Concrete frame question

[jayrod12]

Good Day All,

My firm has in the past done a few projects that involved a single parking garage below a wood framed structure. All of the previous ones were based off of a design completed 20 years ago by an engineer that no longer is employed at the firm and probably doesn't remember his calculations.

Either way, we have concrete beams that run below the corridor walls between concrete frames. The concrete frames are to transfer the corridor load from their locations (3' off the centre of the building) to the edge of the drive aisle (approximately 11 feet off the centre of the building). I've been tasked with re-analyzing the concrete frame to determine if we can reduce the reinforcing as the spec that has been used for years seems quite ridiculous.

My question is about the connection between the top concrete transom beam (top member of the frame) and the columns. In digging through the old files the only reference I can find is that the connection can transfer approximately 765 kft of moment. I've determined that they calculated this by determining the moment couple of the reinforcing that is continuous around this corner (two layers of 5-35m bars spaced approximately 18" apart).

Is this the correct way of designing this connection or should we be taking into account the concrete (or some other aspect) in our design?

I'll try and submit a sketch tomorrow when I'm back at my computer but hopefully I've described the problem well enough to get the discussion going.

Thanks,

 

[JAE]

jayrod12,
I'm not sure I understand your question. Transfer of load from a beam to a column involves both moment and shear load transfers.
These are basic concrete design issues. What specifically don't you understand about it?
I suppose your sketch tomorrow may help.

 

[jayrod12]

Here is a sketch of the connection, I'm no artist so hopefully it gives you an idea of the corner I'm looking at.

Link

Now my question is, For the moment transfer capability around the corner of the frame is it conservative, or not conservative, or completely incorrect to use the moment couple between the two sets of 5-35m corner reinforcing, centre to centre of bar is roughly 18".

As I said previously the only documentation I have found in the previous projects is a scribble on a paper about the corner can transfer roughly 550 kft (I erred earlier when I said 765, its roughly 765kNm). I played with some numbers and came up with my previous point about the moment couple between the reinforcing.

As = 5000 mm^2
As'= 5000 mm^2
d-d'=18"=450mm
fy=400MPa
fieS=0.85

Mr=0.85*400*5000*450=765kNm=565 kft.

My brain tells me that there is more that should go into the design of this connection but in any of my concrete classes and my experience so far I haven't had to specifically design moment connections. All connections were designed as pins and as such only designed for shear and tension development length.

Part of me thinks there should be some contribution from the concrete but another part of my brain is concerned about the concentrated stress in the concrete right at the inside corner and the high tensile strength at the outside corner.

I think that's enough typing for now.

 

[jayrod12]

tensile strength should read tensile stress.. we need an edit function

 

[JAE]

jayrod12,

The way I understand you is that this is a moment "bent" - i.e. a column, beam, column arrangement. If so, the loading on these should be applied in your analysis as a total frame which would take into account the relative stiffnesses of the beam vs. the columns. The beam typically is assumed at some reduced moment of inertia (in the US - the ACI code uses Ie - an effective moment of inertia based on how much moment it receives). The column, under axial load, many times doesn't approach cracking moment and is used at a higher effective moment of inertia.

ACI suggests for analysis using 0.7Ig for the columsn and 0.35Ig for the beams.

This will have a big effect on how much moment is actually occurring at the end of the beam (and top of the column).

Once you have this analysis completed, you would have shear, moment and possibly axial load on your beam end. You would also have axial, shear and moment at the top of the column - same numbers assuming there are no other members framing into the intersection.

The design of concrete sections in flexure is comprised of a compression block (whitney stress block) and a tension force at the rebar forming a couple. You typically do not use a couple made up of the opposite bars only. If there are bars on the compression side, you can use them in the calculations but only along with the compression block of the concrete as well.

For the column, with the axial load, there is a bit more complexity as well because you have the bars all in compression or perhaps some in tension. It all depends on the relative depth of the column vs. the moment and axial force.

It sounds like you don't know how to do these reinforced concrete calculations. It is not - as you indicate was done before - simply taking the bars and calculating a force couple off the rebar.

Also - the inside vertical column bars extend up out of the column and into the beam - then turn immediately horizontal along the beam bottom face. This is incorrect.

The bars should all be taken up to the top of the beam and hooked there. The beam bottom bars should extend all the way to the far end of the beam and be hooked there.

 

[jayrod12]

Thanks for the input, that was my worry exactly. You are correct I do not know the intricacies that go into concrete design in moment connections. Running beams and columns given standard type loading, i.e. axial and shear and lateral bending, I am confident in my abilities. This concrete frame is a new challenge for me to design correctly and efficiently.

The thing that triggered my looking into this deeper was when I analyzed the frame to get the moment, shear and axial load on the column assuming full moment transfer the value I got absolutely killed the column in bending (like 200% overstressed) so when I inquired with the higher ups they said that the connection isn't going to provide full moment transfer and to look in the older projects.

I've hunted and can't find and definitive calcs so my next step was to ask the guys here how the majority design the capacities at the corners of their concrete frames.

 

[JAE]

Well - ultimately you really aren't designing the corner. Rather - you design the beam just off the face of the column and you design the column just below the beam.

Then make sure the required reinforcement is developed into the corner.

 

[jayrod12]

So if I'm understanding what you are saying, The column will take as much moment as it is capable (assuming the beam is capable of transferring that magnitude of moment to the column) and the reinforcing that continues through the corner needs to be capable of transferring those same loads.

There's no concern about extremely high local stresses at the inside corner?

Do the bars from the inside face of the column contribute anything to the connection when extended to the top of the beam or do I ignore them completely and design the corner as a singly reinforced concrete beam (Essentially the same moment resistance as the column assuming the reinforcing from the outside face is continued into the beam enough to develop full tensile resistance?

 

[WARose]

You may want to look at ACI 352. It covers the checks required for beam-column joints.

 

[jayrod12]

I think I also need to confirm that this frame is not part of the lateral load resisting system of the building nor is it designed for any significant lateral load. It is designed as a frame (instead of just a beam supported on a column) in an attempt to reduce the total positive moment on the beam by transferring some of the moment into the column.

 

[jayrod12]

Another clarification. If people could direct me to CSA clauses as the design is in Canada. I can use the ACI as guidelines but if I need to cite where I got my design information then I would need CSA references.

 

[WARose]

I'm not familiar with anything but ACI code (for concrete).....never worked with Canadian code.

 

[jayrod12]

It has been recommended to me to analyze this closing corner using a strut and tie model. I've got some information from MacGregor "Reinforced concrete Mechanics and design 4th ed". And the code references for Canadian design would be CSA A23.3 clauses 11.4 (Strut and Tie model) and clauses 7.7, 12.5.3 and 12.5.4

What is the general consensus on the Strut-and-Tie modelling technique. I know it is highly subjective and so was curious to everyone's opinions.

 

[CELinOttawa]

I am a very big fan of strut and tie for deep beams or very short spans. When you are designing something that looks like a traditional beam profile, I do not apply Strut and Tie.

It competed with our standard techniques at the dawn of RC engineering. If it lost the battle, it was because of something more than calculators...