Varying length PS girders

[MIKE_311]

Due to some site constraints I have a bridge that may require a skewed abutment. The beams are PC bulb tees. The bridge is 3 spans, 62'-75'-62', simple made continuous. The end span would now vary from 62' to 75'.

Are there any issues with varying length PS beams made continuous that couldn't be worked out with proper detailing? e.g. adding additional continuity reinforcement in the deck for negative moment.

At this point, i'm just looking for options and trying to identify any significant design challenges, and whether this is a viable option to present.

 

[Agent666]

I'm not a bridge guy by any means but I guess aiming for similar magnitude deflections on all three girders would be my first thought to keep things reasonably compatible. Whether that means larger sections for larger spans or higher levels of PT reinforcement and conventional reinforcement.

 

[STrctPono]

Ideally, you would at the very least like your abutment and piers to have parallel skews. However, that is not always the case, and when site constraints or curved bridges necessitate the need for girders with varying lengths then so be it. I don't see a lot of issues with the strength considerations. This can (as you pointed out) be detailed accordingly relatively easily so long as you run an accurate analysis. The main 2 issues I see are as follows...

1. Calculating out the camber growth and deflection of each girder relative to each other and making sure that they both fit within the constraints of the expected and tolerable concrete haunch over the girder. Since your girder closest to the acute corner of the deck will have the longest span and deflection this will be the most critical when figuring out if it will work geometrically.

2. If you have a curve, make sure that your creep blocks at the abutment don't lock up your bridge for daily use (traffic, temperature, shrinkage) since all these effects could potentially have a displacement component perpendicular to the face of the creep block. If curved, we always like to leave a large enough gap between the creep block and girder to allow for this motion but not too large such that you fail your elastomeric bearing prior to engaging the creep blocks for a seismic event. Even if there is no curve you will also notice that your skewed abutment will have a unique behavior when you run your response spectrum analysis. You may get a very large out of plane bending moment at the top of your abutment wall that you will need to address.

This is all I can think of for now.

 

[wiktor]

@Mike
There will be no issue with the design - the decks on the PC bulb tees are designed as a hinged slab, so some extra reinforcement in the slabs would be required - but the increase of the spans will be gradual.
With the hinged slab approach, the issue of the uplift in the acute corner is almost non existing, so there is only issue of proper modeling and detailing.