Seismic analysis of a single span bridge according to AASHTO LRFD

[Hassaan_16]

Hi! We are designing a single span bridge comprising spill-through abutments at the ends instead of retaining wall abutments. The abutments are initially planned as pile bent piers with height (till depth of fixity from NGL) approximately 15m. The piers will protrude from the fill for 3m till the pier cap.
AASHTO LRFD suggests no seismic design be carried out for single span bridges but I think this one should be analysed and designed for seismic forces. My reason for this is that unlike abutment retaining walls the spill-through abutments might be able to move separately from the ground (due to insufficient restraint provided by the retained fill to the piers) and will require seismic analysis and design to be carried out. Am I right with my concerns? Please guide.
Thanks.

 

[bridgebuster]

Generally speaking, we're allowed to be more stringent than AASHTO.

Is a spill-thru abutment being used because the bridge could be lengthened in the future? If so, a seismic analysis would be needed because in the future your abutment becomes a pier.

When you say "spill-through" aren't you providing a backwall behind the stringers or are you constructing an integral abutment? From your third sentence, I assume you're driving piles into existing ground, then backfilling 3m. That's why I asked if you're actually constructing an integral abutment. Anyway, in the pile analysis, you would have an extreme load case.

 

[BridgeSmith]

A sketch of the proposed configuration is needed for us to understand how seismic displacements might affect the connection.

That said, if it's going to be on 15m piles, the typical abutment type would be fully integral. We wouldn't even consider a different substructure type for a single span bridge on piles.

 

[Hassaan_16]

Thanks for the replies. I see my post didn't portray the situation properly, so I've attached a handmade sketch of the proposed configuration.

For bridgebuster's queries:
-The bridge isn't planned to be lengthened in the future.
-You have correctly understood the situation you mentioned in your last paragraph. A backwall is provided behind the girders (RC/Post tensioned) and it isn't an integral abutment. For now, we're not considering providing integral abutments here.

 

[BridgeSmith]

With that type of configuration, we would most likely restrain the elastomeric bearings top and bottom, and provide shear keys to limit the transverse movement to a displacement less than the shear deformation limit that would damage the pad. That limit is larger than you might think. I asked what that limit was for some LRB seismic isolators preliminarily designed by a designer/manufacturer (DIS), and the answer I got surprised me - 9" lateral for a 7.375" tall pad. Of course, if you use keepers instead of bonding the pads to the top and bottom plates, in extreme circumstances, the pad would just roll over.

As I said, it's unlikely we'd ever consider anything other than fully integral (with elephant ear wingwalls) for that situation. If necessary due to shallow depth to rock, we would predrill to give the the steel H-piles about 3m of depth to flex. Although, I'm not sure I understand where the proposed soil backfill extents are.

 

[bridgebuster]

Hassaan - what's the purpose of the diagonal brace, is that permanent? Your piles look to be robust; they should be able to take up the lateral load.

 

[BridgeSmith]

I was wondering about the diagonal brace also. If this bridge is over a creek, the diagonal brace would need to be embedded adequately below the scour elevation. We've sometimes done 2 rows of H-piles with the front row battered sort of like that, except only on a 4 to 1 (vertical to horizontal).

 

[Hassaan_16]

My apologies for the confusion. As I mentioned, this is a spill through abutment. The diagonal "element" as it seems in the sketch is actually just stone pitching for slope protection of the retained fill. I forgot to mention this in the sketch.

 

[bridgebuster]

OK, thanks for the clarification.

 

[BridgeSmith]

I'm still not clear on where the proposed ground surface is and where or if there's something acting as a retaining wall.

 

[DanielSgeo]

I think if your pile spacing is 3B or less, for longitudinal loading, you effectively have an abutment wall.

 

[VoyageofDiscovery]

Yes, design the abutment for seismic. As a minimum the concrete bearing area should have sufficient end and edge distance to not allow the girder end to slip off.

 

[tatox]

So if this bridge in SDC D for example, the force from earthquake to design connection in substructure to upper structure usually some dowels = As x Dead Load Reaction / Response Factor (0.8) will be transferred to pile right ?.
So the pile will need to be designed for the earthquake design connection force by Response Factor elastic (1) for SDC D? with maximum displacement for earthquake criteria?
I saw the Seismic AASHTO is stated the soil fill behind us must using 0.5 As for force passive earthquake, not much about the earthquake for foundation. The flowchart B2 said only Determine design forces (4.5) -> Determine min support length (4.12) -> Design Complete.

If we got SDC D, and high PGA the foundation will be quite large.
shear key ? not mentioned i think shear key is refer to the connection also

 

[BridgeSmith]

For single span bridges, regardless of SDC, AASHTO basically only requires adequate bearing length to prevent unseating of the superstructure from the substructures. My understanding is that this is primarily due to the assumption that there is no differential substructure movement or lag, as there would be for a multi-span bridge with intermediate substructures. IOW, it pretty much all moves together.

 

[tatox]

@BridgeSmith true. The whole bridge will act together, since the dowel connection will also placed in the rolled bearing also (assumed as hinge bearing - rolled bearing).
Even though there is small differential in the rolled bearing that dowelled with some space gap.
So the force that hold the connection, Is this EQ force need to be transferred by foundation to the soil?
Is there any potential failure at the pile foundation perhaps near pile bent or pile cap for dissipating the holding connection EQ force.