electrodynamic forces on busbars

[tomatoma]

Hi,

in case of short-circuit inside or downstream an electrical cabinet (supplied from a low voltage 50/60 Hz grid), the electrodynamic forces amongst the conductors inside the cabinets can be relatively high.
Here is an example of a paper that deals with this subject:
https://www.studiecd.dk/cahiers_techniq ... ystems.pdf

Is there a way to estimate with elmer:

1. the short-circuit currents (these could be computed with other means and given as an input to elmer)
2. the electrodynamic forces generated by the short-circuit currents
3. the deformation of the busbars during the event
4. the forces acting on the busbars supports (isolators, metal brackets...)
5. during the event, the temperature of the busbars can raise by let's say 100-200 K:
5a) can elmer estimate the power losses and temperature rises of the busbars during the short-circuit event?
5b) the temperature increase can change the mechanical properties of the busbars. Can this be taken in account by the solver?

While the currents can be assumed to be independent on what happens inside the cabinet, the electrodynamic forces depend on the distance between the conductors and thus on their deformations. So forces and defomations of the conductors are a coupled problem. (if deformations are small compared to the distances amongst the conductors, than the deformations can be neglected for the sake of calculation of the electrodynamic forces).
Point 5 of the previous list goes into another level and could be neglected. Alternatively, the busbar temperature rise could be estimated in other ways, knowing the current and the resistivity of the conductors.

Has anyone had any experience with elmer on this kind of problems?
Are there any example available?

Thanks,
Matteo

[kevinarden]

viewtopic.php?t=8269

[tomatoma]

Hi kevinarden,

from what I understand, the post you linked deals with the electric field.
The forces between conductors are generated by the intercations between the currents and the magnetic filed generated by the currents themselves. The effect of the electric field is negligible. I believe the problem is magnetostatic + "mechanical". Forces are pulsating at twice the grid frequency. There can be mechanical resonances in the system, so the actual forces acting on the busbars supports are different than (can be considerably higher than) the electrodynamic forces.
The output of the simulation shall be an estimation of the stress (or forces) acting on the key elements of the busbar system and ideally an estimation of the deformation of the same elements (particularly of the busbars).

[kevinarden]

This project couples magnetic force to displacement, although the magnetic field is not electrical induced
but you should be able to go from electrical field to magnetic field to mechanical force
https://github.com/mrkearden/magnet

[raback]

Hi

The way to proceed is indeed to

1) Use AVSolver to solver for electric/magnetic fields
2) Use CalcFields solver to compute nodal forces
3) Some up nodal forces and moments for each piece.

Now 1) and 2) are shown in many test cases (as that of Kevin in force.sif). The step 3) is a minor technicality which I can give pointers to when I have a little more time. It involves components and some simple summation.

-Peter

[tomatoma]

Hi,

Please bear with me, but I'd like to better understand the three steps:

1) Use AVSolver to solver for electric/magnetic fields
Currents are non-sinusoidal time-varying waveforms which can be computed by other means (not by elmer): they depends on the caracteristics of the utility grid upstream the busbars under examination. So, as I understand, this is a transient type study and from this step we obtain the electric and magnetic fields as function of time. The deformation of the mechanical structure is not considered by the model during this step. Is my understanding correct?

2) Use CalcFields solver to compute nodal forces
The output of this step are the nodal forces acting on the conductors. Knowing the magnetic field and the current density in the conductors, the electrodynamic forces are computed at each mesh node of the conductors for each time instant considered during the transient analysis. Again, I believe the deformation of the mechanical structure is not considered to derive the nodal forces.

3) Some up nodal forces and moments for each piece.
The output of this step, as I understand, is a sort of integration on the conductor domain of the nodal forces.

The result of this study, as far as I understand, are the excitation forces as function of time acting on the conductors. From these forces (i think from the nodal forces) one should simulate how the mechanical structure responds to obtain deformations, stress and mechanical forces on some items of the structure (e.g. on the busbars and on the isolators). Note that these forces are different than those obtained in steps 2 and 3 above.

Thanks for your help

[raback]

Hi

Sorry, I was reading hastily. So if you want to really solve the displacements also you should apply the nodal forces directly for the elasticity solver. Here conforming mesh is assumed. For example see,

https://github.com/ElmerCSC/elmer-elmag ... e/case.sif

The key lines are

Code: Select all

  Displacement 1 load = equals "nodal force 1"
  Displacement 2 load = equals "nodal force 2"
  Displacement 3 load = equals "nodal force 3"

That associate the computed electromagnetic forces to the displacement solver in a discrete level.

-Peter