Main Bridge and Viaduct Approaches

[X-Wing]

Okay we have this bridge that is curved in plan. The approaches were made to be viaduct style instead of introducing massive retaining walls and embankments.

Question: Does the seismic design of the "Main Bridge" be separated with the viaduct? The main bridge will be on Type VI AASHTO Girders and the viaducts will be on flat span, 6m spans (I suppose)

Thanks!

Very Truly Yours,

- andru18

 

[Lomarandil]

I suppose that depends on what type of joints you have between the approach and main spans.

 

[X-Wing]

@Lomarandil

I suppose the connection between the main bridge and the joining viaduct will be on simply supported with joints. The approaches will be on curved plan.

Very Truly Yours,

- andru18

 

[Lomarandil]

I'll admit, this is only moderately informed speculation on my part.

I would first attempt to size the joints between the main span and viaduct to be large enough to take up the longitudinal movement of the main and viaduct spans under seismic load (e.g. so they can't contact each other).

If that isn't feasible, I suppose you would either have to get more detailed in your analysis of the separate structures (if they had the same period, you might argue that longitudinal motions are in phase so the gap never closes under seismic load) or design the structures together.

I recall ASCE7 had some provisions for buildings with separations less than their seismic drift. That might be a start for your design philosophy.

 

[yakpol]

In transverse direction the pier supporting the main bridge and the approach will be affected by both structures. Typically I model bridge from end to end regardless of the type of the superstructure.
Yakpol

 

[X-Wing]

@lomarandil
That would be a good start, limiting the drift/displacement and force in the ends of the main bridge so not to affect much the viaduct

@yakpol
One question sir, since the main bridge is curved, and the viaduct is also curve, will there be a problem in the earthquake force application in transverse and long direction?

To all: I think this is entail modelling the main bridge separately with the viaduct, and modelling it end to end, and the viaduct span per span.


Very Truly Yours,

- andru18

 

[yakpol]

Andru,
Depending on the level of engineering (preliminary or final) and the earthquake zone, you can implement some simplifications in your analysis. My experience working in AASHTO seismic zones C and D, that I hardly ever got away with the separate models for lateral analysis. Especially with curved bridges the three-dimensional model is the must.
In response spectra multimode dynamic analysis you will need to apply acceleration spectra in two ortogonal directions and then to combine results using standard combinations X+0.3Y and 0.X+Y, which will take care of the bridge curvature and will envelop dynamic forces.
Hope it helps!