ANSYS Meridian stress

[Kanterbrau]

Hello, I'd like to know how to obtain the meridian stress of a structure, in a circular part, but its shape is not regular, so with a polar coordinate system, it's impossible. I can only have the circumferential/hoop stress, normal to y!

Many thanks

 

[Kanterbrau]

I'm sorry I wanted to say cylindrical and not polar!

 

[flash3780]

You'll want to set up a cylindrical coordinate system and scope your results to that coordinate system. Ansys can do that.

I think that I already answered this question for you, though. You'll simply have to create a cylindrical coordinate system to scope the results to.

If you're using Ansys Workbench it's fairly straightforward. Click on "Coordinate Systems", "Right Click>Insert>Coordinate System", and select "Cylindrical Coordinate System".

Now, to output a solution in cylindrical coordinates, click on "Solution", "Right Click>Insert>Stress Component", and select the coordinate system you just created.

Hope that gets you squared away.

 

[Kanterbrau]

Yes you did indeed! But for the hoop stress, no problem, I can have that, but for the meridian stress, I need to have a coordinate system which follows the shape of my part! Meridian stress cannot be calculated in a cylindrical system!!
Look at this:

http://www-mdp.eng.cam.ac.uk/web/li..._dvd_only/DAN/pressVessels/shells/shells.html

 

[flash3780]

I'm sorry, I misread your original post. You're looking for membrane stresses, then. You can use element tables to output the element membrane stresses and bending stresses. However, they will be in the local element coordinate system, which may or may not be in the desired direction. To output element stresses, look at the ETABLE command in the ANSYS help. You'll have to use Ansys/Classic to retrieve the element data.

Is the direction important? Are you working with an anisotropic material? Can you give more details about what you're working on?

If you're interested in a particular direction that cannot easily be defined by a coordinate system, perhaps the best thing to do would be to output the stress components in an available coordinate system and then compute the stresses in the desired direction with a simple coordinate rotation.