Splicing of bundled bars

[bugbus]

General question regarding proper splicing detail of bundled bars in piles and columns:

AS3600 and AS5100.5 (and presumably all other relevant codes) require the individual bar splices in a bundle to be staggered by no less than the lap length. In our design drawings, we typically show this as a specific detail, usually at most for 2 bars in a bundle (I have not had to design bundles of 3 or 4 bars, which can usually be designed around and are in any case not allowed for seismic reasons in most cases by AS5100.5).

However, for a 2-bar bundle lap splice, on two occasions (despite our drawings) I have seen a contractor simply provide a double-length lap splice for both bars in the bundle and terminate both bars at the same location (presumably to simplify the reinforcement).

Is there any harm in simply providing this double-length lap splice for both bars in the bundle and cutting off both bars at the same location? I am in two minds about this. On one hand, this detail basically takes the preferred staggered lap splice arrangement and 'adds a bit more length' to two of the bars. It could be argued that this detail still satisfies the requirement that the lap splices are staggered. I am generally of the opinion that adding additional reinforcement will not reduce the strength of the member.

On the other hand, the staggered lap splice arrangement would basically 'force' the two splices to act at two separate locations and reduce the possibility for requiring too much (what I would call) 'bond demand' at a single location, which might encourage splitting/cracking.

Any thoughts?

 

[bugbus]

I should mention that this relates mostly to circular piles and columns with straight lap splices, where a cranked bar wouldn't be required to anchor the corner of a fitment.

For a rectangular column, I would imagine that having two cranks at a single section would require some more careful detailing of the fitments at that location to resolve the outward force component in the bars at the bends.

 

[r13]

It may not be that critical, but I see two potential problems. 1) Bond. 2) The staggered detail has less stress concentrated in one specific location (weak plane), but the even length splice might introduce weal plane within the splice zone.

 

[Agent666]

The way I see it is if you effectively have four bars in your bundle in the middle by double length splices. Then it is not really the same as having staggered splices, you are developing two bars in contact at the same location with no stagger.

Also given for a 4 bar bundle you are also required to increase the splice length by 33% if shows the intent, more bars = less bond = more development length. To avoid weaknesses at individual splices you must stagger them according to the code.

The bond on a bar without splices or not within a development length region is comparatively low when compared to bond stresses in laps/splices/development lengths. Consequently, at laps the stresses are comparatively higher, given the lack of concrete around the entire bar in a bundled lap, the limit is only developing one bar at a time. Double length splices does not achieve this, anecdotally you'll get higher bond stresses developing effectively with two bars in contact at the same location.

While we have this notion of development length, providing twice as much length does not halve the bond stresses. I think that is the crux of the issue here.

The code (AS3600) couldn't be clearer in terms of the intent:-

"Individual bar splices within a bundle shall not overlap".​

NZS3101 contains the same statement. ACI318 contains the same statement. Eurocode 2 has a similar/more onerous requirement (requires 1.3 x development length stagger). Notice the theme here...

In saying that I've seen it to in others designs I've reviewed and always made them change it for the reasons noted. The two are not equivalent in my opinion.

In my own designs I always try real hard to not bundle, it is just a royal pain on site! If you are blessed with the worlds best contractors then go for it, otherwise you might get what you're noticing.