality
#
OOFEM=$1
echo "target executable: $OOFEM"
pwd
echo "Command: $OOFEM -f context01.in.0 -c"
# run target on input and store context file
$OOFEM -f context01.in.0 -c
echo "Command: $OOFEM -f context01.in.0 -r 1"
# run target on the same file, but restarting from step 2
$OOFEM -f context01.in.0 -r 1
Best regards
Posts found: 16