Cantilever Pipe Support - Effective Wall strip

[EDub24]

Hi, I'm working on a project under construction rehabbing an existing externally wrapped post-tensioned concrete tank that was built in the '60's. The Contractor has submitted a pipe support design that's basically a cantilever steel beam support off the existing 10" thick concrete core wall (pipe is on the interior of the tank). The wall has minimal reinforcement (likely shrinkage only) and does not have much moment capacity at all. We did testing of the structure during the design phase and found the concrete strength to be in excess of 6000 psi and the steel yield strength to be 44 ksi. Is there a way to determine the effective width of the wall that would resist the moment from the beam support? So if the support has a moment of 20 kip-ft and we use a 2' strip that comes down to 10 kip-ft/ft of wall.

My preference is to have the Contractor provide some type of kicker support so that there is no moment transfer from the support into the wall but I want a backup plan in case I get push back from the Contractor.

 

[WARose]

Depends on the connection. I assume we are talking something like a bolt group (i.e. Hiltis) connecting the cantilever to the wall. In such a case, it's going to depend on the bolt spacing to determine how much of the wall is engaged to resist the moment.

 

[EDub24]

Depends on the connection. I assume we are talking something like a bolt group (i.e. Hiltis) connecting the cantilever to the wall. In such a case, it's going to depend on the bolt spacing to determine how much of the wall is engaged to resist the moment.

Correct. They have two pipe supports both with a standard 4-bolt pattern. One pipe support has the bolt spacing at 7" and the other support has them at 12". The support with the 7" bolt spacing I'm not too concerned about because the forces are small however the larger support (naturally) has much larger forces.

My thinking is to use the 35-degree cone that ACI 318 Appendix D uses to calculate the tensile strength of an anchor (diameter = 3*embedment depth) and figure out the effective width from that. So basically 2*1.5*embedment depth + bolt spacing = effective width.

 

[WARose]

My thinking is to use the 35-degree cone that ACI 318 Appendix D uses to calculate the tensile strength of an anchor (diameter = 3*embedment depth) and figure out the effective width from that. So basically 2*1.5*embedment depth + bolt spacing = effective width.

Makes sense to me.

 

[dhengr]

EDub24:
With ACI 318 Appendix D you check to determine if you have a cone type failure, or the strength of that A.B. prior to that failure. If you don’t have a cone failure, then your effective width can be something greater, which good sound engineering judgement, based on the rebar and conc. strength would allow. Maybe something like the cone dia. plus 1.5 – 2 times the wall thickness. The contractor came to you with ‘A’ design, nobody said it was a good design or a practical one for this situation. You, as the engineer, get to select the design which best fits the situation, and if there are any questions, your get to explain to him why yours is the preferred solution. There is no need for some great debate if he wants you to sign off on it. Your top beam and kicker should be a fine solution, with fewer moment issues, and not appreciably different in price to a beam with a moment end pl. But, I’d suggest a three-piece fab. for your solution: a top beam, the kicker and a vert. bar or member to the tank wall, which connects the beam and the kicker, at the wall. Now, you have a top tension, a bot. compression and a bar which distributes the shear to all the A.B’s. and both main members.

 

[EDub24]

With ACI 318 Appendix D you check to determine if you have a cone type failure, or the strength of that A.B. prior to that failure. If you don’t have a cone failure, then your effective width can be something greater, which good sound engineering judgement, based on the rebar and conc. strength would allow. Maybe something like the cone dia. plus 1.5 – 2 times the wall thickness. The contractor came to you with ‘A’ design, nobody said it was a good design or a practical one for this situation. You, as the engineer, get to select the design which best fits the situation, and if there are any questions, your get to explain to him why yours is the preferred solution. There is no need for some great debate if he wants you to sign off on it. Your top beam and kicker should be a fine solution, with fewer moment issues, and not appreciably different in price to a beam with a moment end pl. But, I’d suggest a three-piece fab. for your solution: a top beam, the kicker and a vert. bar or member to the tank wall, which connects the beam and the kicker, at the wall. Now, you have a top tension, a bot. compression and a bar which distributes the shear to all the A.B’s. and both main members.

I agree. I tend to be a bit more conservative with existing structures but yes I've gone out farther for new structures without cone failure. The three-piece solution would definitely work and that's something i've seen/done before with strut supports. In this case I don't want to direct the Contractor as to what he should do since this is his design/stamp and instead I've given them parameters to work with. I've been in situations where the Contractor will demand that I design/stamp something if I give them explicit directions so instead I gave them a maximum moment allowed on the wall which effectively will force them to go with a braced solution.

 

[JStephen]

There are design methods for analyzing local loads on a cylindrical shell- see WRC 107, WRC 297, etc. The methods tend to be cumbersome, and generally assume "thin" non-composite shells (IE, steel plates, from the pressure vessel industry). Some of the pressure vessel handbooks have simplified methods also. If it's just one support, it'd probably be quicker to add some metal than it was to prove you didn't need it.

 

[BridgeSmith]

My concern would be a local failure at the anchorage, unless the vertical reinforcement in the wall is developed with hooks, etc. The anchors would have to go deep enough to "lap" with the vertical reinforcement, wouldn't they?

Would it be acceptable to run a tension strut down the outside of the wall and anchor into the wall lower down? This would give you a composite system and continuity with fully developed reinforcement.