Spacing of torsional longitudinal rebar

[WARose]

I am reviewing a existing beam/girder. It's about 7' deep. They have more than enough longitudinal steel (for torsion) anchored at the joint. But what worries me is: there is about a 2' gap between [longitudinal] bars near about mid-depth. (This is not counting skin reinforcement which is not anchored.)

ACI 318 limits spacing to 12" (around the perimeter) for such steel........so, should I worry about this? Thoughts?

 

[KootK]

How wide is this bad boy?

 

[WARose]

About 5 feet.

 

[Tomfh]

Spacing seems a bit excessive for tensile steel, but spacing rules are more about crack control etc aren’t they? Are there cracks?

 

[WARose]

Are there cracks?

Don't know. May have to go out there and have a look.

 

[KootK]

It's distributed reinforcement meant to match what we envision to be a distributed phenomenon. I'd not be willing to damn an existing beam to comply with a spacing that's surely pretty arbitrarily established to begin with. It sounds as though you've got the sides divided up into thirds by the anchored bars. That doesn't seem unreasonable to me.

 

[rapt]

Are you sure it is for torsion. Longitudinal reinforcement for torsion is normally required in the corners of the torsion ties. Not distributed over the side face.

I agree side face reinforcement should not be grater than about 1'

 

[WARose]

Are you sure it is for torsion. Longitudinal reinforcement for torsion is normally required in the corners of the torsion ties. Not distributed over the side face.

Sect. 11.5.6.2 of ACI 318-11 says that The longitudinal reinforcement required
for torsion shall be distributed around the perimeter of
the closed stirrups with a maximum spacing of 12 in.

In the commentary, it sounds like they prefer the extra rebar (in addition to what is needed for flexural) near the centroid.

 

[Agent666]

Maybe look at the code version at the time of original construction, see if it would of complied. The 12" is probably a matter of good detailing and not specifically saying you need it for strength provided you have some distributed reinforcement and the required stirrups. If the numbers worked out but the spacing wasn't satisfactory and the original code said this was ok then I think you'd find 1001 other buildings in the same situation. Members don't fail because the code suddenly changed. Sometimes though changes are a matter of fixing deficiencies so you might want to look into why it may have changed.

Beam may have never complied is another option. Similarly without knowing the reason for the torsion, the original designer may have followed the compatibility vs equilibrium torsion requirements if the requirements for its consideration were satisfied.

Most codes have the distributed recommendations/requirements, I'm guessing though as rapt noted the corners see a higher stress than those along the middle of the sides.

 

[KootK]

I have one additional insight to add to this conversation that will, admittedly, be of little practical value.

Longitudinal torsion rebar is really and indirect way to provide distributed, longitudinal clamping to our imagined network of torsion struts. Conceivably, there could be other ways to accomplish that:

1) For a beam spanning between, say, two gravity dams, the clamping force could come about via the axial rigidity of the supports such that no longitudinal rebar is required.

2) For the very common case of torsion reinforcement embedded into an adjacent member/column, that embedment would likely result in thee clamping force from concentrated reinforcement strutting back to the beam in question such that the clamping force wound up being just as distributed as it would have been were the reinforcement not concentrated. I'd be inclined to say that hooked bars at 24" o/c and developed 16" into a column would provide clamping just as distributed as bars 12" oc terminated at the face of support with u-bars etc. And, certainly, this benefit would be available at all interior locations where the longitudinal bars would be generously developed past the cross section being considered.

 

[rapt]

WARose

Interesting, Australian code says

"As close as practicable to the corners of the cross-section".

Eurocode agrees with ACI. CSA specifies a minimum bar diameter for the corners but the rest is distributed.

I will have to find out why!

 

[Settingsun]

Rapt, it might be worthwhile tracking how old the corner reinf requirement is. I suspect it matches the space truss assumption. I wonder what the difference is to crack control.