Elastomeric Bridge Bearings with

[dscott]

I am modeling a concrete deck/steel girder 4-span bridge for seiesmic response in the piers using GT-Strudl. So that I only need a single expansion joint, I am going to fix the superstructure at the third pier and use elastomeric bearings at the other 2 piers and 2 abutments. Using seismic design examples from Federal Highway Administration and from co-workers, I have assembled a model that reasonably approximates the bridge. Elastomeric bearings are modeled as springs. Results show that lateral translation of the superstructure is predicted to be 8" at one of the piers. To minimize this, I propose keeper plates that are placed 0.5" to either side of the elastomeric bearings. These will transfer lateral forces down to the pier.

THE QUESTION - How do I model this? What type of member should I introduce to the model to represent something that will transfer force only after it has deflected 0.5"? The only option that immediately comes to mind is to define the lateral elastomeric bearing restraint as a pin rather than a spring - not 100% accurate, but more conservative than what I have now.

 

[Bhat165]

You are using an element which is activated depending on deoformation of teh structure, that means, your problem is of nonlinear in nature. At present I dont know if any software can handle this. If you write code, its straight forward.

 

[Qshake]

You can get around this by using two models, one with the lateral restraint out of the picture thus allowing the structure to move and the second with a shear key in place (which really should be used if that much displacement is expected - it helps distribute the load more uniformly from pier to pier).

Often times we have to model bridges with more than one model to encompass the overall seismic behavior. More proof of that statement maybe found in the FHWA and Preistly's book on the matter. Another area of concern is the abutment and the non-linear action there, first due to possible soil yielding and second is the tension/compression nature.

Another note is why the expansion joint within the bridge - this leaves a major substructure unit vulnerable. It would really be best to place the joints (if necessary) at each end at the abutment, where if something needs to be replaces its not impossible to do with regular maintenance workers. Joints at the piers will lead to premature aging of the pier or bent and may cause access problems for the workers if its over road could restrict traffic below.

Also, many states now use jointless bridges for lengths up to 600'. See Tennessee DOT.

 

[dscott]

The expansion joint is not within the bridge - it is at Abutment A. Pier 3 is fixed. Abutment B will have the deck over the backwall.

FHWA suggests keeping the bridge symmetric. The spans (85'-105'-105'-85'), however, are such that if the bridge were symmetric it would require a finger joint at Abutment B as well, therefore, the decision was made to fix Pier 3 in order to minimize potential thermal expansion at Abut B. The idea is that there is less risk of damage from an earthquake than from a failed joint.

We occasionally use jointless, but our guidelines recommend not to use integral for steel spans > 200'.

Thanks for your previous response.

 

[corus]

If you're using a spectrum response method then you're basically running a static analysis. You'll have to assume that you have the restraint of the bearings if you've already shown that you will achieve the required displacement to make contact.

 

[Qshake]

Wow, you're really putting alot into the FHWA guidelines. I've been designing bridges for quite a long time with most in moderate to high seismicity zones.

Most bridges are layed out to have symmetry but most cannont due to anyone of a number of constraints. Hence, the bridge and spans are located such as to optimize where possible but not always. Mostly we consider the exterior to interior span ratios.

It also seems that finger joints as you noted are a bit much for this bridge and that some consideration should be given to the use of compression seals of 4.5 to 5" range.

Just some thoughts.

 

[pbeng]

At 8" of displacement, isolators or dampers might be an attractive option - keep the displacements down with added damping and you might not need keeper plates. I've heard that GT STRUDL was coming out with a bilinear hysteretic element, that may be of help. You've got to be careful when modeling nonlinear events as linear ones (like the stop). I've had a lot of trouble over the years trying to do that with GT STRUDL - this is something SAP2000 could probably handle pretty easily using a nonlinear spring element with a time history analysis.

 

[pbeng]

Also, on the topic of joints, there are a couple of companies out there that have designed and tested (UC@Berkely) seismic modular expansion joints. Joints tested at velocities of 40 inches/second, large transverse and longitudinal movements. The testing was done in order to fulfill Caltrans requirements for the new East Span SF Bay Bridge project.