How to create a rubber bearing with link element

[Vincentzili]

Hi,
I am new to Seismostruct.
I am modelling a ramp bridge and going to analysis the response of the bridge under earthquake.

I have a problem with creating the Rubber Bearing that link the piers and the beam. I am not sure what kind of value I should enter inside the F1, F2, F3, M1, M2 and M3 to create a rubber bearing link element. And I am not sure how to link it in Element Connectivity.

Besides the questions above, I am not sure how to make the beam - pier connection to be flexible or fix neither.

I hope if you can help me out. Thank you!!

[huffte]

You'll need two coincident nodes. The first one will be connected to the substructure, the second one to the superstructure. These two nodes will be the first two node specifications in the definition of your bearing element, which as you know will be a nonlinear link.

A third node should be located vertically above, or below, the first two. This will define the "1" , or axial direction of the link. A 4th node defines the 1-3 plane of the link.

As for properties, I am afraid that you'll have to decide what you wish those to be yourself depending upon the type of bearing you wish to model. You might consult come manufacturer's literature for the range of loads and movements you expect.

See "Element Classes" -> "Link Element - Link" sections of the help system, which is excellent. Also see subtopic "Bilinear kinematic curve" within the "Link Element - Link" section of the Help system. This particularly element is one with which I have had much success modeling lead-rubber isolator bearings. If you are interested in friction - pendulum bearings, I have found it possible to model these using the "Plst" link element.

Best of luck Vincentzili - modeling a curved bridge with bearings is quite an interesting project.

[Vincentzili]

Thanks for your help huffte.

The third and 4th nodes are non structural nodes right? and if I am using circular bearing, I could just define the axial direction of the link with one node. Is that right?

So what about defining the rigid connection for the beam-pier link?

[huffte]

Vincentzili,

The 3rd and 4th nodes may be either structural or non-structural. If your model already has structural nodes at the locations which will provide the proper orientation of the link, then there is no need to create additional non-structural nodes for this purpose.

I'm not certain what you mean by the phrase "define the axial direction of the link with one node". A link will always require 4 nodes. A circular bearing would have identical properties in the 2 shear directions for both translational and rotational stiffness.

Yes, the bottom of an isolator or bearing (link node 1) would typically be attached to the pier with a rigid link, while the top of the isolator (link node 2, coincident to link node 1 in the unloaded state) would be typically attached to the girder/superstructure with a second rigid link.

Of course, you are the one who knows the system you wish to model, so there could certainly be exceptions to this. But your proposed method sound realistic for typical bridges.

Best of luck Vincentzili.

[Vincentzili]

thank you again huffte.
But I am still not really understand the part that make the Beam-Pier connection became rigid or flexible.

[huffte]

Vincentzili,

I have e-mailed one of my bridge model files to you. This model incorporates isolators at the bearings and will hopefully clarify some things for you.

Please note that my isolator (link) properties will certainly not be applicable to your particular case, but the modeling strategy will become evident and will hopefully shed some light on the subject for you.

[Vincentzili]

Huffte thank you very much !!!!
I am an Uni student, so I am not really very good at this.