Torsion Stiffness Modifier Code Reference 3

[NewbieInSE]

Dear Engineers,
Is there any reference from any Code such as ACI, regarding totally neglecting the Compatibility Torsion by distributing it to the adjoining slab, and beams perpendicular to each. In other words, is it allowed to redistribute Compatibility Torsion without any special detailing, except for taking its (redistribution) effect in designing adjoining members?

 

[HTURKAK]

Calculate the cracking torsion Tcr as per ACI 318 Table 22.7.5.1. If the torsion is less than 0.25*Tcr,
the effect of torsion will be ignored. However the cross-section shall be checked as per 22.7.7 ( Cross-sectional limits).

 

[NewbieInSE]

HTURKAK. I understood what you meant.
But I did not mean it. What you said is neglecting torsion if under threshold torsion.
What i wanted to know is whether I can complete redistribute Torsion, resulting in no special long and shear reinforcement that would be required for solely torsion.

 

[JoshPlumSE]

Section 22.7.1 of ACI 318-14 is all about this subject. In summary:

If the demand torsion is greater than the cracking torsion and the demand torsion is REQUIRED to maintain equilibrium then it MUST be designed for (22.7.3.1)

However, if you have a statically indeterminate structure, then you can reduce the demand torsion down to the cracking torsion and re-distribute that moment into other members. There is some good commentary on this as well. 22.7.3.2 and 22.7.3.3

 

[slickdeals]

I don't think there is a factor for "torsional cracking". I usually set it to 0.01 and run the analysis and provide the reinforcement required for threshold torsion in the girder with some extra stirrups at the ends. That way, I know the positive moment in the beam framing into the girder is higher than what it might need to be.

 

[HTURKAK]

HTURKAK. I understood what you meant.
But I did not mean it. What you said is neglecting torsion if under threshold torsion.
What i wanted to know is whether I can complete redistribute Torsion, resulting in no special long and shear reinforcement that would be required for solely torsion.

I just see your comment.E-mails from this forum are always falling in junk mail box.I understand what you meant..It is not possible totally ignore the torsion for statically indeterminate structures however,the code ACI 318-14 suggests ( copy and paste from R22.7.3 ' (b) The torsional moment can be reduced by redistribution of internal forces after cracking (22.7.3.2) if the torsion results from the member twisting to maintain compatibility of deformations. This type of torsion is referred to as compatibility torsion.'..).
IMO, the procedure for indeterminate structures is,
i) Calculate the cracking torsion without considering the effect of torsional reinforcement as per ACI 318 Table 22.7.5.1.
ii) If the torsion is less than 0.25*Tcr, the effect of torsion can be ignored,

iii) 22.7.7.1 is implicit upper limit to the torsional resistance of a concrete beam. Check the Cross-section for 22.7.7.1. if not satisfied, increase cross section .

 

[NewbieInSE]

I will go through the articles referred by you all. I will come back after that. Thanks all.

 

[KootK]

In other words, is it allowed to redistribute Compatibility Torsion without any special detailing, except for taking its (redistribution) effect in designing adjoining members?

1) I suspect that there may be some terminology confusion in that statement but the only way to find out is to just dive and see...

2) As I understand it, ACI basically gives you two paths for dealing with torsion:

a) Make a realistic estimate of torsional demand based on stiffness etc and deal with the torsional forces that implies, either as:

i) a torsionally uncracked member (0.25 Tcr) OR;

ii) a torsionally cracked member reinforced using the 3D truss analogy.

Accurate estimates of cracked torsional stiffness are hard to come by but, in general, you can count on stiffness dropping off massively after cracking (hence slickdeal's approach). I have most of the seminal ACI docs that deal with that and those references indicate that cracked stiffness modifiers on the order of 0.05 to 0.20 for many practical situations.

b) Provide enough torsional reinforcement in your member such that torsional cracking can develop and allow the redistribution of torsional actions to other parts of the structure without inducing dire consequences in the process. The most common "dire consequence" is basically a crap ton of torsion cracking near your beam supports that would be unsightly and possibly compromise shear resistance.
Admittedly, a torsional crack is quite different from a shear crack and the degree of interaction between the two is limited.

3) You can redistribute your realistic estimate of torsion when that torsion is not required for equilibrium. You cannot, however, redistribute the compatibility torsion. The compatibility torsion demand is unrelated to the actual torsional demand (realistic estimate) and is simply the value intended to force you into providing an amount or torsional reinforcing that would allow your member to torsionally crack, rotate, and resdistribute its torsion without inducing the aforementioned dire consequences. So there is no third option which you seem to have suggested as, effectively: redistribute the actual torsion and do not design for compatibility torsion.

Do let me know if I've misinterpreted the true meaning of your statement.

 

[KootK]

Maybe a useful way to envision this is as follows: compatibility torsion design leads to providing the minimum torsional reinforcement required for the rotation of a beam having had its torsion redistributed to be compatible with the notion of that beam not cracking to bits torsionally. That's really the "compatible" in compatible torsion I feel.

 

[slickdeals]

The following clarification has been added in ACI 318-19. However, I don't think I agree with the statement about modeling the torsional stiffness of edge beams (especially in compatibility torsion scenario).

 

[KootK]

However, I don't think I agree with the statement about modeling the torsional stiffness of edge beams (especially in compatibility torsion scenario).

Agreed. That might actually be an outright error given that it seems to be implying that an edge beam is an example of equilibrium torsion. Some might be but the overwhelming majority are not. And your classic edge beam is pretty much the most useful place to make use of compatibility torsion. Not that we get a lot of edge beams to begin with these days...