HARIS (Highly Adaptable Rubber Isolation System) is to improve and validate alternative design layouts of elastomeric bearing for seismic isolation of bridges. The overall objective is to design, build and analyze economic and technical advantages of an adaptable non-isotropic layout, able to achieve different stiffness in in-plane directions. The main part of the research effort focus on the experimental determination of mechanical characteristics of adaptable non-isotropic layouts of high-damping natural rubber bearings. Different sizes and different layouts are considered to investigate compressive and shear behavior, the strain at failure, energy dissipation and equivalent damping properties. Innovative designs and layouts are always compared to standard bearings for a clear and objective interpretation of test data. Evaluation of test results clearly proof that inclined steel plates produce an increase in stiffness for shear deformations up to approximately 100% strain. The behaviors of all bearings tend to converge for larger strain amplitudes to the standard design due to flattening of steel plates.
The Artificial Mass Simulation method is used to scale a one-span concrete girder of the proposed Taiwan High Speed Rail System. The 1:6 geometric scale model of the girder is used for dynamic characterization of the adaptable non-isotropic bearing design. Shaking table tests are performed to determine natural frequencies and damping properties for harmonic base excitations. Following forced harmonic motion, real acceleration data from past seismic events are used for tri-axial seismic excitation to determine acceleration and displacement response spectra. The non-isotropic bearing considerably reduces the relative displacement induced by medium amplitude earthquakes when compared to the standard design. The reduction is less notable for high amplitude motion where the steel plates flatten, thus reducing the stiffening effect.
Constitutive equations for isotropic hyperelastic materials are used for numerical characterization of the proposed design. Different strain energy formulations are evaluated for stability and accuracy of the numerical algorithm. It is found that the Ogden model is capable to provide a numerical stable response up to high strain values and in general a better fit to given experimental data of high damping natural rubber. A pressure dependent bulk modulus is used to formulate a cavitation model for rubber like materials. In fact, rubber in seismic bearings may be subjected to negative hydrostatic pressure due to rotation of the steel reinforcing plates and/or due to inclined laminae. Thus, it is recognized that cavitation is an important issue in rubber bearings and must be implemented in any numerical algorithm to be able to capture the physical response of the device.