Redistribution of negative moments

[db7]

Instead of redistributing a maximum of 20% of the negative moment at the ends of a concrete beam, can I just design a beam as simple support and assume no end moment transfer to the column? Any justification in the ACI code?

 

[CANPRO]

That might be a conservative design for the beam, but not the column. I think this was talked about recently here about the effects of ingoring the fixed ends of the beam.

How is the beam connected to the column? If its monolithic you will have moment in your column and it will be undersdesigned.

 

[msquared48]

If there is no negative moment steel there to develop the moment, which is unlikely, then yes. Otherwise you are just kidding yourself.

Mike McCann
MMC Engineering

http://mmcengineering.tripod.com

 

[IsaacStructural]

msquared, if you leave out the negative moment steel, won't you get crazy uncontrolled cracking? If the beam is cast monolithically, the moment will still develop, just the concrete won't be able to take tension from the flexure.

M.S. Structural Engineering
Licensed Structural Engineer and Licensed Professional Engineer (Illinois)

 

[db7]

The beam is cast against an existing concrete column. Shear is to be transferred by shear friction dowels. We are concerned that the presence of the dowels will restrain rotation and draw some end moment.

 

[PicoStruc]

Welcome in the eternal debate between what people does in practice (plastic design - 100% redistribution) and what code prescribes (elastic design with 20% limited redistribution).

A common practice is to design beam as simply supported and still put a certain amount of negative reinforcement to prevent cracking at service level.

With this 'rational' method - that is in reality a plastic design, no moment connection detailing with stirrup in the joint is required (shear transfer).

There is a lot of debate between engineers about if it is a good or bad practice. Any opinion ?

In my opinion it is good because, Imagine a transfer beam where the column cannot take the moment or the negative reinforcement is badly developed (anchored), the joint capacity is reduced and maximum negative moment cannot be achieved in the beam. Thus, the beam moment capacity can be dangerously reduced at mid span after redistribution of moment if there is not enough bottom reinforcement (designed with lower moment due to negative moment at support).

 

[msquared48]

IsaacStructural:

Yes, you will. I was not suggesting to remove it, only saying that the moment will be there if the steel is there. Kinda like the "if you build it, they will come analogy".

Mike McCann
MMC Engineering

http://mmcengineering.tripod.com

 

[CANPRO]

Well no matter what the connection looks like, you're going to have an eccentricity equal to half the depth of the column...so you do have some additional moment in the column.

Can you detail the dowels to limit moment transfer? Maybe sleeve the dowels in the top half of the beam to take away the tension in the dowels?

 

[hokie66]

CANEIT,
That approach might work if this were a bearing connection, but the OP said he is using shear friction to connect to the face of an existing column, and the theory of shear friction is that the load is carried by friction, which requires tension development of reinforcement to provide the clamping force.

I don't believe in shear friction for that type connection, and would suggest to db7 that he needs to at least key into the column to provide a beam seat.

 

[slickdeals]

There is some good information in the commentary of ACI 318 Section 11.6.7 that talks about moment and shear friction.
Depending on the loads you are dealing with, I agree with hokie66 that providing a "positive" connection by means of beam seat would be better.

It is a question of chicken or the egg. You are trying to create a simple shear connection, but doweling the bars provides fixity.

Can you provide some more information? Is this a "flying beam" or will it have a slab on top. If so, how is the new slab attached to the existing slab. You can add top bars in the flange of the beam, cut grooves in the existing slab and develop that reinforcing better than epoxy into the column.

 

[db7]

The beam will be placed on top of an existing slab (no access underneath. The slab will be doweled to the beam. So the top reinforcement at the column is whate
ver gets dowelled in.

 

[db7]

I am wondering why some hesitate to use shear friction? Any bad experiences or specific concerns?

 

[TXStructural]

Shear friction can only happen with fully developed bars or tendons. It is not the shear on the dowels, it is the friction across a cracked plane restrained by the reinforcement across the (potential) crack. Using the dowels as the shear transfer is just that, like a bolt. Design the beam as singly-reinforced, and make serviceability a primary consideration (that is, make it as stiff as you think you need.) Have bars at each corner and stirrups/ties along the length. It will act somewhat like a precast element would. You may be able to use the slab for the shear transfer at columns, or you may want/need to create a key into the cover of the existing column. Since it will not be integrally cast and will not have top edge lateral support, avoid narrow sections.

I have done this, and we "hung" the slab below using threaded rods and plate washers to connect the two. If you truly do not have access below, be cautious about your connections to the slab. Be sure mechanical anchors are pretensioned, or the adhesives are properly selected, mixed, and placed.

 

[db7]

Thanks, TXStructural. Our understanding of shear friction is similar to what you've said. My concern still remains that unless the ends of the beams are allowed to rotate, that there will be some negative moment transfer at the beam ends, for two reasons: 1) the shear friction dowels and 2) the trapped condition of the beam between the column. I understand the with the beam designed for the full simple span moment that it will not fail, my concern is pulling the dowels out of the concrete column, similar to the failure of an adhesive anchor with concrete pull out. We are finding it impossible to have the dowel bars yield prior to pull out failure unless very small bars are used, and these aren't enough for the shear friction. We are using the HILTI profis to design the dowel embedments.

 

[hokie66]

It would be helpful to know the sizes of beam, column, and existing slab we are talking about. Shear friction theory requires development of the bars, so you can't avoid that requirement if you are going to depend on friction.