Anchors + Friction vs Seismic w/ Overstrength

[P1ENG]

I'm looking for some thoughts. I have an elevated tank with granular fill supported on concentrically braced columns (R=3, Omega=2). I have run the load combinations without overstrength and there is no net uplift (actually minimum compressive force is ~230 kip [(0.9-0.2SDS)*D+E]) and worst-case shear is ~45 kip.

I have determined anchorage for shear without overstrength, however there is no way to do ductile yielded plate (unless someone wants to enlighten me), so technically I should be using overstrength for shear. If I do that, then shear loads would be ~90 kip (45 kip * Omega).

Now for your opinions (or a reference to the code that will crush my dreams). Can the 50 kip capacity of the anchors be combined with the 230k*0.2 = 46 kip friction resistance (steel on concrete assumed mu=0.2) to satisfy the overstrength design? I've seen the friction question posed before, but it was always to resist 100% of seismic with no anchorage, so I didn't know what the community thought about using friction as a redundancy or to supplement the overstrength design.

Juston Fluckey, SE, PE, AWS CWI
Engineering Consultant

 

[Rabbit12]

I'd say no. In order to engage the anchor you'd have to slip. I'd be hesitant to sum the two mechanisms.

 

[bones206]

I say no as well. For the same reason that slip-critical connections are designed as bearing type even though they are detailed as slip-critical.

 

[JStephen]

"The net base shear to be resisted by the anchorage Vnet shall be that portion of the calculated base shear that exceeds the calculated frictional resistance. For elevated tanks, anchorage for base shear (sliding) need only be provided when Vnet exceeds zero."- AWWA D100-11, 3.8.7.1, applicable to municipal elevated water tanks. Note that other provisions require you to be able to develop yield strength or 4/3 of yield strength in the cross-bracing, assuming the use of sway rods, which may be a problem if you have rigid (oversized) bracing.

 

[bones206]

If you are trying to develop ductility in the brace, wouldn't you have to ensure your anchor bolts were designed for that overstrength so they don't fail first? Reducing their design capacity by taking credit for friction only makes them more of a weak point in the system.

 

[bones206]

Also, check out 15.7.10.4:

For post-supported tanks and vessels that are cross braced... (d) The connection of the post or leg with the foundation shall be designed to resist both the vertical and lateral resultant from the yield load in the bracing...

 

[P1ENG]

Rabbit12: What is wrong with slip? Just because there is slippage doesn't mean the friction force goes away. Static vs dynamic friction and my 0.2 is pretty conservative for either. Friction plus mechanical anchors still exceeds overstrength.

bones206: Thank you for taking the time to comment a couple times. You've given this some thought so I will have to review your comments with more thought. I've never designed a connection as slip-critical (always snug-tight), so I will have to determine what you meant in your first post. I am not trying to develop ductility in the brace or connection, that is why I am opting for the omega factor. I'll have to look into 7-16's section you referenced. My first thought is if there is no net tension in the post/leg (and far from it), why would I design my anchor to develop the vertical yield load of the brace?

JStephen: I do not have that standard, but it looks like you posted something that validates my approach. Unless I am reading that incorrectly? My bracing is L3.5x3.5x1/2, so it is pretty rigid. However, it is concentrically braced, so the yield strength of the x-bracing would cancel each other out. I assume sway braces are not meant to carry any compression and that is why it is done, but my braces do carry compression.

Juston Fluckey, SE, PE, AWS CWI
Engineering Consultant

 

[bones206]

I guess my thought on the slip critical analogy was that your base plate could slip into bearing on the anchor bolts and then maybe the next jolt in the earthquake is being delivered directly into the anchors. But I see what you are saying about friction theoretically always being present and only resisting the net shear with the anchors. I thought in general that codes discourage taking credit for friction resistance with seismic loads, so I was surprised that AWWA D100 allows it.

As far as the bracing yielding, I believe that is a code requirement independent of the load magnitudes. The R factor for this type of structure is probably based on bracing ductility being the energy dissipation mechanism, so you have to bake that mechanism in your design to justify use of the R factor. It’s more about the relative strengths of each element in the load path and consistency of structural behavior with the underlying assumptions of the R value, rather than designing only for the actual calculated loads. At least this is my understanding of the code intent.

 

[Rabbit12]

Juston, I guess I'd just be concerned that the two mechanism aren't exactly sharing the load similar to what bones206 describes above.

I certainly could be convinced that they can be summed, but my first inclination would be to not do that.