Push-and-Release Tests of the New Seismically Isolated Research Centre at the University of Camerino

The isolation system consists of 28 High Damping Rubber Bearings (HDRBs) arranged on the plan perimeter to maximize torsional stiffness, and 36 Low Friction Sliding Bearings (LFSBs) in the interior to support the higher vertical loads. The system was designed according to the general criteria described in [1] and to provide an isolation period of 3.60 s, to ensure no superstructure or contents damage up to the maximum design demand. A shear modulus of 0.4 MPa and a conservative damping ratio equal to 10% (10%-15% is the typical range for rubber compounds commonly used in Europe) were assumed in the preliminary design. The HDRBs are all 600 mm diameter, with a total rubber height of 184 mm leading to a horizontal stiffness of 0.62 kN/mm for each bearing. The displacement capacity is 360 mm (corresponding to the shear strain limit of 200% suggested by the Italian Seismic Code [2] for elastomeric bearings) and which is slightly larger than the Collapse Limit State (CLS) displacement demand for a return period, T_R of 1950 years, which is 324 mm, including the effects of torsion and considering the lower damping ratio value. The LFSBs have a displacement capacity of 400 mm and a dynamic coefficient of friction less than 1% for the seismic vertical load condition. The low friction coefficient is provided by the PTFE interface material and the lubricated sliding surface of the devices. Type tests were carried out according to EN15129 [3] before manufacturing of the production devices, and and factory production control tests (FPCTs) carried out on six HDRBs (20% of the total number of devices) confirmed the design properties, with  a damping ratio of 14 %. It is worth noting that in the European context when LFSBs are used only to support vertical loads (and not provide energy dissipation for the isolation system) the manufacturer can qualify them according to EN1337-1 [4] which only requires material tests and not complete device tests.. However, the Italian Seismic code [2] allows the designer to prescribe additional tests on completel devices. For this project the seismic test prescribed by the EN15129 [3] for Concave Curved Surface Sliders (CCSSs), were required in order to verify that the dynamic behaviour was satisfied and that the coefficient of friction was equal to or lower than the design values (1% and 2% for the dynamic and breakaway friction coefficients, respectively). Further details of the design process can be found in [5]. A unique aspect of this project is that a testing system was designed together with the building to allow testing of the isolated building at the completion of construction, as well as periodically throughout its life. The testing system that was developed involves an innovative removable, push-and-release device, together with a reaction wall and a push loading system. The push-and-release device consists of a quadrilateral articulated steel system equipped with a sacrificial, replaceable steel rod made of high-strength steel in the middle of the central tie element (Figure 2), that acts as a fuse. To avoid excessive transverse displacement of the quadrilateral system after the fuse failure (due to the sudden release of the accumulated energy) a looped cable with a friction dissipative device was put in parallel with the fuse (Figure 2c).  Two hydraulic jacks with a maximum stroke of 300 mm and total maximum force of 10 MN (5 MN for each jack) and two load cells are positioned in series with the quadrilateral system. The building was equipped with accelerometers at each floor, strain gauges on steel braces and displacement transducers at the isolation level and first floor. In this article, only the response of the isolation system is discussed, further information regarding the other aspects of the experimental data processing and building response can be found in [6].

Figure 1: (a) design floor plan and section, (b) construction phase and (c) completed CHIP building.

Figure 2: (a) Quick-release device with hydraulic jacks and load cells, (b) high-strength rod fuse after rupture, and (c) friction dissipative device (looped cable).

The experimental tests were performed when construction of the steel structure was complete (before installation of partitions, external walls, architectural elements, and laboratory research equipment). At this stage of construction the total estimated mass of the structure above the isolators was 4633 tons. Several tests were performed as summarized in Table 1: a preliminary test (S1) to verify the load equipment where a force of about 500 kN was applied (without moving the building), two quasi-static tests (S2 and S3) characterized by a slow loading phase followed by an slow unloading  phase where increasing and decreasing displacements were imposed respectively, and four dynamic tests (D1, D2, D3 and D4), consisting of a slow loading phase (up to the rupture of the fuse) followed by a sudden release of load and free vibration. The maximum horizontal displacements reached in the quasi-static and dynamic tests were 285 mm and 227 mm, respectively, which were very close to the design displacement. The tests can thus be considered an in-situ verification of the design-level seismic behaviour of the isolation system. It should be noted that when HDRBs are used (alone or in hybrid isolation systems) a push-and-release testing procedure  is not entirely equivalent to an earthquake excitation, due to the viscous behaviour of this kind of bearing causing a loss of the accumulated elastic energy during the extended duration, pseudo-static loading phase, due to relaxation phenomena of HDRBs characterized by a viscous behaviour sensitive to the load velocity. The velocity-dependent behaviour of the high-damping rubber can be seen in Figure 3a, where the isolation system horizontal force-displacement curves of all the release tests are shown. For dynamic tests the force during the release has been derived from the acceleration recorded at the isolation level and the mass of the structure and is indicated by  thinner solid line in the figure. The following observations can be made: all the tests are substantially overlapping except for test D1 which was influenced by the breakaway friction force of the sliders and by the first cycle effect of the virgin rubber of most of the HDRBs [7-11] (only six bearings had been previously tested for FPCTs, and for these the virgin behavior was mostly recovered); in the subsequent tests lower friction forces were observed (due to the lubrication of the sliding surfaces), compatible with the design coefficient of friction of the sliders; the unloading branches of the dynamic tests are significantly different from those of the static tests due to the different velocities involved and the viscous behaviour of the rubber. Furthemore, it can be observed that there are marginal residual horizontal displacements after the tests, thus, apart from the very first test the building did not exactly start from the same initial position. Residual displacements recorded at the end of each test and those measured at the begin of the following test are reported in the last two coulums of Table 1. The final residual displacement measured after one month after test S3 was equal to 25 mm, in accordance with the expected friction force of the sliders, once viscous over-stresses are completely exhausted. This final offset was considered acceptable, thus the building was not loaded back to its original position.

	Test	Starting time	Duration [min]	Max force [kN]	Max displacement [mm]	Displacement at test start [mm]	Displacement 30 min after the end of the test [mm]
3rd July 2020	S1	12:22	6	518	0	0	0
	D1	13:47	16	2729	177.3	0	15.7
	S2	16:36	21	3206	232.4	15.2	26.1
6th July 2020	D2	12:24	8	1756	109.4	22.1	22.4
	D3	13:36	18	3122	226.9	22.1	23.4
	D4	15:56	9	1786	121.8	22.1	22.9
	S3	17:07	25	3834	284.6	22.6	36.2

Figure 3b shows a close-up of the release phase of the dynamic tests, where it can be observed that the system oscillates around a non-zero displacement during the release phase, exhibiting a highly-damped cyclic behaviour. To better understand this behaviour a model able to simulate all the viscous effects of the HDRBs as well as the energy dissipated by the LFSBs was developed and calibrated based on test D1 (Figure 4). The model consisted of three in-parallel elements: a Kelvin model characterized by the stiffness k₀ and the damping constant c₀, a Maxwell model characterized by the stiffness k₁ and the damping constant c₁ and a friction model characterized by the friction force F_fr. With this model, the over-damped cyclic behaviour can be simulated as well as the viscous displacement accumulated during the loading phase, causing the system to oscillate around a non-zero displacement during the release phase. A comparison between the observed behaviour and the response predicted by the model is shown in Figure 5, together with the lost displacement, , and the lost velocity, , of the Maxwell model of Figure 4.

Summary

This article has given an overview of the experimental testing of the CHIP building at the Camerino University (Italy). The CHIP building is a seismically isolated, braced-frame steel structure with a hybrid system consisting of 28 HDRBs and 36 LFSBs. The global response of the building was tested by means of a push-and-release device designed together the building, up to displacements similar to those induced by design extreme seismic events. However, the applied testing procedure is not fully equivalent to an earthquake excitation, mainly due to the different deformation velocities applied during the loading phase and the not negligible viscous effects of HDRBs. This research work was aimed at understanding and evaluating the observed experimental behavior by assessing and calibrating a model able to simulate all the viscous effects as well as the energy dissipated by the sliding bearings.  

References:
1. Micozzi, F., Scozzese, F., Ragni, L., and Dall’Asta, A. (2022). Seismic reliability of base isolated systems: sensitivity to design choices. Engineering Structures, 256.
2. DM 17 gennaio 2018 (2018). Aggiornamento delle «Norme tecniche per le costruzioni» (in italian).
3. European Committee for Standardization (2009). EN 15129:2009 Antiseismic devices, European Committee for Standardization, Brussels.
4. European Committee for Standardization (2000). EN 1337-1:2000 Structural bearings – Part 1: General design rules, Brussels
5. Dall’Asta, A., Leoni, G., Micozzi, F., Gioiella, L., and Ragni, L. (2020). A resilience and robustness-oriented design of base-isolated structures: The New Camerino University Research Center. Frontiers in Built Environment, 6, 50.
6. Dall’Asta, A., Leoni, G., Gioiella, L., Micozzi, F., Ragni, L., Morici, M., Scozzese, F., and Zona, A. (2022). Push-and-release tests of a steel building with hybrid base isolation, Engineering Structures, vol. 272.In 18th World Conference on Seismic Isolation (18WCSI).
7. Dall’Asta, A. and Ragni, L. (2006). Experimental tests and analytical model of high damping rubber dissipating devices, Engineering Structures, 28 (13), 1874-1884.
8. Ragni, L., Tubaldi, E., Dall’Asta, A., Ahmadi, H., and Muhr, A. (2018). Biaxial shear behaviour of HDNR with Mullins effect and deformation-induced anisotropy. Engineering Structures, 154, 78-92.
9. Tubaldi, E., Ragni, L., Dall’Asta, A., Ahmadi, H., and Muhr, A. (2017). Stress softening behaviour of HDNR bearings: modelling and influence on the seismic response of isolated structures. Earthquake Engineering and Structural Dynamics, 46, 2033-2054.
10. Thompson, A. C. T., Whittaker, A. S., Fenves, G. L., and Mahin, S. A. (2000). Property modification factors for elastomeric seismic isolation bearings, Proceedings, 12th World Conference on Earthquake Engineering, Auckland, New Zealand: New Zealand Society for Earthquake Engineering.
11. Grant, D. N., Fenves, G. L., and Whittaker, A. S. (2004). Bidirectional modelling of high-damping rubber bearings, Journal of Earthquake Engineering, vol. 8, pp. 161–185.