Prestressed concrete girders and camber

[MIKE_311]

So I'm more of a steel guy...

When designing prestressed concreted bridge girders, I know you want to keep the cambers of the beam pretty close. So what is the acceptable difference tolerance?

If I have:

An interior girder design the has a release camber of 2" and a final after SDL of 1/4"
The exterior girder design the has a release camber of 2" and a final after SDL of 1"
Beams are PCBT-29 with a 7'-4 1/2" spacing.

These are two independent beam runs. The deflection difference is the result of the Deck + Haunch loads in the interior and exterior beams. So before the deck hardens, the interior girder will deflect more than the exterior. Do I need to treat this like steel and pre-camber for this deflection or is 3/4" difference acceptable provided I provide enough of a haunch to accommodate it?

I can add more strands to the interior girder but I don't have enough experience with PS beams to know how much deflection difference is acceptable.

 

[OSUCivlEng]

I've never worried about PC beam camber. I give the contractor the deflections due to the deck, diaphragms, stay in place forms, rail and haunch. The size (span) of the beam determines the haunch thickness I use. If it's a standard shape then it's been done before and shouldn't be a problem.

How accurate are the camber predictions for prestressed beams? I've never heard of anyone trying to pre-camber a pc beam before. Also, don't use the long term deflection modifiers, they're a joke.

 

[MIKE_311]

The beam shapes are standard but we need to design the strands. The stage agency says no negative camber after superimposed dead loads so we need to provide enough prestressing in the beam.

During the QC, a co-worker of mine brought up the issue of the different final camber values on the beams saying he had this issue on another project he worked on (different agency). He had to design the beams to have the same final camber after dead loads. Every existing PS bride Ive seen has the same beam and strand pattern, i cant recall a different pattern for interior/exterior beams.

He agreed that he didn't know where the threshold was for what was an allowable difference.

 

[STrctPono]

Mike,

Normally you provide yourself enough play with the concrete haunch between the girder top flange and deck that this haunch can make up for any variation between adjacent girder camber/deflection differences. There are other roundabout ways you can help reduce the difference in camber differences between adjacent girders such as changing the seat heights. For standard precast, pretensioned girders, we would certainly calculate the expected camber but it needs to be taken in context with everything else. You start with finish deck elevation, and subtract deck thickness. You then work your way from bearing elevation on up. So long as both girders with their camber can fit within the standard haunch height then it doesn't really matter if they are 3/4" difference from one another. I don't have an allowable tolerance for this but this is exactly what the haunch is for, so as long as your camber isn't eating into the deck, then you should be fine.

I would not omit strands from the exterior girder. I can't seem to find the exact provision, but I believe there is an AASHTO requirement that the exterior girders have the same load carrying capacity as the interior girders. This makes sense, since if the bridge were ever planned for a widening, this would create complication for future design with a weak interior girder.

There are some bridges where calculating and designing around the camber can be cumbersome. Precast plank bridges can be obnoxious since there is no haunch, and thus no room for tolerance. However, the issues here are rarely due to differences in expected camber and rather differences due to inconsistencies with the precaster. These are typically addressed via allowable camber tolerances in the specifications. Bridges without haunches are another one. I did one bridge like this and it wasn't fun. The bridge was on a skew so it made it even more difficult.

How accurate are the camber predictions for prestressed beams? I've never heard of anyone trying to pre-camber a pc beam before. Also, don't use the long term deflection modifiers, they're a joke.

The camber certainly can vary wildly than what is calculated. It's hard to predict. I don't think the issue is so much pre-cambering a girder intentionally but rather trying to calculate what the expected camber will be. But I think the point you are trying to make is to not play with the strands to produce a desired camber, which I agree on.

For spliced/staged post tensioned girder bridges, we will always check the camber using a construction stage analysis in an FEA program that can take into account the different stages, stressing, concrete shrinkage, and creep. For one particular bridge, since there was no deck, we had the Contractor let the concrete age 90 days before post-tensioning to reduce the camber growth since we didn't have the benefit of the deck to counteract it.

 

[OSUCivlEng]

I also think putting less strands in the exterior girder would be a violation of AASHTO LRFD Bridge Design specs. The exterior girder is not supposed to have less load carrying capacity than interior.

 

[MIKE_311]

Its not really a violation, AASHTO says you are allowed if you are pretty darn sure the bridge wont or can't get wider in the future.

 

[BridgeSmith]

You may be able to get the same capacity for interior and exterior, while equalizing the camber, possibly through the use of more draped or harped strands in the interior girders and/or sacrificial (temporary) strands in the tops of the exterior girders. I suggest working with a prestressed girder supplier about these issues - they seem to have ways of doing magic to adjust camber.


Rod Smith, P.E., The artist formerly known as HotRod10