Section 6.2.3 Robustness - Minimum Lateral Resistance of Connections & Ties

[Drapes]

With reference to section 6.2.3 and 6.2.5 under Structural Robustness in AS1170.0, does the 5% of the value of G and G+ycQ refer to the total vertical load from all slabs above the connection in question, or just the vertical load from a single slab?

 

[blihpandgeorge]

i read it as 5% of the load being transferred, ie the local load from floor not the total load from above

 

[Drapes]

Thanks blihpandgeorge, thats great.

Do others also agree that it can be based on the load being transferred from the floor? Is there any general guidance outside of the code which I can refer to that explains the logic for this?

 

[Drapes]

Any further comments on my previous post folks?

Also, when designing the column top and bottom connection for this minimum load, can we rely on the existing longitudinal column reinforcement that extends and is fully developed through the slab joint to resist this load (via shear friction provisions), or is the expectation to provide additional horizontal tying reinforcement in the slab as well? If additional horizontal tying reo is required (which I can understand for edge and corner columns) can we rely on the existing flexural reinforcement in the slab to act as this horizontal tying reo? How far would this tying reinforcement need to extend into the slab? And over what width would you distribute this reinforcement?

If horizontal tying reinforcement is required (in addition to flexural reo) to withstand this 5% minimum lateral load from the column into the slab, I have my doubts that designers are adopting it or even aware of this clause.

 

[Agent666]

Well in NZ concrete code we need to tie in the column to the floor for 5% of the total axial load on the column or 20% of the seismic shear force in the column. You're effectively applying the restraint conditions like you would in steel design to ensure the effective length for buckling is constrained to a single storey.

Commentary

As to why people don't account for the 1170 requirements explicitly, there only seems to be a couple of practical situations where it might become critical that I can think of (otherwise typical detailing just covers stuff without thinking specifically about it). One common one being edge columns without a tying beam heading off internally to the floor slab, and as I said in NZ we have a more onerous and in my mind more correct consideration of the total load and restraint of columns than the 1170 requirement of implying only considering some proportion of the vertical load transferred from a beam/slab to a column.

Though you could interpret the 1170 requirements as considering the vertical column load and 5% of this requiring tying into the floor. I believe given in NZ we have this requirement, that that's the actual intent. You're trying to stop a column peeling off the floor, so the incoming vertical slab or beam loads at each level are somewhat irrelevant, only the column load is relevant isn't it.

https://engineervsheep.com

 

[Drapes]

Thanks for providing such a comprehensive summary Agent666, just what I was looking for.

Taking 5% of total axial compressive load seems intuitive to me.

A few follow up queries:

1. Would it be reasonable to extend the tie reinforcement a lap length inboard from the slab edge, in order for the reinforcement to be of sufficient length to allow effective dissipation of the design tensile force within the diaphragm? Or would you extend to the next column grid?
2. Do you confine this tie reinforcement over the width of the edge column? Or is there some flexibility to locate beyond the column face?
3. I suspect the horizontal tie reinforcement will need to be in addition to any existing flexural reinforcement?

 

[Agent666]

1. Depends on the load, they are asking you to transfer the force. So you need to engage sufficient diaphragm reinforcement. Review using strut and tie unless your diaphragm reinforcement is higher quantity than your tie. I suspect in most practical arrangements a development length won't cut it.

2. I've always considered the reinforcement needs to go into the column. We wouldn't really consider it a good detail to tie into the potential hinge region of the edge beam. But I guess some judgement can be used provided you can demonstrate a strut and tie mechanism to engage some proportion of the reinforcement being placed just outside the column.

3. I've always considered it as additional to any other requirements. But I'm not 100% sure. Obviously if you have excess flexural or diaphragm reinforcement you can utilise anything over the requirement.

https://engineervsheep.com

 

[Drapes]

Will locate the tie reinforcement over the width of the column and extend it far enough into the slab to mobilise enough diaphragm reinforcement. To determine the amount of diaphragm reinforcement engaged, thinking of a strut and tie model where say a 45 degree strut angle fans out in two directions starting from the termination point of the tie reinforcement back towards the slab edge. The number of bars contained within this fan strut zone determines the diaphragm reinforcement that has been engaged. Does this sound reasonable?

Also, just playing devil's advocate here, but surely there would be instances where the 5% load is low enough that you could justify the load path without having to even rely on any supplementary tie reinforcement. For example, you would simply rely on the longitudinal column reinforcement passing through the joint to transfer the load directly into the slab (via shear friction), not dissimilar to anchor theory where the shear is transferred directly into an unreinforced slab. Of course I appreciate you would possibly have edge distances, cone breakout failure modes etc to contend with here. What are your thoughts on this?

 

[rapt]

My thoughts on this are the same as Agent666, but I am being told by others more involved in AS1170 development that it is specifically referring to the specific floor load only transferring to the column and not to lateral bracing of the column/wall. There is general agreement that the wording needs work if this is the case!

Apparently, according to them, lateral bracing is meant to be handled by the Materials code even though the wording in 6.1 of 1170.0 would suggest it covers everything. AS4100 has 2.5% for steel structures. Not sure if it has anything specific for earthquake.

AS3600 is silent!

I have passed on a request for more information from those who should know more than I obviously do not.

 

[rapt]

Slightly wrong above, AS3600 clause 11.3 has 2.5% lateral restraint requirement for walls. Presumably it should also apply to columns.

 

[Settingsun]

I read AS 1170.0 clause 6.2.3 the way the code developers are explaining it to Rapt. It's a lateral load of 5% of the (G + Q_combination) vertical load that the connection transfers. I would personally clarify in the clause itself that this is referring to a lateral load rather than relying on the heading, but that's just style preference.

Bracing of the column would be based on ultimate load, not the 'instant in time' estimate of an average load.

 

[rapt]

Section 5.4 of AS1170.4 also gives minimum connection forces for seismic conditions, again for floor load connection, not lateral bracing of the column.

 

[Drapes]

As I have understood then, in summary -

The minimum lateral connection at the interface b/w the wall and slab (considering floor load only) is given by:
5% of G+Q_combination for robustness
3% of G+0.3Q for seismic

And minimum lateral connection to justify lateral bracing or restraint at the interface b/w the wall and slab (considering total axial load on the wall or column from all the floors above) is given by:
2.5% of 1.2G+1.5Q

In light of this, it would appear the minimum lateral connection load used to satisfy bracing or lateral restraint would govern in all cases except if you have only one level of slab in your building, in which case it would be dictated by robustness. Unsure if you would ever have a situation when the seismic case would govern.

Also any thoughts on the 2x queries I posted earlier?

To determine the amount of diaphragm reinforcement engaged, thinking of a strut and tie model where say a 45 degree strut angle fans out in two directions starting from the termination point of the tie reinforcement back towards the slab edge. The number of bars contained within this fan strut zone determines the diaphragm reinforcement that has been engaged. Does this sound reasonable?

Also, just playing devil's advocate here, but surely there would be instances where the 5% load is low enough that you could justify the load path without having to even rely on any supplementary tie reinforcement. For example, you would simply rely on the longitudinal column reinforcement passing through the joint to transfer the load directly into the slab (via shear friction), not dissimilar to anchor theory where the shear is transferred directly into an unreinforced slab. Of course I appreciate you would possibly have edge distances, cone breakout failure modes etc to contend with here. What are your thoughts on this?

 

[Agent666]

To determine the amount of diaphragm reinforcement engaged, thinking of a strut and tie model where say a 45 degree strut angle fans out in two directions starting from the termination point of the tie reinforcement back towards the slab edge. The number of bars contained within this fan strut zone determines the diaphragm reinforcement that has been engaged. Does this sound reasonable?

Yeah more or less, but you will need to consider development past the nodal zone typically in strut and tie, in practice this just means fan out from the point where the tie reinforcement is fully developed (not the end of the bar).

https://engineervsheep.com

 

[Drapes]

Thanks Agent666, appreciate it.

Do you have any thoughts on my second query re: justifying the load path without any supplementary tie reinforcement?