AASHTO 3.10.9.2 Horizontal Connection Force

[jrw501]

I’m looking at designing anchor rods at a fixed pier of a 3 span continuous bridge (E-F-E-E) to meet AASHTO 3.10.9.2 requirements of the horizontal design connection force of 25% of the vertical reaction due to tributary permanent load and live load assumed to exist during an earthquake.

Because the fixed pier connection has to transmit 25% of the vertical reactions for the whole bridge, it’s seemingly very difficult to meet some of the ACI-related anchor provisions (particularly regarding breakout shear unless the pier is made to be super wide or stirrups/hairpins are provided around the anchor group and if we’re to assume only a few anchors in the group take the full load). The anchors themselves are no problem to design for the load.

Are there some typical assumptions people make to get around these in a pier cap/bent cap? Maybe counting on cap stirrups to restrain failure surfaces in an extreme event for example? Using friction to reduce applied horizontal forces on the anchors? I've looked at two other projects (one with virtually the same span arrangement and pier width, but a narrower bridge with few girders) that were designed for the same DOT, but it doesn't look like any special considerations needed to be made at the fixed bearings. I'm not particularly concerned, but I'd certainly still like to make it work on paper.

Thanks!

 

[bridgebuster]

I've never seen anyone design for "ACI forces". If there was a large uplift force, it could warrant consideration. AASHTO's concern is preventing unseating of the superstructure. If you have sufficient bridgeseat width you're OK.

 

[Guest090822]

If you are considering only shear breakout of the concrete then you’ll never get it to work. The anchor bolts are cast in and the pedestals generally have hoops with vertical bars that go well into any pier cap. Everything would have have to fail ( anchor bolts, hoops, verticals, and any forces to concrete/reinforcement below the pedestal) so you need to consider the reinforcement too. ACI kinda sort of mentions this but then gives absolutely no guidance.

 

[jrw501]

Thanks, the bridge seating meets the requirements and I'd have to check, but I don't think we see any uplift, it's just that 25% provision. AASHTO C6.13.2.12 refers us to ACI for the global design of anchorages to concrete (although I'm not a big fan of mixing and matching codes, especially in the LRFD age with different load and resistance factors having been calibrated for different purposes) -- and I too have never actually seen someone perform those kind of calculations for bearing anchorages on bridge piers and perhaps it's for the reason Rick mentions (the riser bars etc... that would have to fail).

I'm glad I'm not the only one who finds the ACI guidance on accounting for the rebar pretty lacking, but I guess I can use it and maybe some information from Widianto's paper (

https://pdfs.semanticscholar.org/d31d/19ed0c5c63c6efa7417ec60fdcd814e728ad.pdf)

to at least show on paper it's no concern.

 

[BridgeSmith]

We've always used 25% of the vertical load on the bearing assembly under consideration, not the entire bridge. I suppose if there were no anchor bolts at the bearings on the other substructures, you might have to consider 25% of the full weight of the superstructure, but you'd also have to consider horizontal rotation of the superstructure, since your fixed pier presumably isn't at the center of mass of the superstructure. Longitudinal movement is presumably limited at the abutments. If the transverse movement is limited by anchor bolts or some other means of restraint at the other substructures, then the restraints at the other substructures will take their tributary portion of the load.

Rod Smith, P.E., The artist formerly known as HotRod10

 

[jrw501]

For the transverse direction I consider the other bearing assemblies since they are longitudinally guided disc bearings, but for the longitudinal direction the friction isn't enough for them to get 25% of their tributary length to the anchors/substructure. It also seems AASHTO wants me to consider the full bridge length since "For each uninterrupted segment of a superstructure, the tributary permanent load at the line of fixed bearings, used to determine the longitudinal connection design force, shall be the total permanent load of the segment."

I should probably also add that this is a 3 span main unit of a much longer bridge (so there's a number of approaches on either side rather than abutments to limit longitudinal movement).

 

[BridgeSmith]

I should probably also add that this is a 3 span main unit of a much longer bridge

Well, that changes things significantly. In that case, you would have to use 25% of the weight of those 3 spans, since that is the only restraint location for that piece of the the superstructure. Hopefully, you're aware that the substructure has to carry that force to the foundation, as well. That's usually where we run into problems trying to design a superstructure that is not continuous. The seat width and length requirements also apply to all of the substructures in that case.

If it were my design, I would give serious thought to making the superstructure continuous, not only for the improved performance during an earthquake, but also keeping the joints sealed and protecting the girders has been a common problem. As the old joke goes, what do all bridge joints have in common? They all leak. Tennessee has done some very long jointless bridges; I recommend you take a look at how they've accomplished it.

Rod Smith, P.E., The artist formerly known as HotRod10

 

[jrw501]

You'd think so, right? But since this is Seismic Zone 1, AASHTO 5.11.2 (per my understanding) only requires the connection between the superstructure and substructure to be designed for that force, but not for the design of other structural components (outside of some column detailing depending on the acceleration coefficients).

I'll take a look at some of the Tennessee bridges, there's definitely a movement toward eliminating joints wherever possible and we're starting to see things like floating bridge/deck concepts in Louisiana that make use of longer span continuous units with link slabs. It's possible we could've eliminated a few more joints, but this bridge is about 3000' long with a ~900' long main span unit.

 

[BridgeSmith]

You'd think so, right? But since this is Seismic Zone 1, AASHTO 5.11.2 (per my understanding) only requires the connection between the superstructure and substructure to be designed for that force, but not for the design of other structural components (outside of some column detailing depending on the acceleration coefficients).

You are correct. In the 8th Edition they have removed the line that was in the 7th Edition: "The horizontal design connection force shall be addressed from the point of application through the substructure and into the foundation elements." It was also stricken from the Seismic Guide spec. in the 2015 Interim.

Of course, that's not the only strange thing about the seismic provisions. If you were in the lower end of Seismic Zone 2, you could potentially be designing the connection for a smaller force than you are for Seismic Zone 1.

Rod Smith, P.E., The artist formerly known as HotRod10

 

[BridgeSmith]

I suggest carefully reading the commentary (C3.10.9.2), particularly the parts that allow for designing to a higher seismic category to decrease the connection force, and the part about other components that need to be considered to meet the objective of preventing the total loss of support for the span. It seems they're basically saying that where there is a very low probability of seismic activity, whatever damage the bridge sustains, as long as it doesn't collapse or otherwise potentially kill people, is acceptable. This is consistent with "Life Safety" standard of design discussed in greater detail in the Seismic Design spec.

Rod Smith, P.E., The artist formerly known as HotRod10