seismic anchor bolt design

[mehr27]

Given the SCDOT Seismic Specification loading requirement of designing the connection between the superstructure and the substructure for a lateral load equal to 20% of the dead load reaction in the restrained directions, do you apply the lateral load at the interface between the superstructure and the substructure?

If the lateral load is the result of the inertia of the superstructure when the substructure displaces under seismic loading, I see room to argue that the lateral load should be applied at the center of gravity of the superstructure instead of at the interface. The vertical eccentricity of this lateral load would affect the anchor bolt design loading by creating a couple between the anchor bolts on either side of the beam base plate.

Thanks.

 

[yack]

The 20% rule sounds like a rule of thumb. The old code (here) used to be 1/3 of the dead load and is still part of the shear key design code. The flaw in your reasoning is that your starting with a "fake" static analysis--then your trying to pretend that it is a real loading and say that the inertia of the superstructure will cause a couple.

My guess is that the rule-of-thumb you are talking about is to protect against shear failure at the interface (sudden and possibly catastrophic), bending failure on the other hand is more ductile and is likely addressed elsewhere in your code, as should many other required analyses and less simplified loading conditions.

Overall you don't really sound like you know what your doing, better get some help.

 

[mehr27]

Basically, the 20% is a made-up static load. I was trying to figure out what the behavior was that would cause that load requirement, and I came up with the inertia of the superstructure resisting the movement of the substructure. It's probably fine to apply the lateral load without any vertical eccentricity.

 

[Qshake]

The 20-25% rule may sound like a rule of thumb but it is noted in the older FHWA publication on seismic design/retrofitting of highway bridges. It was generally for bridges in Category A and B.

Reviewing the limits on the response spectra from AASHTO you will note that the coefficient, Cs is limited to 0.25. Thus it's no wonder that you'd be conservative in using this value times the reaction at the bearing.

I've never seen this value applied at the c.g. Assuming that positive connections are required vertically, the moment resulting from that eccentricity would be resisted by the abutment/pile or bent. Generally speaking these systems are large enough to do that. Thus the lateral load simply would be moved down from the c.g. to the bearing.

Regards,

Regards,


Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[mehr27]

Thanks QShake (as usual). Can you picture the couple acting on the bolts and creating uplift? I like your reasoning though.

Thanks.

 

[Qshake]

Sure, I can picture the uplift on the bolts resulting from the couple. Looking at the superstructure as a rigid body ( a simple calculation of the moment of inertia about the longitudinal axis should confirm this) and applying the couple, the couple's load path to the abutment beam or pier cap will need to be resisted by compression on the bearing and uplift on the anchor bolts. Again, you are typically using 1" to 2" anchor bolts (usually swedged) and embedded probably 12-18" in the concrete. In many cases, the anchor bolt has a small spiral around it which will aide in concrete pullout failure mode. Since most bolts that large are of A588 steel tension failure on the bolt is typically not a problem.

But don't misunderstand what I've written for a reason not to check it. I think its a good idea to do so and experience this for yourself. I think its a good part of the logic that your working with. Keep up the good work.

Regards,



Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.