Elmer Discussion Forum

Hi Peter,

Thank you! I was able to resolve the issue. I now get mass as an output in a .txt file (using save scalars).

I am also looking to get stress - in particular, stress zz - as an output in a .txt file (using save scalars). I tried "stress 3" in the variables section, but I get this error "requested variable does not exist". How should I proceed?

Thanks,
Anjali

Also, in the linear elasticity solver, would it be possible to use Force instead of pressure?

Hi

How about "stress zz" ?

-Peter

Hi Peter,

I am still getting the same error for stress zz - "requested variable does not exist".

- Anjali

Peter,

If we are using displacement volume to find the mass, are we then assuming that the mass is going to increase with displacement as opposed to being constant? I compared the displacement volume x density from Elmer versus density x volume (hand calculations), the numbers were close - of the same order - but not close enough.

- Anjali

Hi

If you want to use save scalars then mass would be "operator 1 = volume" times the density. I think you can give volume "coeffcient 1 = density" to get the mass.

As forces go, you could compute stresses and the take an "boundary int" over the desired stress component. Or you could calculate the reaction forces by "Calcalate loads = True" and sum over the resulting nodal forces to get force on a boundary,

-Peter

Hi

In the paraview (vtu file) do you see "stress zz" depicted?

Basically the volume stays constant only when Poisson ratio is 1/2. However, the change in density is not accounted for since usually the effect is minimal.

-Peter

Hi Peter,

There was a minor mistake that I had made in save scalars - used "stress zz" instead of "stress_zz". This resolved the issue. Thank you!

In regards to displacement, would there be a change in density that would compensate the change in displacement volume, keeping the mass constant? Since density is a constant in Elmer, is that why there is a change in mass?

Thanks,
Anjali

Hello,

I am looking to reduce the discrepancies between cantilever beam numerical model and Elmer model results. I am using Elmer linear elasticity solver to determine mass, stress and displacement.

The displacement values compare well. But there are discrepancies in the mass and stress values.

Mass
Numerical model, mass = density x volume
Elmer save scalars, mass = density x displacement volume; since the density for both the numerical model and Elmer model is the same, is the displacement volume not the same as volume?

Stress
Numerical mode, stress = equation in picture file attached
Elmer save scalars - stress = max. stress_zz
It's not rounding off error, because I increased the decimal places to 12, and I still get the same discrepancy in the stress values.

Thank you for any information you can provide.

- Anjali

Those stress results is a close as any FEA code will get to that numerical solution due to Poisson ratio effects in the fea that is not in your numerical solution. Poisson ratio causes the body to deform in the orthogonal directions when deformation occurs in the loaded direction. Your numerical solution is not including these orthogonal deformations due to Poisson ratio. You would get closer to the numerical solution if you made Poisson ratio zero, or nearly zero since some codes can't handle zero.

In solid mechanics I don't not think there is a change in volume or density so the mass should be constant. The body deforms and changes shape, but mass is neither created or destroyed.

Hi Kevin,

Thank you for your response. In regards to the mass discrepancy, for the same dimensions (L,w,t) and density, shouldn't the Elmer mass be the same as Numerical model mass?

- Anjali