Deep beam S&T with biaxial bending

[bkal]

A colleague of mine is designing a deep beam using S&T method. The beam also experiences some minor axis bending and he asked me how to deal with the biaxial bending. My initial thought is to make sure not to claim same reinforcement bars twice, but apart from that I could not think of anything else. We are continuing to look into this, but any pointers from this forum would be really helpful.

 

[KootK]

My initial thought is to make sure not to claim same reinforcement bars twice, but apart from that I could not think of anything else.

That's what I would do if the stresses from the minor axis effects are smallish-relative to those from the major axis effects.

I'm not sure that an STM model accounting for both actions is tractable without the help of AI. For all its awesomeness, STM has numerous practical limitations:

1) Tough to deal with combined actions.

2) No agreed upon method for dealing with deflections.

3) No explicit consideration of stability other than, perhaps, that of individual struts.

I find STM to be super fun to talk about, super useful to think about (detailing), but rarely of practical value in design office situations requiring execution.

 

[centondollar]

Reinforced concrete design is both art and engineering in the sense that RC behaviour is poorly understood even to this day. Anything more complicated than a column or beam is a hurdle to analyze, and many design methods (particularly for shear and bulky structures such as deep beams) boil down to strut-and-tie models combined with empirical testing with a very limited scope. For example, there still does not exist an empirically thoroughly validated and widely accepted standard method to design something as simple as a two-way reinforced concrete slab based on elastic plate moments (you will also find that most codes of practice make no mention of this); some common software downright ignore the twisting moment output of the FE plate model. Don't get me started on reinforced concrete shells or panels with very large shear loading.

You could avoid the biaxial bending or try to design it as a column. Typically, a deep beam should not be bent around its weak axis (much smaller bending stiffness) in the first place.

 

[bkal]

Thanks for your responses.
centondollar - I am intrigued and puzzled by your idea to treat it as column, I cannot envisage how to do it.

 

[centondollar]

Well, if its proportions make it more a wall than a beam, then the column calculation (I meant to write beam in biaxial bending") will not be very accurate. On the other hand, the reinforcement obtained by strut-and-tie models and beam theory are quite similar for deep beams with reasonable proportions (say, length/height > 3,...,5) and which are not thin enough to fail due to plate buckling or LTB.

If you don't design it as a beam in biaxial bending, the next resort is finite element analysis.