Renovating Box Beam Non Composite Deck to Composite Deck

[minorchord2000]

Gentlemen:

I am analyzing an old 3 span non-continuous, pre-cast box girder bridge, with simple spans of 49.25 feet each. The superstructure is 27 feet wide and consists of 9 voided box sections 3 feet wide and 27" deep. Currently the driving surface is only about 2" of asphalt. Our scope of work is to see if we can make the 3 spans continuous for live load be removing the asphalt and replacing with a 5" concrete deck and reinforcing the deck at the supports to resist the tension due to negative bending. I am using the AASHTO LRFD code. My issue is how to get composite action between the proposed deck and the top of the existing precast girder. The top flange of the box is only 5 inches thick. Is there another way to get composite action without the use of rebar dowels drilled and epoxied into the top flange of the girder which is only 5" thick? I do not think that this is feasible since I cannot develop the bars in that short distance. The bridge was built in 1958. I know I can roughen the top flange of the girder by scabbling, but I do not think that will be enough to get true composite action. Does anyone think that epoxy adhesive applied to the top of the deck will help? Any ideas will be greatly appreciated. Thank you.

 

[BridgeSmith]

The dowels would not need to be developed for tension, since they would be in shear, but you have a couple of other issues with doweling into the top slab. First, adhesive-anchored dowels usually aren't considered to be stiff enough to provide adequate composite action. The other issue is that analyzing the shear flow from the top slab to the sidewalls for the increased loading carries a lot of uncertainty.

If I was faced with this situation, I would approach it by providing continuity from the sidewall reinforcing to the new reinforcing in the top of the new deck. This most likely would required chipping away the concrete above the sidewalls, mechanically splicing the vertical steel extending up from the sidewalls to new bars that are lapped with new reinforcing in the top of the deck. Alternately, you could remove the entire top slab and cast a new continuous top slab for the box girders.

The third option, and possibly the most cost-effective, is to replace the superstructure in its entirety. Prestressed girders that have bars or strands extended so that they can be spliced, or continuous steel girders could work for this option.

Rod Smith, P.E.

 

[wiktor]

The bridge is the most likely BI-36 typical AASHTO design. These were reinforced only with the prestressing strands (

https://www.pci.org/PCI_Docs/Design_Resources/Transportation_Resources/Box Shapes.pdf).

As Hotrod10 pointed epoxied dowels are out of question, and so is any epoxy applied in between the girder and an overlay concrete, as the epoxy have a tendency to creep under load.
Short dowels with maximum 4" embedment grouted into top flange may work. The crucial issue will be the condition of the prestressing strands - if good you may squeeze another 40-50 years of service.
You may lift the bridge at the piers prior to making in continuous, and lower after casing of the new overlay, to post tension the zones above the piers.

 

[minorchord2000]

HotRod10 and wiktor: Thanks so much for your responses. The prestressing tendons are straight and are all within the bottom flange of the girder. There are 26 straight 3/8" diameter strands (0.08 in^2/strand). The shear reinforcing is located within the centerlines of the vertical stems of each girder. According to LRFD, I need a minimum interface shear area of steel of 0.36 in^2/ft per girder or 0.18 in^2/ft per girder stem. Each of the 5 spans is 49.25' long and 9 boxes wide. Nine boxes includes 18 vertical stems. Five spans = 5 x 49.25 x 18 stems = 4433 dowel locations. It does not seem cost effective at all to make this 5 span bridge continuous by trying to provide suitable interface shear reinforcing.

Tony G

 

[BridgeSmith]

Retrofitting a prestressed girder bridge to make it continuous for live load is rarely an economical proposition, especially with that many spans. Beyond making the top composite, it can also be a trick to get the compression in the bottom flange to work out without crushing the grout or the bottom flange.

Getting new prestressed girders of the same size, with cast-in reinforcing stirrups extended for the cast-in-place continuous slab, could be a good option.

Btw, when we rehab bridges, any design work is done per the spec. for which the bridge was designed. For now, that's typically the AASHTO Standard Specifications, not the LRFD.

Rod Smith, P.E.

 

[minorchord2000]

Yes HotRod:

I thought of that. This old bridge uses 250 ksi tendons f'c = 5 ksi for the girders and grade 40 for the mild steel. I will use the oldest AASHTO we have. Thanks again for your timely and thoughtful responses. The existing stirrup layout could be the deal breaker in trying to add a 5" concrete slab to the existing girders even non-compositely.

 

[BridgeSmith]

If the spans are less than 55', a prestressed tri-deck section could also be a viable option for full superstructure replacement as simple spans; the 27" deep section is a very common size, with variable flange widths to accommodate almost any overall width. They might be slightly heavier than the existing box girder, but possibly not. It can be tough to keep the joints sealed over the long run, so if water leaking onto the pier cap would be detrimental, the tri-deck may not be the best solution.

If the loading on the substructures is a concern, you may want to consider steel girders to cut down on the weight of the superstructure.

Rod Smith, P.E.

 

[wiktor]

You may also find this article useful -

http://www.concretebridgeviews.com/i79/Article1.php

This option may be used if the spans have enough capacity to accommodate 5" slab, still working as simple supported.