Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations waross on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Solid[95] element with curved edges 2

Status
Not open for further replies.

aminjamali

Civil/Environmental
Sep 28, 2008
62
0
0
CH
I want to simulate a one inch diameter steel rod with only four elements across its section. I wonder if it is possible to define a Solid[95] element with curved edges? I could use a fine mesh so that the the curvature of the rod could be fairly simulated, but I don't want to do that, I thought there might be a way to define elements with defined curved edges?

Thanks for your help
Amin
 
Replies continue below

Recommended for you

Solid 95 is 3 node per side element so u can make 4 elements on section. But this mainly depends on the accuracy u want. Is quarter of a circle simulated by only one element side is enough?.. And by the way why u want only four in a section ??
 
The figures of Ansys help on Solid95 gave me an impression that a Solid95 element can have curved edges (I wanted to simulate one quarter of the steel bar with only one element and because of run time problems.) What I am understanding form your note is that Solid95 can not have curved edges and you recommend using more elements to cover the curve of the circle?

 
Hi,
No, it's not a matter of SOLID95 not being able to "cover" curved edges.
SOLID95 IS a "parabolic" element, its edge-functions are 2nd-order polynomials, so it can approximate curved edges fair well...
NOW, recall basic plane geometry, and consider that you want to approximate an arc of circle with a parabola. The parabola will give exact approximation (i.e. null positional error) only in three points: the first corner node, the second corner node and the midside node. Everywhere else, the curve described by the parabola is obviously different from the curve described by the arc of circle.
Bagarbilla's point is thus the following one: is this approximation acceptable for you? By standard meshing, Ansys tryes to keep the arcspan angle within 25° for a 2nd-order element, you want this to become 90°: is that coherent / acceptable for your analysis?

Regards
 
Status
Not open for further replies.
Back
Top