axisymmetric

[jerrylucas]

Hi all,

I have a problem with the axisymmetric option. I explain:

I want to model a spire. I use plane13 with the axisymmetric option (I paint a circle which center is in x=50,y=50). I coupled all the nodes of the spire and I introduce a costant current. When I view the result, in the circle that represent the spire, the direction of the current is different in the right part compared with the left part. Are opposite.

To what reference do ANSYS the axisymmetric?

Thanks,

Jerrylucas

 

[cbrn]

Hi,
in axisymm, the y-axis MUST be the axis of symmetry, the x-axis is the "radial" one and the model MUST be contained in the first quadrant (i.e., x>=0 and y>=0).

Regards

 

[jerrylucas]

Hi,

all the model is contained in the first quadrant. When I going to launch the simulation, in the ANSYS Output Window appears the following text:

"Axisymmetric elements present. Nodal loads input with the F commands and the reaction forces are on a full circle basis. This differs from pre-ANSYS 5.0 Interpretation"

What it that mean?
I don´t want that in the spire appears reaction forces.

Regards

Jerrylucas

 

[Stringmaker]

I think what Ansys is attempting to tell you is that if you apply 5 lbs to a node that 5 lbs really acts on a 360 degree basis. Typically, it's better to use pressures to apply loads in situations like this. This is easily done by using P=F/A where A is the area of the surface (360* remember) and F is the total acting force.

I'm guessing that Ansys may have changed something in axisymmetric element formulation back in V5.0 that would cause some incompatability with newer 2D axisymmetric elements (eg. PLANE182) if used in parallel with each other.

 

[cbrn]

Hi,
the message you see is informative only and absolutely generic. Never mind, it only reminds you the "new" (new if compared to v.5, that is 8 or 9 years ago...) way in which the force loads are considered. In current version, if you input "1 [N]" on an axisymm model, it means "1 [N] over the full 360 [°] of the "real" model". In other terms, it's "like if" an "EXPAND" was made on 360 [°] and the load applied on this expanded model. The math reason behind that, I presume, is for compatibility with the Fourier decomposition of periodic non-axisymmetric loads (see "harmonic elements", or "axisymmetric models with non-axisymmetric loads").

Regards

 

[jerrylucas]

Hi,

Sorry CBRN for my ignorance.
But I'm working with current flow instead of force. In this case, in the rigth part on the circle modeled as spire, the current is in the z direction and in the left part of the same circle the current is in the -z direction (action-reaction). It is possible that in the circle that I've modeled the current flow migth be in the z direction and in the part that I haven´t modeled because I've used axisymm model the current migth be in the z direction?

Thanks for all

Jerrylucas

 

[cbrn]

Hi,
No, because the z-axis is the "tangential" one in an axisymmetric analysis (if you imagine it going "inside" the monitor when you are in the 1st quadrant, you should imagine it coming "towards you" in an hypothetic 2nd quadrant). In fact, your "desired" load is, as far as I understand it, axisymmetric as well.
See 2.13 "Axisymmetric Elements with Nonaxisymmetric Loads" for a detailed explanation of the "meaning" of the directions (and, in this "complicated" case, also of the harmonic indexes...).

Regards