cyclic symmetry

Moderator: GiD Team

Post Reply
sreelatha
Posts: 7
Joined: Tue Dec 15, 2015 6:37 am

cyclic symmetry

Post by sreelatha »

Hello
i just want to know how to give the condition cyclic symmetry for segmented ring. From two segments with opposite tangential polarization, cyclic symmetry condition has to be given such that the geometry considers the whole ring by 360 degree rotation of those two segments. Also how to check that after assigning condition whether the complete ring GiD has considered or not from cyclic symmetry condition. Y-axis is the axisymmetry for the GiD version we have been using.

Thank you
sreelatha.
User avatar
escolano
Posts: 1922
Joined: Sun Sep 05, 1982 10:51 pm

Re: cyclic symmetry

Post by escolano »

I don't understand exactly the kind of feature that do you require from GiD.
I will comment some things that maybe are related with it.
A GiD geometry or mesh don't know it its shape has revolution symmetry or not.

You can create for example a segment from the coordinates (1,0) to (2,0) and create with the copy tool, by revolution of this line, a extrusion of 6 copies each 30 degrees around a center like (0,0) in 2D.

Code: Select all

Mescape Geometry Create Line 1,0,0 2,0,0 escape
Mescape utilities Copy Lines DoExtrude Surfaces Mcopy 6 Rotation 0,0,0 TwoDim 30 1 escape
revolved_surfaces.png
revolved_surfaces.png (13.29 KiB) Viewed 8183 times
not_revolution_mesh.png
not_revolution_mesh.png (21.34 KiB) Viewed 8183 times
This shape is a revolution shape, but if do you generate a unstructured mesh it won't be of revolution.

Setting structured quadrilateral t iis possible to obtain a 'near-to-revolved' mesh. In fact the inner nodes are interpolated with a general algorithm, its location won't be exactly radial.
approximatelly_revolved_structured_mesh.png
approximatelly_revolved_structured_mesh.png (16.53 KiB) Viewed 8183 times
About 'cycling conditions' I don't know what do you mean.
In GiD is possible to create a 'separated contact surface' between two 'similar curves' to force its mesh to be 'similar' (a one-to-one relationship), you can force this 'contact surface' to generate line elements joining pairs of related nodes.
This is used sometimes to force 'periodicity' of the mesh for some simulations.

You can also attach GiD conditions to mark some parts of your geometry/mesh to define your simulation.

If this don't answer your question you must explain it a little better, e.g. with some sketch or your GiD model and what do you want to have written in the input file for your solver.
sreelatha
Posts: 7
Joined: Tue Dec 15, 2015 6:37 am

Re: cyclic symmetry

Post by sreelatha »

whatever the geometry you have considered is half of the segmented ring. when we give translation symmetry condition, GiD will consider the complete ring and shows the results for complete ring.
i.e; condition -> displacement -> y-axis -> clamped
this is condition is to be given to th egeometry you have sent.

In the same way, what is the condition i have to give to run with two segments only such that GiD considers the complete ring.
sreelatha
Posts: 7
Joined: Tue Dec 15, 2015 6:37 am

Re: cyclic symmetry

Post by sreelatha »

whatever the geometry you have considered is half of the segmented ring. when we give translation symmetry condition to the gemetry you considered, GiD will consider the complete ring and shows the results for complete ring.
i.e; condition -> displacement -> y-axis -> clamped
this is condition is to be given to th egeometry you have sent.

In the same way, what is the condition i have to give to run with two segments only such that GiD considers the complete ring.
Attachments
Half part of the segmented ring<br />Condition -&gt; displacement -&gt; y-axis -&gt; clamped<br />This condition is to be given to the geometry, such that GiD considers the complete ring by translation symmetry
Half part of the segmented ring
Condition -> displacement -> y-axis -> clamped
This condition is to be given to the geometry, such that GiD considers the complete ring by translation symmetry
p1.jpg (22.41 KiB) Viewed 8162 times
2 segments of the ring<br />What condition to be given for the end lines such that GiD considers the complete ring by symmetry?<br />This is to reduce the problem size and run time.
2 segments of the ring
What condition to be given for the end lines such that GiD considers the complete ring by symmetry?
This is to reduce the problem size and run time.
p2.jpg (9.94 KiB) Viewed 8162 times
User avatar
escolano
Posts: 1922
Joined: Sun Sep 05, 1982 10:51 pm

Re: cyclic symmetry

Post by escolano »

I think that you are confusing between GiD and a problemtype+simulation code
GiD is a neutraL pre/postprocessor.
A problemtype is a collection of customization files plus a simulation code to do a specific kind of simulation

The conditions are defined by the problemtype, not by GiD. Then GiD doen't know nothing about 'clamped displacements' or 'symmetry boundary conditions'.
When to you load a problemtype (Data->problem type) then the names and fields of conditions, materials, etc are loaded in GiD to be attached to geometric/mesh entities.

I don't know what is the problemtype that you are using and with are the conditions defined by them and its meaning.
It seems that you are using some kind of 'structural analysis solver'.
And it seems that you want to simulate only some segments of the ring and require to set for the simulation that the displacement is radial only in a couple of boundary lines (set a 'null displacement in a y-local direction)
I don't know if it exists in your problemtype this condition to set forced displacements in local axes (from the point of view of GiD could be implemente with a condition 'over lines' with a field of kind '#LA# allowing to select the automatic local axis of a line/surface or other defined axis)

And from the postprocess point of view it is possible to calculate for example only 1/8 of the ring and in GiD postprocess use the copy tool create the other parts, including its results, to visualize the full ring.
sreelatha
Posts: 7
Joined: Tue Dec 15, 2015 6:37 am

Re: cyclic or circumferential symmetry

Post by sreelatha »

Hello Mr. Escolano
am again stucked with the same problem of how to get full segmented ring performance by modelling two segments only. watever u have suggested in the last post i have tried everything. but i couldn't get it.

we are using ATILA problem type as a solver and doing structural analysis. In this problem-type, for translation symmetry and to avoid structural translation displacement we will be giving the condition

displacement -> clamped -> along required axis assigning condition.

so can u please help me in assigning condition for two segments such that it takes radial symmetry and also the structure should not displace from its position.
User avatar
escolano
Posts: 1922
Joined: Sun Sep 05, 1982 10:51 pm

Re: cyclic symmetry

Post by escolano »

In fact, I think that probably you cannot simulate the full ring only with one or two segments, by symmetry,
because to do it not only the geometry must be symmetric, but also all boundary conditions and forces must be symmetric!!
e.g: a pipe with inner pressure is a revolution-symmetric case.
I think that your case does't has symmetry because of your clamped displacements (you said nothing about applied forces, are they also symmetric?)

I had a look to the problemtype Atila 3.0.27 and it has a condition-line displacement that could be set in each direction x y z to a forced value or to clamped (value=0), but it seems that is only allowed in global coordinates, not in local coordinates (e.g. to fix the movement along the radial or tangential directions)

In any case this is not a question to be asked to this GiD forum, it must be asked to the Atila developers.
Post Reply