Jointless abutment details for long curved bridges?

[MIKE_311]

Does anyone here have any experience using a jointless abutment on a long, curved bridge? The structures im looking at exceed 400' and one is over 800' long.

These lengths, plus the fact that they are curved, exclude the use of integral or semi-integral abutments, Deck extensions are also out as they tend to be limited for curved structures at the 300' range.

VDOT has their Virginia abutments that places a the joint and a drainage trough behind an end diaphragm and in front of the backwall that supports the approach slab. They allow these for curved structure that exceed 300' and expansion lengths length than 1.5".

This is the only detail I have seen that would make a bridge "jointless" for a curved structure. They require periodic maintenance and present constructability challenges.

 

[BridgeSmith]

Not sure why the length or curvature would preclude the use of semi-integral abutments. I designed one 485' long with semi-integral abutments. It wasn't curved or skewed, but with proper detailing, it seems like it could work in those situations, as well.

Due to the expansion length required (and the seismic considerations) we utilized a compressible inclusion behind the end diaphragms.

We had a consultant design one over 600' on a 45 degree skew, but I'd have to look at what abutment type they used. I'll take a look and get back to you.

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

 

[STrctPono]

I will admit that I am at a bit of an advantage (or disadvantage depending on how you look at it) when it comes to integral/semi-integral abutment design as I don't have to deal with the massive temperature swings you would have to however seismic would be a key consideration for me. 400ft long bridge would be less than 1/2" thermal movement at each abutment in my area for a straight bridge. Either way, I agree with BridgeSmith that the length or curvature shouldn't preclude the semi-integral option, but rather necessitates that you detail it carefully and properly. Less than 300ft radius is tight and will create some out-of plane displacements at the abutment. Building a thorough FEA model with the curve and running a temperature analysis could help you understand the movement you need to detail for at the ends. See attached for a semi-integral abutment design that I might consider. Not sure what Virginia DOT detail looks like but I would think that they need to place some restrictions on their standard details and that anything outside of those parameters needs a specific design. You will need to run the model and see how the bridge is actually behaving.

Is this a new bridge or a retrofit? You don't mention anything about the connection at the piers. Are these not a concern?

 

[BridgeSmith]

Tennessee DOT does some very long jointless bridges. They used to have some details available on the web, too.

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

 

[MIKE_311]

According to VDOT, 700'+ straight is acceptable for semi-integral, I've designed straight bridges up to 450' with semi-integral abutments before. The issue here is the bridge is curved and curves over 300' start to get large arc above the chord length and expansion is more difficult to control? Expansion follows along the chord length, maybe moving joints to end of the approach slab creates problems at the end of the approaches? Straight bridge with larger skews need substantial wing haunches to resist thermal forces from the skew.

 

[STrctPono]

Mike, I like the idea of moving the joint to the end of the approach slab (we do this on long straight bridges).... the problem is that when the curved/skewed bridge expands and contracts along the chord length this could cause your approach slab to jam up against your wingwalls or any retaining walls that you might have adjacent to the slab.

 

[BridgeSmith]

...maybe moving joints to end of the approach slab creates problems at the end of the approaches?

It could, especially if the bridge is not fairly wide, but it's not necessary to move the expansion joint to the end of the approach slab. You'd need a bond-breaker on a fairly wide corbel to support the approach slab, but you could have the expansion joint on the end diaphragm. Given the depth of the modular expansion joints I've seen that accommodate lateral movement, having it on (or in) the end diaphragm could actually save some concrete, since housing the joint in the sleeper slab would require a fairly massive sleeper.

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

 

[MIKE_311]

having it on (or in) the end diaphragm could actually save some concrete, since housing the joint in the sleeper slab would require a fairly massive sleeper

This is a really good point that I didn't consider.

 

[3Fan]

Iowa State did some field monitoring of curved bridges with integral abutments. Their report can be found here:
Link

 

[bridgebuster]

3Fan thanks for the report. A few questions, if you don't mind? Does Iowa have height limitations on the abutment stems? In NY, there's an empirical deisgn for stems <20' tall but a refined analysis is required for heights >20'.

What type of piles are used? In NY in concrete filled pipe piles or H piles for spans <165' and H piles for spans >165'.

Thanks. I never designed an integral abutment other than some preliminary studies. They're becoming quite popular in NY.

 

[3Fan]

bridgebuster,

I do not work in Iowa. Our DOT (Ohio) was looking into utilizing integral or semi-integral abutments on horizontally curved bridges and I was doing some research and stumbled across that document. Typically here, we do not use integral abutments if the abutment height is more than just a stub abutment. Even then it is suggested to run LPile analysis on the piles to check for bending in the pile and making sure we can get fixity. We are not limited to span lengths on what piles we use. Typically we use H piles to rock with the pile turned for weak axis bending. Pipe piles are usually used if no rock is in the area.

We cannot use turned back cantilevered walls if it is a state project. But non-state projects we do cantilever a turned back wall off the back of the abutment.

VTrans, FHWA (Design Step 7.1 of the PSC Girder Bridge Example), INDOT, and MASSDOT all have very good detailed examples of integral abutment designs available.

 

[bridgebuster]

3fan, thanks for the reply. I will look at the references you cited. Presently, I'm reading "Integral & Semi-Integral Bridges" by Martin Burke, Jr. Until now, I thought Tennessee DOT pioneered integral abutment bridges; now I know it was Ohio.

 

[BridgeSmith]

Fully integral is our most common type of abutment, since we do alot of short bridges. Typically they have a cap that is 2'-9" minimum, supported by piles driven to refusal. If we have stiff soil or bedrock at shallow depths, we predrill holes and backfill the top 15' around the piles with sand.

We did one fully integral abutment with a 12' high backwall, but the stresses on the abutment and superstructure were difficult to determine with a sufficient level of accuracy. Due to the variability of the soil reactions on the piles, the soil loads to the abutment and wingwalls, and stiffness of the moment connection between the superstructure and he abutment, the range of possible stresses in the abutment was huge. That resulted in the necessity for very robust connections between the superstructure and the abutment, and alot of reinforcing to confine the piles. We opted for a semi-integral abutment the next time.

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

 

[BridgeSmith]

Typically we use H piles to rock with the pile turned for weak axis bending.

I've been trying to convince our head honchos to orient ours with the weak axis in the movement direction, to reduce the moments transferred to the cap under thermal movement, but I have been unsuccessful so far, and we still orient them with the strong axis perpendicular to the abutment. It's resulted in some strange movements on highly skewed bridges.

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

 

[3Fan]

Sounds like an awesome book to cure insomnia! On semi-integral conversions our DOT is using a very similar detail to what STrctPono posted above. A geogrid wall under the approach slab with a gap between the diaphragm and the geogrid wall. Relieves the passive pressure. WVDOH also uses a very similar detail for both integral and semi-integral abutments.

 

[bridgebuster]

yes indeed


Interesting observation Rod. Did the bridges rotate laterally in plan? In NY, the DOT used to orient H piles for weak axis bending. Recently, switched to strong axis bending, saying there isn't any difference. What's piqued my interest again in these types of bridges, one of my co-workers is designing a 170' signle span integral abutment bridge with a 35-degree skew. I suggested that he orient the piles for weak axis bending in the longitudinal direction. My thought was there would be more stiffness transversly to prevent or at least minimize rotation of the structure. His response: Bridges don't rotate laterally. I said to myself, maybe you should inspect at some skewed bridges.

 

[BridgeSmith]

Did the bridges rotate laterally in plan?

Yes. When the bridges expanded, the abutments moved sideways, along the path of least resistance.

The movement sheared and destroyed the joint material between the abutment and the approach slabs.

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

 

[3Fan]

When we work for the WVDOH, we can turn a single row of H piles the strong way and call it semi-integral. I haven't figured that one out yet. In Ohio, you have a rigid foundation for a semi-integral configuration like 2 rows of piles, shafts, or spreads.

BridgeSmith, were there and guides to force the expansion to be parallel to the beams and not along the path of least resistance?

 

[BridgeSmith]

BridgeSmith, were there and guides to force the expansion to be parallel to the beams and not along the path of least resistance?

It was fully integral abutment. Thee are no guides or bearings. The girders are embedded in the abutment. It all moves together.

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

 

[MIKE_311]

His response: Bridges don't rotate laterally. I said to myself, maybe you should inspect at some skewed bridges.

They absolutely do. The passive pressure on the that develops on the backwall acts perpendicular to the face of the backwall (not longitudinal to the expansion direction) and a resultant transverse load results that attempts to rotate the superstructure. This is why semi integral abutments have large wing haunches or shear keys in the end diagram to resist this force.

VDOT has a great example of deriving this force:

http://www.virginiadot.org/business/resources/bridge/Manuals/Part2/Chapter17.pdf

(sheet 62 of 68)

For high skews a semi integral is better since you disconnect the movement and forces and can design the abutment pile cap to resist the forces. Integral abutment are intended to translate with bridge expansion. It's hard to design them to translate and be stiff enough to resist movement!

Does this bridge being design require a design exception? Most DOTs will set a limit for length and skew of integral abutments if they allow them. If your peer is paving new ground I highly recommend they reconsider a semi -integral instead. following the link above, VDOT is pretty generous with length and skew and this bridge is even outside their limits.