plate bending stress in Staad Pro

[senthil83]

Dear all,

i am using the Staad Pro software for the structural analysis,
i am herewith attaching my model of simple plate element fem model fixed at the both ends,

i just checked bending moment at the center of plate from Staad and manual calc for the beam fixed at both ends
but it is matching.

can anyone pl let me know how it is changing.

regards
senthil

 

[senthil83]

it is not matching with the maunal calc.

please help

 

[dirnad]

I haven't used Staad Pro but for general debugging:

Does it converge towards the correct value when you refine the mesh? If so then just refine until it doesn't change much.

Does the deformed shape look correct - ie shows that the ends are clamped not pinned, etc?

Are the reaction forces correct? Maybe you're not applying the same load that you think you are.

Is the distribution of stress qualitatively OK? Ie maximum positive in the middle on the outside of curve, negative inside, similar but opposite near the ends. If so then maybe the loads are too low/high.

Is the hand calc for an ideal plate which may be less accurate than a solid model especially if it's thick relative to it's length/width?

 

[JoshPlumSE]

What type of output are you using? Plate Forces which are based on the plates local axes? If so, then you have to make sure your local axes are pointed in the right directions. I don't think they are. But, it's difficult for me to tell because I cannot open the model in STAAD.

Also, are you looking at plate contours? If so, the local axis issues will become even more important. Contours may not display well if the local axes of your plates aren't consistent. After all, if the Mx moment for one plate means longitudinal bending and for some of the the adjacent plates it means bending about an axis 30 or 40 degrees off from that then what does a contour mean for those plates?

Some ideas:
1) Use corner forces instead (if those are even available in STAAD).
2) Switch from using Tri plates to quads and align them so that their local axes line up properly.

Caveat:
I actually work for a competitor of STAAD's (RISA Technologies). Just so you are aware of any personal bias I may bring to the discussion.

 

[IDS]

What do you get for the bending moment?

I don't have STAAD, but I imported the model into Strand7, which gave me a bending moment of about 0.0412 for unit load and unit span. That's about 99% of the beam theoretical value of 1/24 (0.04167).

Was the subdivision into 3396 random triangular plates done by STAAD? For practical applications rectangular elements are easier to work with and a much smaller number will give acceptable accuracy, at least away from discontinuities.


Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[senthil83]

thank u