Risa3D deep concrete beams 1

[David Hanson]

Are Risa3D solid elements appropriate for designing concrete deep beams? The ACI strut and tie method is very useful, but it can be quite time consuming. I just don't know enough about Risa's finite element analysis to be confident using it here. Would the non-linear strain be modeled correctly?

 

[JoshPlumSE]

David -

I believe this is the case. But, you have to be cautious of how you mesh the deep beam. I believe I put together the original verification problems for the solid element. It was to verify that the results of a beam (not deep) made out of solids matched the flexural results we know are correct. IIRC, it was highly dependent on the number of element through the height of the beam.

I believe in the attached files, I use a 4 x 3 mesh. So, the vertical deflection (where we have used 4 elements through the cross section height) matches very well. However, if we tested it in the weak direction (where we only had 3 elements through the width), the deflection would NOT have been good.

https://files.engineering.com/getfi...4&file=solid_element_verification_problem.pdf

https://files.engineering.com/getfi...7-a225-7dfeea5bd8d5&file=Solid_Cantilever.r3d

Caveat: I used to be the VP of engineering at RISA and worked there for 16 years in the Tech Support / Training / Documentation / Verification of their various programs. However, I left (was pushed out) in late 2017 when the original owner retired and sold the company to some foreign corporate entity that seemed to have little respect for what made RISA software so good. In fact, I can only think of 3 technical employees that still remain from the people I worked with. Debbie (though she's mostly sales / management now), Adam and Brighton (who are developers). Regardless, I now work for their biggest competitor (Computers and Structures, Inc). Hence, my opinions on their software may be a little biased. And, may even be a bit out of date because I haven't really used their programs for about 4 years.

 

[larcix]

The strut and tie is necessarily a simplifying assumption regarding concrete behavior. If you build a full solid model of the beam with sufficient granularity, you should achieve more accurate results.

Having never done anything with solid elements or reinforced concrete in RISA, I'm well out of my wheelhouse here. But, this FEM software is relatively powerful, so it should have no problem modeling this.

EDIT: A way to check for 'sufficient granularity' is as you increase the number of elements, are the results converging? Or are they still changing a lot? If they keep changing, the model itself may not be the best representation of the situation, and updating the mesh size is doing more to the model than just refining the results.