Hanger Stirrups in Beam-to-Girder Joints 1

[NeilL286]

My co designers at office just determine if the girder portion where beam frames into it has sufficient V to resist the shear of the beam. They assume the reaction from the floor beam is more or less uniformly distributed through the depth of the interface between beam and girder. But in references, there is this compression strut forming from beam to lower part of girder which has to be provided for by means of hangers stirrups inside the joint that serve as tension ties to transmit the reaction of the beam to the compression zone of the girder, where it can be equilibrated by diagonal compression struts in the girder. And so these hanger stirrups are supposed to be put inside the joint itself (see illustration provided). Do you put them too, as not all follow it?

 

[NeilL286]

No one designs beam-beam joints with internal stirrups here? Maybe most misses the crucial aspect of it. Here are the exact details.

According to the book "Design of Concrete Structures" by Arthur Nilson, David Darwin Charles Dolan (and illustration attached):

"Commonly in concrete construction, secondary floor beams are supported by primary girders, as shown in Fig. 11.13a and b. It is often assumed that the reaction from the floor beam is more or less uniformly distributed through the depth of the interface between beam and girder. This incorrect assumption is perhaps encouraged by the ACI code "Vc+Vs) approach to shear design, which makes use of a nominal average shear stress in the concrete, vc=Vc/Bw d, suggesting a uniform distribution of shear stress through the beam web.

The actual behavior of a diagonally cracked beam, as indicated by tests, is quite different, and the flow of forces can be represented in somewhat simplified form by the truss model of the beam shown in Fig. 11.13c (Ref. 11.7). The main reaction is delivered from beam to girder by a diagonal compression strut mn, which applies its thrust near the bottom of the carrying girder. Failure to provide for this thrust may result in splitting off the concrete at the bottom of the girder followed by collapse of the beam.

Proper detailing of steel in the region of such a joint requires the use of well-anchored "hanger" stirrups in the girder, as shown in Fig 11.13 a and b, to provide for the downward thrust of the compression strut at the end of the beam (Refs. 11.8 and 11.9). These stirrups serve as tension ties to transmit the reaction of the beam to the compression zone of the girder, where it can be equilibrated by diagonal compression struts in the girder. The hanger stirrups, which are required in addition to the normal girder stirrups required for shear, can be designed based on equilibrating part or all of the reaction from the beam, with the hanger stirrups assumed to be stressed to their yield stress fy at the factored load stage".

Is this mentioned in the ACI or Eurocode at all? How many implement it?

 

[frv]

What you are referencing, I believe, is "Hanger" reinforcement. It is not covered in the code, but should be. Check out MacGregor for more information. He actually provides design guidance.

 

[NeilL286]

It's not in the codes and even my colleagues haven't designed internal stirrups in the beam-beam joint. But I noticed hairline cracks underneath it. Please see.

http://img713.imageshack.us/img713/7444/zdz1.jpg

Owing to unfamiliarity with 45 degree compression struts in the joint. Colleagues can't state definitely whether it is lack of internal stirrups. The beam was just under service load and I'm sure it's not overload.

Has anyone noticed cracks like it too? Is it common?

 

[JAE]

I don't think this is really an issue in design or construction. There's been thousands/millions? of beam to beam joints designed without hanger stirrups all over the world and I've yet to hear of or see any shear failure at the joints.

The mechanism for failure that is implied is a rupture through the side of the supporting beam. Usually supporting beams have a lot of longitudinal reinforcement where a shear friction mechanism can develop and avoid a "bite out of the side of the beam" failure.

 

[NeilL286]

Quantify "a lot of longitudinal reinforcement". For instance, in midspan of a supporting girder are 3 #3 (3/8") main bars with a beam with 2 lower bars of same size framing into it. Supposed the 2 beam bars only rest on 2 of the girder lower bars. Won't the 2 girder bars reach yielding strength in horizontal orientation or buckle.

You may reason the girder top bars are also connected to the beam framing into it with bent up bars. But the bent up bar will not help to share the load since the load distribution of secondary beam shear on main beam is assumed to be dispersed at 45 degree on either side .

The top bent is usually for bond and anchorage requirement. So how would the 2 lower girder bars support the load of the entire beam and supported slabs?

We usually compute the yielding of a bar in vertical orientation. How about sideways or horizontal. Won't the yielding be the same as vertical oriented?

 

[IDS]

It's not in the codes and even my colleagues haven't designed internal stirrups in the beam-beam joint. But I noticed hairline cracks underneath it.

I don't think this is really an issue in design or construction. There's been thousands/millions? of beam to beam joints designed without hanger stirrups all over the world and I've yet to hear of or see any shear failure at the joints.

To quote from the Australian Bridge Code, AS 3600:

8.2.11 Suspension reinforcement
If forces are applied to a beam in such a way that hanging action is required, reinforcement or tendons shall be provided to carry all of the forces concerned.
Where, for some situations, suspension reinforcement is required, it shall be in accordance with the procedures given in Appendix D.

If I recall correctly, similar provisions were introduced around 25 years ago, in response to real cracking problems.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[IDS]

AS 3600 (Australian Concrete Structures code) on the same subject:

8.2.11 Hanging reinforcement
Loads applied to a member other than at the top chord of the member shall be transferred to the top chord, within the load application region, by the provision of hanging reinforcement of area consistent with strut-and-tie modelling.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[hokie66]

This depends a lot on the geometry of the connection. Arthur Nilson is certainly a respected source, and he is correct in pointing out the potential problem. I agree with JAE that hanger stirrups within the width of the supported beam is not common practice, except perhaps in high seismic areas. Maybe the reason they don't crack in practice is the clamping action of the longitudinal bars in the supporting beam as JAE suggested, although there might be little top reinforcement near midspan. Or maybe the flanges, along with the stirrups just outside the beam, carry the load into the supporting beam. Structures are often smarter than their designers.

 

[JAE]

IDS - I'm not familiar with your code - but those quoted provisions appear to be talking about load applications along the length of a beam - such as a spandrel beam with a ledge for brick veneer where the load along the beam is applied on the lower half of the beam. That doesn't appear to apply to a simple intersecting beam condition. I would agree that for spandrel/ledge load applications that hanger steel is needed along the beam length.

My experience with many beam-beam connections tells me this isn't a problem.

NeilL - I never would use #3 bars in a beam for flexural reinforcement - for main girders carrying other joists or beams you would typically see heavier reinforcement such as #6 to #9 bars (3/4" to 1 1/8" bars).

 

[IDS]

JAE - see attached extract from Appendix D of AS 5100 for typical situation they are talking about; i.e. the intersection of longitudinal beams with transverse beams transferring the load to a support.

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[JAE]

That appears to be more in depth than anything ACI 318 has provided...in fact I don't think that 318 has any reference to hanger stirrups for conditions of beam to girder connections that I'm aware.

 

[NeilL286]

I think we missed something. Internal Hanger stirrups are only required if V (shear) at the end of the supported beam is beyond a certain threshold that can cause diagonal cracking. Dolan Et Al mentioned "Hangers will also be unnecessary if the factored beam shear is less than 0.85Vc (as is usually the case for one-way joists, for example), because in such a case diagonal cracks would not form in the supported member. The predictions of the trust model would thus not be valid, and the reaction would be more nearly uniform through the depth."

Another author Mcgregor stated " These provisions can be waived if the shear, Vu2, at the end of the supporeted beam is less than 3*sqrt(fc')bd, because inclined cracking is not fully developed at this shear.

What I wish to know is what is the provision if the supported beam (beam framing into girder) has stirrups and Vs resistance is added to Vc. Would the provisions for hanger stirrups still hold?

In other words. Would compression fan still be form for beam with sufficient stirrups that prevent the diagonal crack widths from opening, or would the mere presence of the diagonal crack widths irregardless of stirrups produce compression fan that dispensed 45 degrees to the girder bottom?

 

[NeilL286]

The drawing of the compression fan attached (from Mcgregor presentation of the 1984 Canadian Concrete Code 6-29) illustrated my description above. If there is no compression fan (no diagonal inclined cracks). Would the load still be dispensed 45 degrees toward the bottom of the girder or would the reaction be more or less uniformly distributed through the depth of the interface between beam and girder? This is important because if it is, then if our loads are below the concrete Vc strength, then we can avoid the hanger stirrups. Or would the beam framing into the girder stirrups Vs add to its total V (but then stirrups are said to be functional only after the concrete cracks)?? Would someone be so kind as to clear this up?

or

http://imageshack.us/a/img13/8156/gdyd.jpg

 

[rapt]

A side loaded beam as you are discussing is treated as a bottom loading as most of the shear comes in at or near the bottom. Hanging shear reinforcement is required for the full load in these cases. And it is required by most design codes. It is definitely required by AS3600, BS8110 and Eurocode 2. ACI discusswes the need for inclusiuon of any load applied below the inclined crack for loads applied near supports. It does not mention it otherwise.

Draw your truss analogy diagram withn the load applied at the bottom and you will see why you need it. Same as you do for structural streel members.

 

[NeilL286]

Where exactly in ACI is this provision mentioned? Nonetheless, according to fellow poster JAE, millions of buildings are built without any hanger reinforcement in beam-girder joints. Maybe millions of buildings are really in danger? Why is ACI not distinct on this requirement?

How do you differentiate between the compression fan forming in inclined cracks and uncracked beam?

 

[rapt]

NeilL286,

R11.1.3.1.

Remember also that a design code is not a design manual or textbook. Engineers are supposed to understand and apply proper design theory. Codes are supposed to define Limits etc, not teach you how to design.

 

[hokie66]

I think this may be a solution in search of a problem. The truss analogy is obvious, but as in many cases of strut-tie theory, there is more than one applicable model. Typical details for stirrup placement show one stirrup in close proximity to the support. In the case in discussion, this stirrup would serve as a tie to "hang" the load nearer to the top of the supporting member.

As JAE pointed out, many structures are standing without purpose designed hanging reinforcement in beam to beam joints, and I don't think there have been reports of in service failures of these joints.

Hanging reinforcement is crucial in some connections, e.g. corbels, halving joints, but in full depth joints like this, I think the required reinforcement is typically there as a matter of course. Needs to be thought about, but usually can be deemed to comply by inspection.

 

[NeilL286]

Hookie, please see images of the joint I designed which gave me great concern as it doesn't have the internal hanger stirrups. Earlier in this thread, I posted the hairline crack below the joint. Here is the file photo of exact bar details of the joint. Have you designed anything like it? How does your design differ from it? Do you believe the top 4 bars of the beam framing into the joint can carry load or only for moment requirements with the load dispensed 45 degrees downward? What are others opinions of this (and how do you design such joint)?

or

http://img849.imageshack.us/img849/9636/8mrd.jpg

 

[hokie66]

Why is there more top reinforcement than bottom in the supported beam? No, I don't think those bars carry the shear. You have two (four legs) #3 stirrups in close proximity to the girder. Those can act as tension ties in a truss analogy, delivering the load at the top, if they are strong enough.

As to the crack, I couldn't see it in the photo, but will have another look.