[Brayanashel]

How can exact stresses distributions (normal & shear stresses) along the left BC be obtained in a cantilever beam under tip shear load, as shown in attached Fig?

[Brayanashel]

The missing paper in the previous question can be found in below.
http://www.sciencedirect.com/science...68874X08000140

[Brayanashel]

Hi,

Is there any solver for shell & plate, please?

[bigphil]

Originally Posted by Brayanashel

Hi,

Is there any solver for shell & plate, please?

Hi,

Currently there are only continuum elements implemented within the OpenFOAM solid mechanics solvers, so there are no shell/plate/beam/membrane/etc elements yet; but it is possible to implement them using the Finite Volume method, see eg Ivankovic, A. & Demirdzic, I. (1994). Finite Volume Approach to Modelling of Plates. Computational Mechanics (PM me if you would like a copy).

As regards your previous question about application of BCs to the cantilever case, it is possible to apply all the BCs within the paper but you will need to code some yourself, in particular for the BCs with a fixed zero normal displacement and an applied parabolic profile shear traction: it would be very similar to fixedDisplacementZeroShear.

Philip

[sfmoabdu]

Originally Posted by bigphil

Hi Giamp,

Hmnnn for transient analysis the beam should never reach steady-state, can you check if a small bug-fix in TL solver makes a difference:
in writeFields.H, change this:

    rho = rho/J;

to this:

    //rho = rho/J;

Also, it would be interesting to check if the small stain solver elasticSolidFoam has same issues.

Philip

hello Philip, i have the same error

--> FOAM Warning :
From function eigenValues(const tensor&)
in file primitives/Tensor/tensor/tensor.C at line 170

but i can not find this line rho = rho/J in writeFields.H to fix !!!
the file writefields.H is that on the link
http://sourceforge.net/p/openfoam-ex.../writeFields.H

which does not contian this line!!

then the question what should I fix?!

[bigphil]

The change mentioned should be fixed in this file:
$FOAM_SOLVERS/solidMechanics/elasticNonLinTLSoldFoam/writeFields.H
You can see it here.

Though, I am not sure if this is related to your warning:

--> FOAM Warning :
From function eigenValues(const tensor&)
in file primitives/Tensor/tensor/tensor.C at line 170

Philip

[sfmoabdu]

ok Philip it is not the solver which I use, but i use elasticNonLinULSolidFoam and after some time steps when my beam is starting to give the desired deformation I got this warning and then it crashes.

--> FOAM Warning : 
    From function eigenValues(const tensor&)
    in file primitives/Tensor/tensor/tensor.C at line 170
    complex eigenvalues detected for tensor: (24.0327 -0.573962 7.72431 -84.7865 0.703912 -23.9374 -2646.74 -70.2494 2.14876)
Floating point exception (core dumped)

in some cases when i restart the simulation with different time step it goes on and crashes later. But at any case it gives this warning and then crashes ,,,, do you have any suggestion?!!

[davidsblom]

Hi Philip,

I have another question regarding the boundary conditions for the solid mechanics solvers. For an FSI run, is it possible to first apply a traction boundary condition, and at the end of the time step a time varying fixed displacement boundary condition?
I'm currently working with something similar to a geometric multigrid setup, and at the end of the time step I need to synchronize all the solid mechanics solvers.
Thanks a lot!

David

[bieshuxuhe]

dear Philip,

there is another way to use timeVaringFixedDisplacement boundary in icoFsiElasticNonLinULSolidFoam with no need for changing $FOAM_SRC/solidModels/fvPatchFields .
for fluid part I change U to Ufluid , and for solid part I change Usolid to U . after wmake , the timeVaringFixedDisplacement boundary works ok in icoFsiElasticNonLinULSolidFoam.


but I found another prolbem !
there is something wrong with moving mesh for fluid part when using timeVaryingUniformFixedValue boundary in motionU

    ground
    {
        type            timeVaryingUniformFixedValue;
        fileName        "$FOAM_CASE/constant/timeVsVelo.txt";
        outOfBounds     clamp;
        value           uniform (0 0 0);
    }

at some time step , the mesh of fluid don't move as "timeVsVelo.txt" required !
I have tried this boundary condition in icoDyMFoam , it works good .

I don't know what is wrong ? maybe it has something to do with pointer ?
I had heard that for strong FSI It's necessary to move mesh(fluid) back to previous point field in each subiteration followed by another update with the new point field, the pointers were switched back into the right position.
do you have any suggestion?
thanks

xuhe

[codder]

Hi Phil -

Can I limit crack origination to only from a specified patch?

Looking at ACP's crackingBiMatDcbLinear tutorial:

system/fvSolution

solidMechanics
{
...
    //crackPropagationFromSpecifiedPatches no;
    crackPropagationFromSpecifiedPatches yes;
...
}

constant/dynamicMeshDict

crackerFvMeshCoeffs
{
    crackPatch crack;
    openPatch  tractionFree;
}

So, does this mean that crack can now only propagate from tractionFree? Alternately, is specification of openPatch supposed to enough for this?

I'm seeing cracking from non-openPatch patches, so wondered about these configuration options...

Best, Eric

[Brayanashel]

Hi,

Could you please provide the following journal paper.

Ivankovic, A. & Demirdzic, I. (1994). Finite Volume Approach to Modelling of Plates. Computational Mechanics

Thanks.

[bigphil]

Originally Posted by Brayanashel

Hi,

Could you please provide the following journal paper.

Ivankovic, A. & Demirdzic, I. (1994). Finite Volume Approach to Modelling of Plates. Computational Mechanics

Thanks.

I will send you a copy if you PM me your email address.

Philip

[bieshuxuhe]

hi,
I know we could use sampleDict to get Usolid at patch when posting results from the two fsiFoam in solidMechanics .
but I wonder if it is possible to use libsampling.so in controlDict to get Usolid at patch ?
I have tried ,but only succeed when sampling fluid field .


thanks!

[iczajka]

There are slides by Robert L. Campbell and Bent A. Craven from 6th openfoam workshop. In their pressentation authors have mentioned about classes: structure and fsiInterface. Do you know, how to get such classes? Are they available somewhere in the internet for download?

[bigphil]

Originally Posted by iczajka

There are slides by Robert L. Campbell and Bent A. Craven from 6th openfoam workshop. In their pressentation authors have mentioned about classes: structure and fsiInterface. Do you know, how to get such classes? Are they available somewhere in the internet for download?

I have the ".tgz" file from that presentation; if you PM me your email address, I will send it to you.

Best regards,
Philip

[bigphil]

Originally Posted by bieshuxuhe

hi,
I know we could use sampleDict to get Usolid at patch when posting results from the two fsiFoam in solidMechanics .
but I wonder if it is possible to use libsampling.so in controlDict to get Usolid at patch ?
I have tried ,but only succeed when sampling fluid field .


thanks!

probe functionObjects should work to sample Usolid.

Philip

[hamsteri15]

I managed to run icoFsiElasticNonLinULSolidFoam from the foam-extend3.0 for a simple case of submerged hydrofoil under an ice sheet and the results are very similar to what I got with an in-house developed code. However, now that I'm trying to dig in more to the icoFsiElasticNonLinULSolidFoam code and since I don't have much experience with C++ or FSI some questions came up.

-First of all, why is the solid mesh not updated within the sub-iteration (outerCorr) loop in icoFsiElasticNonLinULSolidFoam.C?

-What does createStressPointMesh.H do and what is the StressMesh created for?

Petteri

[Jibran]

Hi Philip and everyone else,

I am using the solid mechanics solvers (elasticSolidFoam etc) on OpenFoam 2.3.0. The solver runs well but the problem arises when I try to plot the residuals with pyFoam by using pyFoamPlotWatcher.py.

I get the following error multiple times:

Time: 5

DICPCG:  Solving for Ux, Initial residual = 7.86012e-08, Final residual = 7.86012e-08, No Iterations 0
 PyFoam WARNING on line 138 of file /usr/local/lib/python2.7/dist-packages/PyFoam/LogAnalysis/GeneralLineAnalyzer.py : Problem doing timelines could not convert string to float:  
 PyFoam WARNING on line 138 of file /usr/local/lib/python2.7/dist-packages/PyFoam/LogAnalysis/GeneralLineAnalyzer.py : Problem doing timelines could not convert string to float:  
DICPCG:  Solving for Uy, Initial residual = 9.42587e-08, Final residual = 9.42587e-08, No Iterations 0
 PyFoam WARNING on line 138 of file /usr/local/lib/python2.7/dist-packages/PyFoam/LogAnalysis/GeneralLineAnalyzer.py : Problem doing timelines could not convert string to float:  
 PyFoam WARNING on line 138 of file /usr/local/lib/python2.7/dist-packages/PyFoam/LogAnalysis/GeneralLineAnalyzer.py : Problem doing timelines could not convert string to float:  
Time 5, Corrector 0, Solving for U using DICPCG, residual = 9.42587e-08, relative residual = 0

Time 5, Solving for U, Initial residual = 9.42587e-08, Final residual = 9.42587e-08, Relative residual = 0, No outer iterations 0
ExecutionTime = 137.95 s  ClockTime = 138 s
 PyFoam WARNING on line 138 of file /usr/local/lib/python2.7/dist-packages/PyFoam/LogAnalysis/GeneralLineAnalyzer.py : Problem doing timelines could not convert string to float:  
Max epsilonEq = 1.34163e-08
Max sigmaEq = 3212.97
ExecutionTime = 138.01 s

I guess the problem is maybe with the formatting of the log file since pyFoam is unable to extract the data.

Please note that pyFoam runs perfectly with the other solvers in the original OpenFoam release. The problem is only encountered with the OpenFoam Extend solvers. Also I am able to execute the command pyFoamClearCase.py successfully in OpenFoam extend.

I will really appreciate if anyone can help me in this regard

Jibran

[bigphil]

Hi Petteri,

Originally Posted by hamsteri15

-First of all, why is the solid mesh not updated within the sub-iteration (outerCorr) loop in icoFsiElasticNonLinULSolidFoam.C?

The solver uses an Updated Lagrangian (UL) approach: in the UL approach, the equilibrium equation is written in terms of the configuration (mesh) at the end of the previous time step.
UL is described a umber of times in literature e.g. Bathe Finite Element Methods, I can send you some FV specific literature if you would like about UL.

Originally Posted by hamsteri15

-What does createStressPointMesh.H do and what is the StressMesh created for?

StressMesh is the mesh for the solid.
mesh is the mesh for the fluid.

Best regards,
Philip

[bigphil]

Hi Jibran,

Yes I suspect you are correct that the format of the log is not correct for pyFoam.

It is very straight-forward to change output format, if we find what format it would like.

Best regards,
Philip