S.D. Campbell/Kinetics Noise Control
EXPERIMENTAL TESTING
An experimental program was designed to verify the theoretical analysis and to investigate the important aspects of the behavior that cannot be mathematically predicted but must be determined from physical testing. The testing program determined the static and dynamic stiffness of the bearings, checked for debonding of the elastomer and carbon mesh under extreme loading, and investigated the long-term deflections through creep testing. All testing was independently performed at the Multidisciplinary Center for Earthquake Engineering Research (MCEER) at the University at Buffalo, State University of New York.
The test specimens measured 200x200 mm in plan and were composed of five elastomer layers (either natural rubber or neoprene) separated by either steel plates (3.2 mm thick) or a carbon fiber mesh (0.8 mm thick) and the bearings had a cover of 6.5 mm of elastomer on each side. The elastomer layers measured 6.5, 25.4, 9.5, 25.4, and 6.5 mm in thickness. Several specimens were obtained by cutting larger samples down to 200x200mm, thereby leaving some edges without an elastomer cover over the reinforcing. A summary of the specimen properties and dimensions is shown in Figure 2 and Table 1.
| Table 1: Bearing test sample properties and description | ||||
|---|---|---|---|---|
| Spec. No. | Elastomer | Durometer | Reinforcement | Description |
| 1 | Natural Rubber | 69 | Steel Plates | Molded at tested size |
| 2 | Natural Rubber | 76 | Carbon Mesh | Molded at tested size |
| 3 | Neoprene | 60 | Carbon Mesh | Molded at tested size |
| 4 | Natural Rubber | 74 | Carbon Mesh | Cut from large sample - 4 sides cut |
| 5 | Natural Rubber | 73 | Carbon Mesh | Cut from large sample - 3 sides cut |
The test samples were based on pads used on a previous project and had a design dead load of 111.2 kN and live load of 89 kN. Testing began with a debonding test wherein the pad was loaded to 150% of the total design load (300.3 kN). After five minutes of sustained load the specimen was visually checked for debonding of the reinforcement and elastomer. A static load-deflection test was performed next, up to a total load of 200.2 kN. Since the static and dynamic stiffness of the pads can be significantly different, cyclic tests were performed at 5 Hz, 10 Hz, and 12 Hz. The upper loading frequency of 12 Hz was based on the capacity of the testing machine. The specimens were subjected to the dead load, followed by cyclic displacements of between 0.5 mm and 1.0 mm, depending upon the frequency. Finally, to allow creep estimations, a constant load (111.2 kN) was applied and the displacements were measured over at least six hours.
Results for all the tests, along with a plot of pad natural frequency versus loading frequency, are presented in Table 2 and Figure 3. All specimen results were adjusted to a durometer of 70 to allow for comparison of stiffness and natural frequency between the specimens. The adjustments in stiffness are based on the change in modulus with durometer as reported in [3] and are approximate. The creep, as a percentage of the initial deflection, was estimated at 25 years. The method uses six hours of load-deflection data to estimate the modulus, and hence creep, as a function of time [4].
| Table 2: Specimen testing results. Stiffness values are in kN/mm, frequencies in Hz, and 25-year creep in percent of initial deflection. | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Static | 5 Hz | 10 Hz | 12 Hz | |||||||
| Spec | Debond | Stiff | Freq | Stiff | Freq | Stiff | Freq | Stiff | Freq | Creep |
| 1 | No | 32.2 | 8.48 | 46.4 | 10.18 | 51.9 | 10.77 | 52.8 | 10.86 | 4.5% |
| 2 | No | 38.9 | 9.32 | 53.5 | 10.93 | 59.3 | 11.51 | 59.2 | 11.50 | - |
| 3 | No | 30.7 | 8.28 | 57.7 | 11.36 | 66.3 | 12.18 | 69.7 | 12.49 | 12.9% |
| 4 | No | 37.7 | 9.17 | 52.2 | 10.81 | 57.8 | 11.36 | 59.9 | 11.57 | 12.4% |
| 5 | No | 35.2 | 8.87 | 51.5 | 10.74 | 58.6 | 11.44 | 60.5 | 11.63 | 5.9% |
No debonding of any specimen was noted during testing, the restraint provided by the reinforcing remained effective for the full duration of the overload. Figure 4 shows the deformations and reinforcement restraint for Specimen 5 during the debonding test. Typical values of allowable creep over 25 years range from 25% to 45% of the initial deflection, depending upon durometer [5]. All the tested specimens easily meet the creep limit.
The stiffness and natural frequency data, approximately normalized to a 70-durometer elastomer, shows the new bearings to behave comparably to steel-reinforced bearings. The static natural frequency of all the bearings was within ten percent of the reference bearing frequency and all the bearings showed an increase in natural frequency with increasing cyclic load frequency. The dynamic natural frequency of the pads ranged from 23 to 31 percent above the static natural frequency for the natural rubber bearings while the neoprene bearing natural frequency increased by 50 percent at the 12 Hz loading. Most specifications call for a dynamic/static frequency ratio of no more than 1.4. Of the tested samples, all the natural rubber isolators meet this criterion while the neoprene isolator stiffened excessively under dynamic loading.
Comparison with the theoretical values calculated as described above is useful for verifying the analysis method. The calculated static stiffness of the bearings, for 70 durometer rubber, is 29.6 kN/mm with a resulting natural frequency of 8.13 Hz. The calculation is sensitive to the elastomer durometer, with an increase in durometer to 71 leading to a stiffness of 33.1 kN/mm and a natural frequency of 8.61 Hz. It is clear from the results that a theoretical analysis can accurately estimate the vibration isolation characteristics of the bearings. However, care must be taken in the use of the results, since the typical specified tolerance on durometer of ± 5 can have a significant effect on the isolation efficiency at certain frequencies.
