Tank Anchorage, Overstrength Factor, and Overturning 2

[AThor]

My question relates to ASCE 7 Chapter 15, particularly 15.7.5. I am designing the anchorage for a saddle supported liquid storage tank. Section 15.7.5 makes it clear that the overstrength factor is NOT to be used for tank anchorage calculations, because they are trying to force a ductile yield in the anchor, requiring the concrete to develop the full steel strength of the anchors in tension. However, do I apply the overstrength factor in my overturning calculations? When I do not use overstrength, there is no net overturning in the tank, i.e. no anchor tension. But, if I do use overstrength, there is a net overturning moment, so I get anchor tension. How would you approach this situation:

a) Do not apply overstrength to overturning calculations, no anchor tension, so shear governs anchor design. But, still apply the detailing requirements of 15.7.5 to design the embedment to develop the full strength of the anchor in tension, and provide 8*diameter of gauge length?

b) Do not apply overstrength to overturning calculation, no anchor tension, so shear governs anchor design, and do not apply the tension detailing requirements of 15.7.5.

c) Apply overstrength to overturning calculations, determining that there is tension in the anchor. Then, anchors would be designed for the governing tension/shear combination, and the detail requirements of 15.7.5.

On a side note, to accomplish the 8 diameter gauge length requirement, we would have to weld on a plate and pipe sleeve to give the anchor standoff above the concrete. The sub who is ordering the tank told me he has seen this detail before, but usually on vertical tanks, not horizontal like this one. So, maybe others have not used that requirement on this type of tank in the past. But, since all installations are site specific, and this is in a SDC D region, I want to determine for myself if this is actually required.

Thanks for any responses.

 

[JStephen]

Not sure on that- but on vertical tanks, you evaluate overturning/uplift without the overstrength, then if applicable, it is applied to the calculated net bolt load (and limited to the yield strength of the bolt).

 

[oldrunner]

Just curious, but are you also considering thermal movement between the saddles?

 

[bones206]

Since you are in SDC D, you must detail the anchors per 15.7.5. That requirement is based on SDC and independent of load magnitude or whether or not you anticipate overturning.

As stated in 15.7.5, do not apply the overstrength factor for the purposes of sizing the anchors. The intent is that the bolts will yield and dissipate energy, while the tank components remain essentially elastic. If anything, I would anticipate the saddle would be designed with an overstrength factor. Although I've never designed a saddle so I'm not sure...

There are also other ways to achieve the required stretch/gauge length, like putting tape around the anchor or using anchor sleeves cast into the foundation.

 

[bones206]

The section governing the anchor-to-saddle connection (or any attachment element between the anchor and vessel) is 15.7.3:

15.7.3 Strength and Ductility.

Structural members that are part of the seismic force-resisting system shall be designed to provide the following:

a. Connections to seismic force-resisting elements,excluding anchors (bolts or rods) embedded in concrete, shall be designed to develop Ω0 times the calculated connection design force. For anchors (bolts or rods) embedded in concrete, the design of the anchor embedment shall meet the requirements of Section 15.7.5. Additionally, the connection of the anchors to the tank or vessel shall be designed to develop the lesser of the strength of the anchor in tension as determined by the reference document or Ω[sub]0[/sub] times the calculated anchor design force. The overstrength requirements of Section 12.4.3 and the Ω[sub]0[/sub] values tabulated in Table 15.4-2 do not apply to the design of walls, including interior walls, of tanks or vessels.​

So I think that confirms you would apply the overstrength factor to design the saddle itself.

 

[AThor]

Thanks all. Oldrunner, the tank itself has slotted holes on one saddle base, which I assume will mitigate any thermal expansion loads. So, I have not considered those in my analysis.

bones206, I tend to agree with you, which is why I made it option a) in my original post. But, do you have a reference or clause somewhere that states that explicitly, that the detailing rules apply independent of analysis? My only doubt comes from ACI 318-14 17.2.3.4.2. While ASCE 15.7.5 doesn't say anything on the matter, it seems to be essentially mimicking ACI 17.2.3.4.3(a), however, in ACI, this only applies when the tensile component of the earthquake force exceeds 20% of the total tensile force. So, in ACI, zero tensile force means these details don't apply. In a perfect world, ASCE would tell me whether or not that is the case for them as well.

 

[Hokie93]

I vote for Option 'C'. I would run two sets of load combinations; one set without the over-strength factor for member and connection analysis and design and a second set of load combinations including the over-strength factor specifically for anchorage-to-concrete design. In response to your last question to bones, it would seem that ASCE 7-10 Section 15.7.5 makes it very clear the detailing requirements for SDC C, D, E, and F are independent of the analysis results.

 

[bones206]

Section 15.7.5 says:

The following special detailing requirements shall apply to steel tank and vessel anchor bolts in Seismic Design Category C, D, E, and F. Anchorage shall be in accordance with Section 15.4.9, whereby the anchor embedment into the concrete shall be designed to develop the steel strength of the anchor in tension. The steel strength of the anchor in tension shall be determined in accordance with ACI 318, Eq. 17.4.1.2. The anchor shall have a minimum gauge length of eight diameters.Post-installed anchors are permitted to be used in accordance with Section 15.4.9.3 provided that the anchor embedment into
the concrete is designed to develop the steel strength of the anchor in tension. In either case,the load combinations including overstrength of Section 12.4.3 are not to be used to size the anchor bolts for tanks and horizontal and vertical vessels.

My reading of that is that the detailing requirements are mandatory in SDC C, D, E, & F. It would be more clear if they ended that first sentence with a colon, then had a numbered list of the exact detailing requirements. Instead of just packing it all into a paragraph where you don't know where the "following detailing requirements" starts or ends.

But the overall message is to ensure the anchors are the ductile yielding element and the other elements of the anchorage load path remain elastic. To guard against the possibility that the actual anchor material is stronger than specified (which is quite probable), the other elements are designed with the overstrength factor to ensure the anchor still reaches yield before they fail. The commentary for 15.7.5 simply says to design the anchor per ACI 318 Equation 17.4.1.2, which is the steel strength of the anchor alone. Personally, I would follow through with ACI 17.2.3.4.3(a) to ensure the breakout strength is greater than 1.2 times the anchor steel strength. Otherwise, the overstrength concept gets left out of the breakout capacity. You can justify it in way, by saying that the tensile earthquake force is artificially set to 100% of the anchor steel capacity, therefore it automatically triggers the 20% tensile force threshold.

The code jumps all over the place and I find it overly complex, so I try to boil it down to the fundamental concepts and go from there.

 

[AThor]

Those arguments certainly makes sense. On the other side of the argument, there is the 15.7.5 commentary, which says "Where tanks must be anchored because of overturning potential, proper anchorage design provides both a shell attachment and an embedment detail that will allow the bolt to yield without tearing the shell or pulling the bolt out of the foundation." So, what does "overturning potential" mean? Is it that analysis shows overturning at the design earthquake level? Or is it that overturning is theoretically possible under any amount of shaking?

What bones206 says about the intent of the code is a good way to think about it, because the intent is definitely to force a ductile yielding anchor before failure of the shell or foundation.

 

[bones206]

There's a lot of uncertainty with seismic loads, so it makes sense to provide detailing that ensures the energy dissipation mechanism that you want to happen in case of an extreme event. The code arbitrarily reduces the maximum credible earthquake by 2/3 for design (S[sub]DS[/sub] = 2/3S[sub]MS[/sub])... I bet your vessel would overturn in the MCE instead of the design earthquake, so that's a good argument for why you want the right detailing in place.

If you are in SDC A or B, the detailing requirement doesn't apply. So in those cases you can just design the anchors for the loads, or in some situations you can design the tank to be self-anchored. Section 15.7.5 kicks you to the applicable reference document to determine whether or not your specific tank or vessel qualifies to be self-anchored.

Check out Table 15.7-2 and the paragraph below talking about the "anchorage ratio" J. This is one quantitative way to determine overturning potential, but I think it's geared towards flat-bottom, ground-supported tanks. I think it was meant to be in a different section too, because it doesn't make sense to be in Section 15.7.4, "Flexibility of Piping Attachments"...

 

[oldrunner]

Slotted holes have to be designed - not asummed. You have to determine your total temperature range and also include the effect of the sun on the tank and then calculate the daily movement. Have heard of bolts in slotted holes being sheared off because of longer thermal lengthening of the vessel than assumed. I am also assuming that only one pedestal resists the longitudinal seismic forces and both pedestals resist the transverse seismic forces.

 

[AThor]

I am not designing the tank or saddles, just the anchors for a pre-manufactured tank. So, when I say I assume the slots are okay for the thermal expansion, I mean that I assume the tank manufacturer designed it correctly. Although to your point, it might not hurt for me to do an expansion calculation just to confirm that my anchors will not see thermal expansion loading. Also, that is a good point to consider about the pedestal with slotted holes not resisting longitudinal seismic forces.