Shear Friction and Composite Construction

[MacGruber22]

We are installing new shear walls between existing columns, and used RAM for our frame analysis and for the shear wall design. The problem was that the doweling into the existing columns and elevated slabs to transfer shear was excessive considering trying to develop bar lengths to meet the shear friction requirements. See attached sketch.

What we ended up doing was utilizing the 80*b*d allowance from the composite construction chapter for the horizontal dowels, and then added lug-style dowels for some "belt and suspenders". Worked nicely to avoid developing the dowels to L sub d. Made sense considering the couple of existing columns act much as flanges, and the shear wall as the web of a beam.

The part that got confusing was the vertical dowels. Now, the ones towards the ends of the new shear wall were obviously working more in tension/comp. so they had to be developed to splice to the wall below/above - no big deal as there is room between the wall perforations and the existing columns to do that. The issue became trying to meet the shear force requirements (output from RAM). We couldn't develop the bars for shear friction at the perforations, and after hours of appendix D design, we couldn't get lugs to work. Frustrating.

There seemed to be two options other than a complete reconfiguration of shear walls (which was not going to happen). 1. Bundle a ton of fully-developed vertical dowels together to handle all of the tension and the shear friction at the left and right ends of the shear walls. 2. Justify a bond strength of the new shear wall to the 6" structural floor diaphragm.

I know the composite construction chapter is using flexural theory of transverse shear (VQ/I) to derive the the 80*b*d allowance of shear at the interface (with a roughened surface). I mentioned that it seems like what that 80*b*d equation is telling me is that their is an 80 psi presumed shear bond strength of the roughened surface. The units of that "80" must be stress, so why not?

I was told that the vertical dowels in this situation are take "vertical" shear, i.e. transverse shear to the flexural element. I agree in definition, but I feel that because this flexural element is oriented orthogonal to a typical composite beam (i.e. gravity is acting parallel through the flexural member) there should be some allowance for concrete shear bond at the shear wall-floor interface.

Actually, now that I am in the middle of typing this - I can see why that may not be OK to assume, considering that the tension bending forces are trying to pry that interface apart, hence reducing the normal friction force. if there is too much tension strain orthogonal to the interface, that is a problem for developing shear. I take it back - it does sound riskier to assume a concrete bond strength there...

haha..thoughts?

 

[TXStructural]

You absolutely have to develop whatever connection you have between the old and new in order to use shear friction. Since you are only doing one of these, do not get too concerned about the depths of the embedded bars or the count (unless we are talking 4" OC). Contractors do this all the time - just look at the highway repair guys drilling dowels for hours a day. They will say that it is excessive, but you did the calcs and know it is correct, right?
Roughen the surfaces before drilling by bush hammer or similar method to expose the coarse aggregate. Hammer drill holes sized for adhesive embedding of rebar. Properly install the rebar dowels with adhesive - you will need to go either one full Ld, or use 318 App D - I prefer using Ld, since this assures that the bars engage properly with bar in the old member (15-18" for #3 or 4 bars, but it may be excessive for large diameter bars.)

 

[MacGruber22]

Thanks, TX.