Performance of a large-scale magnetorheological elastomer–based vibration isolator for highway bridges

Abstract

This study presents an experimental investigation on the magnetorheological effect of a new magnetorheological elastomer–based adaptive bridge isolation bearing system. Two identical magnetorheological elastomer–based adaptive bridge bearings (isolators) were designed and fabricated. Electromagnets were incorporated to create a closed-loop magnetic path in the magnetorheological elastomer layers. A double-lap shear and compression test setup was utilized to characterize the mechanical properties of the system subjected to scaled structural cyclic forces and strains. Experimental results demonstrated that the effective stiffness of adaptive bridge bearings increases with increased applied magnetic field and a compressive force resulted in larger apparent shear stiffness. Also, increasing loading frequency resulted in larger apparent shear stiffness and lower magnetorheological effect and similarly, however, a compressive force resulted in smaller magnetorheological effects.

Keywords

Magnetorheological elastomer adaptive bridge bearing isolator vibration control magnetorheological effect

Introduction

Various mechanical systems may be susceptible to high-frequency vibration, which can result in metal fatigue. Similarly, vibration-sensitive equipment such as medical and electronic equipment may malfunction due to short duration as well as sustained vibration, even low intensities. Furthermore, structural systems may be subjected to adverse ambient conditions due to wind, traffic, and rare events such as earthquakes, which impose various levels of deformation and acceleration demands. Vibration isolation is recognized as an effective means to mitigate shock- and vibration-induced damage in mechanical and structural systems. Over the last 80 years, the concept of isolation has evolved and a large amount of research culminated in categorically three sets of strategies and devices/systems. These systems may be passive, semi-active, or active; while passive devices are well established and most commonly used in various applications, research on the latter two continues to develop alternatives that are more versatile than their passive counterparts. Passive devices can be designed to accommodate various vibration-induced translational and rotational deformation demands. However, they cannot be tuned in real time for different loading and operating conditions. Semi-active vibration isolation systems may overcome such limitations since their apparent properties can be tuned while they consume significantly less power than the active systems.

There are a variety of materials and technologies suggested for semi-active vibration isolation, such as magnetorheological elastomers (MREs). MREs are composed of micron-sized magnetically permeable particles and elastomeric medium, and can have other additives such as carbon black and carbon nanofibers. They exhibit a magnetoviscoelastic behavior and, upon application of an external magnetic field, the force between the particles increases which results in an increased stiffness and damping. However, damping change is typically smaller compared to stiffness change (Behrooz et al., 2016). When MREs are subjected to a variable magnetic field, changes in stiffness can be realized in relation to the field intensity. In general, material tests conducted on MREs with various compositions showed that the moduli of MRE samples can be significantly increased by applying a magnetic field (Gong et al., 2007) and the optimum particle concentration for increasing the shear modulus is approximately 27% by volume (Davis, 1999). In addition to the applied magnetic field, the magnetorheological (MR) effect (percentage of increase in stiffness due to the applied magnetic field) is sensitive to the loading conditions such as loading frequency, shear strain level, and applied compressive force. For example, simultaneous application of compressive force can result in reduced (around 50%) MR effect in shear mode (Dong et al., 2012). Furthermore, Stepanov et al. (2007) showed that the MR effect is higher when the samples were subjected to small strains. Nonetheless, the experimentally observed fast response times (7 ms under impact loading (Fu et al., 2013) and 60 ms under large strains (Nguyen and Ramanujan, 2010)) make the MREs potential candidates for use in vibration isolation applications over a wide range of frequencies. MRE vibration isolators have been studied in vehicles’ engine mounts (Ginder et al., 2000), vehicle transmission system (Hoang et al., 2010), vehicle seat suspension (Du et al., 2011), and adaptive tuned vibration absorbers (Deng et al., 2006; Deng and Gong, 2008; Liu et al., 2017). Recently, MREs have been proposed for use in structural vibration isolation applications. Li et al. (2013) demonstrated the design of an MRE-based isolator by replacing traditional rubber layers in a bearing with MRE layers and achieved a 30% increase in effective stiffness. Behrooz et al. (2013a) presented a variable stiffness and damping isolator for seismic base isolation of buildings and modeled its behavior using a phenomenological model with springs, viscous dampers, and a hysteretic Bouc–Wen element. Behrooz et al. (2013b) used the current-dependent relations of the MRE base isolator with a control algorithm to control the performance of the isolator and reduce vibrations in a scaled model of a building. A design methodology was presented for an MRE base isolator by Yang et al. (2014, 2015) that incorporates a permanent magnet and can achieve both positive and negative stiffness changes based on the direction of the applied magnetic field. Tu et al. (2014) investigated the field-dependent shear modulus of an MRE isolator and concluded that the shear modulus increases 40% before reaching magnetic saturation. Xing et al. (2015) designed and tested the performance of a bridge bearing using MRE layers and implemented a field-dependent model to consider the dependency of the stiffness and damping of the isolator on the applied magnetic field.

Both isotropic and anisotropic MREs were studied experimentally and analytically for different applications and loading conditions to evaluate their feasibility. These studies concluded that the mechanical properties of MREs are influenced by volume of iron particles, curing conditions, and magnetic field intensity (Ginder et al., 1999; Jolly et al., 1996; Jung et al., 2009). Chen et al. (2016) theoretically and experimentally determined that the optimal design of an MRE isolator depends upon the particle volume fraction distribution. The reported maximum and minimum MR effects in the past research varied between 1,629% (Li and Li, 2014) and 5% (Yarra et al., 2017a) which depends on magnetic field intensity, thickness of sample, iron particle shape and size, and viscosity of matrix and additives. Zhao et al. (2017) designed a miniature MRE isolator for lateral vibration suppression of bridge monitoring equipment and showed a maximum increase of 114% in effective stiffness through experimental results when current was increased from 0 to 3 A. Wahab et al. (2016) developed a natural rubber (1 mm thick)-based MRE isolator, tested to evaluate compression force, and concluded an increase of 14.5% during static and an increase of 7% during dynamic experiments (5 Hz). MREs with carbon black content showed enhanced off-state shear modulus (Chen et al., 2008; Yarra et al., 2017b). However, studies related to MRE use in civil engineering applications, such as bridges and buildings, with combined shear and compression loads calculated by following standard codes were slim to none.

This study presents an experimental investigation on a new MRE-based adaptive isolation bearing particularly for highway bridges. The prototype geometry and properties of the bearings were selected based on a comprehensive analytical investigation of typical steel-reinforced elastomeric bearings with comparable passive characteristics. A geometric scale factor of 4 was selected based on the test setup and equipment limitations. Two bearings which were designed and fabricated for this study incorporated eight electromagnets per bearing. Electromagnets were designed to generate a closed-loop magnetic field and optimized to produce an average magnetic field of 1.3 T through four MRE stacks. Each stack consisted of alternating layers of MRE material and steel shims. TAP platinum silicone (Side A base and Side B catalyst with 1:1 ratio) and carbonyl iron particles were used to fabricate MRE layers. Electromagnets, steel shims, sole plate, and masonry plate were fabricated using steel 1018 due to its high magnetic permeability. A specially designed double-lap shear test setup was used to evaluate the performance of the bearings under different applied current, shear strain levels, loading frequencies, and compression loads.

System requirements for the design of bearings

To establish realistic force and deformation demands on the MRE-based bearings, steel-reinforced elastomeric bearings with similar passive characteristics were considered first. The baseline passive (off-state) properties were determined after a comprehensive analytical evaluation of 24 steel plate girder benchmark bridge configurations. The bridges vary in span length, bridge width, and girder spacing as shown in Figure 1. A complete statistical analysis of geometric properties of steel girder bridges across the United States was presented by AmiriHormozaki (2013). It was found that most of the common steel plate girder bridges have one of the following geometric configurations including three different average span lengths (80, 130, and 170 ft), four different bridge widths (34, 58, 82, and 106 ft), which represent two- to eight-lane bridges, and two different girder spacing (10.5 and 12 ft). Each geometric configuration of reinforced elastomeric bearings was analyzed for three different shear moduli (100, 175, and 250 lbf/in²) to quantify the effect of stiffness changes on the various response quantities.

The steel-reinforced elastomeric bearings for all the benchmark bridges were designed in accordance with American Association of State Highway and Transportation Officials (AASHTO) LRFD bridge design specifications (AASHTO, 2012). Based on the service limit criteria in AASHTO provisions (AASHTO, 2012), permanent and transient loads are considered. For this purpose, the analytical models of the bridges were developed using CSiBridge (2015) software. The software has the capability of analyzing bridges under different types of loading.

The diameter of the bearing and elastomer height were considered as the main design parameters. These two parameters are selected in such a way that the bearing design would be applicable for three different shear moduli in each benchmark bridge. The total rubber thickness is determined to ensure that shear strain due to service load conditions remains under 50% and hence it is controlled by the minimum anticipated shear modulus (softer bearing will result in smaller shear stiffness and hence larger deformations). On the other hand, the surface area of the rubber was determined to ensure the sufficiently high axial stiffness to limit the axial strain below 4.5% under gravity loading as well as the smallest possible diameter as per AASHTO LRFD; therefore, it was governed by the largest shear modulus.

Finally, the prototype bearing diameter (24 in) and elastomer height (5 in) with an average shear modulus of elasticity, G, of 175 lbf/in² were established based on the deformation and forced demands on the bearings. These prototype dimensions, with a geometric scale factor of 4, were adopted in view of the test setup and equipment limitations as well as other requirements, for example, electromagnetic field design. The test protocols for the component tests were determined to satisfy the minimum requirements of AASHTO (2012). Accordingly, the maximum anticipated forces were 23 (in compression) and 5 kip (in shear).

Design and fabrication of the bearings

The bearings were designed such that they resemble traditional passive isolation bearings commonly used in highway bridges. Overall dimensions, rubber area, shear stiffness, and deformation demand were determined from an extensive analytical study of different bridges around the United States. For the analytical study, AASHTO M251 (AASHTO, 2016) which is referenced in Section 14 of AASHTO LRFD bridge design specs (AASHTO, 2012) and Section 18 of AASHTO LRFD bridge construction specs (AAS HTO, 2015) were considered as the basis for the laminated bridge bearing design and properties. A geometric scale factor of 4 was adopted for the bearings that were experimentally evaluated in this study. The MRE-based bearings feature four stacks of alternating MRE and steel shim layers, eight coils, a sole plate, and a masonry plate. Each stack consisted of 3-in-diameter of 10 1/8-in-thick MRE and 9 1/8-in steel shim layers. Figure 2 shows the overall dimensions and geometry.

Fabrication of MREs and steel shims and preparation of stacks

The MRE material consisted of 35.3 vol% (80 wt%) spherical-shaped iron particles (Grade-R-2410; provided by ISP Technologies Inc., New Jersey, USA), 3.7 vol% (2 wt%) SR303 carbon black (provided by Sid Richardson Carbon Company, Texas, USA), and 30.5 vol% (9 wt%) base A and 30.5 vol% (9 wt%) base B of Taps Platinum Silicone. The average diameter ranging from 5 to 8 µm of iron particles was used for the MR effect. SR303 was used as an additive to improve the base passive stiffness of the MRE material. The mixture was prepared using the above composition and poured into a pre-fabricated mold designed to produce 1/8-in-thick and 3-in-diameter MRE; then the mold was placed into a vacuum chamber and suctioned to −70 kPa to remove air bubbles. Next, the mold was placed in a vise and squeezed until the desired thickness of MRE was produced. Finally, the mold was placed into the electromagnets consisting of two coils wrapped with a superior Essex 19 Heavy AWG magnet wire, with each coil ranging from 13 to 17Ω and capable of producing 1.2 T of magnetic field and cured for 4 h at room temperature. Two power supplies (Sorensen SGA 400-25) were used to apply a constant 5 A current which generates the desired magnetic field. Figure 3 shows the electromagnetic system used for the fabrication of MRE layers.

Steel shims (1/8 in thickness and 3 in diameter) were made from high-magnetic-permeability steel 1018. Steel shims were sand blasted for better bonding. Extensive testing was performed to determine the bonding strength of three commonly used adhesives used for bonding silicone and rubber materials to steel. The three adhesives were Loctite 380 Instant Adhesive, Titebond Construction Adhesive, and Dow Corning 832 RTV Sealant. Finally, MRE and steel layers were connected using an adhesive (Dow Corning 832 Multi-Surface Adhesive Sealant), which also included iron particles with a 1:1 ratio to ensure better magnetic conductivity through the adhesive.

Fabrication of electromagnets for bearing

The coils used to induce the magnetic field in MREs of the bearing were made from a 2.25-in-long and 6-in-diameter 1018 steel rod. Computer numeric control (CNC) milling was used for precision and to fabricate 3-in-diameter and 1.5-in-long electromagnetic cores with 6-in-diameter and 1/4-in-thick plates at the top and bottom. The fabricated coils are shown in Figure 4. Coils were double insulated with non-magnet conductive CP high-temperature red varnish insulation spray and Teflon liners as they were wound using a Superior Essex 15 Heavy AWG Ultra Shield Plus magnet wire to achieve the desired magnetic field. Each electromagnet was built to have the same number of windings which was 515, subsequent to the fabrication and winding of each electromagnet, and the measured resistance of each electromagnet was 2.5Ω. Therefore, all electromagnets were thus considered as identical bearings. The coils were designed to carry 19.5 A. The detailing and the material properties of the electromagnets result in capacities of several folds larger than the capacity of the elastomer stacks that ensures safe load path from the top and bottom plates to the MRE stacks. An image of one of the fabricated bearings is shown in Figure 5.

To ensure sufficient magnetic field for the activation of the MRE layers, three-dimensional magnetic field analyses were performed by modeling the bearings shown in Figure 7 using ANSOFT/Maxwell3D finite element analysis (FEA) software package (ANSYS Maxwell Computer Software, 2016). A direct current (DC) of 10,000 amp-turns was applied through insulated copper wire gage 15 with a resistance of 2.5Ω. Figure 6 shows the analytically predicted magnetic field distribution along the height of each stack (total of eight stacks for two bearings) with an average of 1.3 T. The alternating magnetic field is due to higher magnetic permeability in the steel shims than MREs. Similarly, the maximum magnetic field of over 1.6 T is achieved in the steel top and bottom load-bearing plates. However, after the designed test matrix was completed, the middle and outermost MRE layers were cut with a blade, size as probe of a gaussmeter and measured the magnetic field at the same applied amperage for experimental tests. This procedure was not an ideal as a gap between the probe and the hole might be filled with some air, thus reducing the magnetic field and the experimentally measured magnetic fields at 1/16 and 1–1/16 in were 0.5 and 0.7 T, respectively. The difference between the analytically predicted and experimentally measured field intensities was attributed to less than ideal physical bonding between the elastomer and steel shim layers which the analytical model does not take into account. In addition, the magnetic permeability of the MRE material used in the analytical models was estimated based on the previously published research (Kashima et al., 2012); hence, it was deemed as one of the factors that contributed to this discrepancy. In conclusion, reduction in the experimentally measured magnetic field was due to magnetic permeability, saturation limits of materials used, and the method used to measure the magnetic field.

Experimental evaluation of the scaled bridge bearings

Double-lap shear and compression experimental setup

The component tests were conducted at the Large-Scale Structures and Earthquake Engineering Laboratory at the University of Nevada, Reno. The main objective of the component tests was to characterize the mechanical properties of the bearings such as axial (compressive) and shear stiffness as well as energy dissipation characteristics under various magnetic field intensities. A double-lap shear and compression test setup was designed and fabricated for this purpose as shown in Figure 7. The setup was used for the application and control of shear and axial force and strains on the bearing system. The shear strain was applied and controlled using a hydraulic actuator (type MTS 244.22) and axial force was applied and controlled by a hydraulic cylinder (manufactured by William S Pine). Hydraulic actuator (MTS 244.22) has a stroke length of 20 in and a force rating of 22 kip in compression and in tension. Hydraulic cylinder has a stroke length of 12 in and a force rating of 360 kip. Two power supplies (Sorensen SGA 400-25) were used to apply varying amperage (0, 7, 14, and 19.5 A) which generates the desired magnetic field. As shown in the Figure 7, aluminum plates were used to prevent magnetic field passing through the entire test setup and to magnetically isolate the bearings, thereby preserving the high levels of the magnetic field in the MRE layers.

Instrumentation and data processing

A total of eight displacement transducers (four per bearing; Novotechnik TR75) were installed to measure axial deformation and potential rotation of the bearings when they were subjected to combined shear and compression loading. Three string pots (UniMeasure PA40) were used to capture the out-of-plane deformations that may result due to imperfections in the test setup. In addition, two thermocouples were installed to measure the temperature of the MRE layers before and after each test. Finally, one accelerometer was used to measure absolute acceleration at the actuator head. The accelerometer was a three-axis Analog Devices Model ADXL326 MEM with the nominal measurement ranges of ±16 g. Each sensor was connected to National Instrument data acquisition system, conditioned, low-pass filtered with a cut-off frequency of 80 Hz, and sampled at 48 Hz.

After each test, preliminary inspection and processing of the recorded data was performed to ensure fidelity. The measured acceleration response was used to apply the necessary correction to the measured bearing forces due to induced inertial forces, particularly during the high-frequency tests. Further post-processing of effective stiffness and damping were carried out using a custom-developed software in MATLAB (2014).

Experimental results and discussion

The experimental program was developed to apply six constant strain cycles and the following test parameters were considered: (1) shear strains (5%, 10%, 15%, 20%, and 50%); (2) axial load (0, 2.5, 5, and 10 kip with the corresponding normal stress on each stack as 0, 88, 177, and 353 lb/in²); (3) sinusoidal loading frequency (0.1, 0.5, 1, and 4 Hz), and (4) magnetic field intensities that correspond to 0, 7, 14, and 19.5 A of applied current to the electromagnets. Effective stiffness was calculated using the procedure recommended by US codes ASCE 7-16 and AASHTO (2015). The effective shear stiffness is defined as shown in Figure 8 and determined using equation (1) (American Society of Civil Engineers (ASCE), 2016) as follows

k_{eff} = \frac{| F^{+} | + | F^{-} |}{| Δ^{+} | + | Δ^{-} |}

(1)

where $F^{+}$ and $F^{-}$ are the positive and negative forces, at $Δ^{+}$ and $Δ^{-}$ , respectively. The component tests were used to evaluate the effect of various magnetic field intensities on the effective stiffness calculated from the force–displacement curves. The MR effect is defined as the change in the properties of MRE under the magnetic field and quantified as the ratio of absolute change in effective stiffness due to a given magnetic field intensity to zero-field effective stiffness (k_o,_eff) and determined using equation (2)

% M R e f f e c t = \frac{k_{M R, e f f} - k_{0, e f f}}{k_{0, e f f}}

(2)

Response under shear-only loading

Experiments were carried out for different strains, frequencies, and magnetic fields using a hydraulic actuator and by setting the hydraulic cylinder at zero position as a pre-load condition. The percent change in effective shear stiffness is determined relative to the 0.1-Hz tests as summarized in Table 1. For a given strain level with varying frequencies, the increase in effective shear stiffness varies between 13% and 66%. While the increase is insignificant for frequencies up to 1 Hz, up to 66% increase in effective stiffness was observed corresponding to 4 Hz and 20% shear strain. The effective shear stiffness values are summarized in Table 1. Figure 9 shows a sample shear force–displacement hysteresis response. The overall behavior of MRE follows the stress–strain relationship of rubber-like materials which show a higher initial effective stiffness. Also, due to the viscoelastic behavior of MREs, a higher stiffness is observed at higher loading frequencies.

Table 1.

Effect of loading frequency on the effective shear stiffness.

Loading frequency (Hz)	Effective stiffness (kip/in)
	Strain (%)
	5	10	20
0.1	4.10	3.70	3.47
0.5	4.13	3.67	3.52
1.0	4.20	3.68	3.57
4.0	4.62	4.99	5.76
Frequency effect (%)	13	35	66

Figure 10 shows an example of shear force–deformation response for a loading frequency of 0.1 Hz and an applied strain of 50% with different magnetic field intensities corresponding to 0 and 19.5 A. For a given test frequency, an increase in effective stiffness due to increasing magnetic field was observed. For example, Figure 11(a) shows that for the quasi-static tests (0.1 Hz) at 20% shear strain and 19.5 A input current, the increase in stiffness is approximately 12%. However, for a given magnetic field intensity, the increase in effective stiffness reduces with increasing strain amplitude. In summary, the percent increase in effective stiffness reduces from 18% at 5% strain to 5% at 50% strain since a larger force is applied at higher strains and the ratio of the generated magnetic force to mechanical force decreases. It can also be seen from Figure 11 (comparison between Figure 11(a) and 11(b)) that the MR effect is reduced with increasing frequencies. This is due to higher forces being applied at higher frequencies and a reduced ratio of magnetic force to mechanical force. Figure 11(b) shows that the maximum increase in effective shear stiffness is approximately 14% at 5% strain for a test frequency of 4 Hz. The effective shear stiffness values with zero axial force and varying frequencies and strains are summarized in Table 2.

Table 2.

Effective stiffness (kip/in) under various loading conditions with different axial loadings.

Applied current and MR effect	5% strain		10% strain		20% strain		50% strain
Applied current and MR effect	0.1 (Hz)	4 (Hz)	0.1 (Hz)	4 (Hz)	0.1 (Hz)	4 (Hz)	0.1 (Hz)
Axial force = 0 kip
0.0 A	3.91	4.72	3.46	4.20	3.35	3.96	2.90
19.5 A	4.61	5.36	4.03	4.72	3.76	4.40	3.05
MR effect (%)	18	14	17	12	12	11	5
Axial force = 5 kip
0.0 A	5.04	6.15	4.86	5.60	4.41	5.20	NA
19.5 A	6.19	7.40	5.54	6.36	4.80	5.77	NA
MR effect (%)	23	20	14	14	9	11	NA
Axial force = 10 kip
0.0 A	8.06	8.41	7.42	7.96	5.90	7.00	3.59
19.5 A	9.58	10.27	8.00	8.69	6.13	7.51	3.93
MR effect (%)	19	22	8	9	4	7	9

MR: magnetorheological; NA: not applicable.

Combined shear and compression loading

The test setup was used to apply simultaneous shear strains and axial forces. A hydraulic cylinder was used to hold a constant axial force and different strains, frequencies, and magnetic fields were applied using a hydraulic actuator. As shown in Figure 12(a) and (b), for a given strain amplitude, the effective shear stiffness of the bearing increases with increasing compression force. It can also be observed that the effective shear stiffness increases for larger magnetic field intensities for any given compression force and strain. Table 2 summarizes the effect of axial force on the shear force–deformation properties of the bearing. Figure 13(a) shows that under quasi-static shear deformation and an axial force of 5 kip, effective shear stiffness enhancement due to magnetic field is 23% at low strain level (5%), but reduces to 9% at a higher strain amplitude of 20%. The corresponding enhancement in the effective stiffness with 10.0 kip of axial force decreases from 22% to 4% when the strain level increases from 5% to 20%. Therefore, it is reaffirmed that larger strain amplitudes usually result in lower MR effect due to a smaller ratio of applied magnetic to mechanical force, as presented in Table 2. Also, as shown in Figure 13, the effect of magnetic field on the effective shear stiffness decreases with increasing axial load. This may be due to the reduced distance between the iron particles, and thus an increase in the base passive shear stiffness values. At higher frequencies, the bearings exhibit a higher stiffness due to viscoelastic properties of the elastomer matrix (Table 2).

Compressive loading

As a pre-load condition, a hydraulic actuator was used to hold the bearing system at zero shear strain position and the magnetic field effect was studied at different axial forces applied using the hydraulic cylinder. These experiments were conducted to investigate the adaptability of the bearings while supporting large structural loads based on AASHTO axial strain limits. The maximum compression force was determined based on AASHTO (2012) as 23 kip, which results in 115 lbf/in². Bearings were tested under different axial forces with varying magnetic field and loading frequencies. The axial force–displacement plots shown in Figure 14(a) and (b) demonstrate a smaller increase in axial stiffness compared to shear stiffness due to increasing magnetic field. Table 3 shows the axial stiffness increment with increased magnetic field. The stiffness increment is 12% for 10 kip axial force and 4% under 20 kip axial load. Therefore, similar to shear stiffness, the relative effect of magnetic field on the stiffness decreases with increasing axial load.

Table 3.

Pure compression test results.

Axial force(kip)	Current (A)	Load rating(lb/sec)	Thickness ofMRE (in)	Axial displacement (in)	Axialstrain (%)	Axial stiffness(lb/in)
10.00	0.00	50.00	1.25	0.056	4.48	178.57
10.00	19.50	50.00	1.25	0.050	4.00	200.00
20.00	0.00	200.00	1.25	0.105	8.60	190.48
20.00	19.50	200.00	1.25	0.101	8.09	198.02

MRE: magnetorheological elastomer.

Temperature

During the experimental testing, the temperature of bearing electromagnets was measured and it varied from 23°C at 0 A to 27°C at 19.5 A.

Mode of failure

Finally, the failure of the adaptive bearings was investigated by subjecting the bearings to 100% shear strain cyclic loading. Figure 15 shows the force–displacement response of the adaptive bearing under the constant strain of 100% and the frequency of 0.1 Hz with zero magnetic field. As shown in Figure 15, failure was observed approximately at 74% strain amplitude. Further investigation of the MRE layers revealed that the failure was due to the bond failure between MRE and steel surfaces, as shown in Figure 16.

Conclusion

A large-scale adaptive MRE bridge bearing (vibration isolator) was designed and manufactured to investigate the effect of structural loads on its MR effect. The effects of magnetic field, strain level, and loading frequency on the effective stiffness of the bearing were studied. The observed MR effect under combined shear and compression loading ranged between 23% and 9% at 5 kip axial load and strains of 5% and 20%, respectively. However, increasing the compression force further to 10 kip reduced the MR effect relative to 5 kip. Effective stiffness increased with increasing frequency for a given strain amplitude. In the absence of compressive force, the maximum and minimum changes in effective shear stiffness (due to increased frequency) were between 13% and 66% for strains between 5% and 20%, respectively. Under compression-only loading, the MR effect was reduced from 12% to 4% when the axial load was increased from 10 to 20 kip. Larger axial force resulted in larger apparent passive stiffness which increased by 7% when the axial force was increased from 10 to 20 kip. The computed damping was insignificant. Adhesive failure was observed between MRE and steel surface at 74% shear strain.

Footnotes

Acknowledgements

The authors would like to express their gratitude to Dynamic Isolation Systems, Inc., and to Scougal Rubber Corporation. They also acknowledge Mr Tony Berendsen, Development Technician of the Department of Mechanical Engineering, Mr Chad Lyttle, Development Technician of the Center for Civil Engineering Earthquake Research, University of Nevada, Reno, and Dr Patrick Laplace, Research Associate Professor and Manager of the Large-Scale Structures and Earthquake Engineering Laboratory, for their assistance and guidance. They also wish to thank Mr Troy Martin, Structures Division of the Nevada Department of Transportation (NDOT), for his valuable feedback and participation during the component testing of the adaptive bearings. Finally, the authors would like to thank undergraduate assistants D Mar, N Pinuelas, and B Muznich for their assistance.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project was funded by the Federal Highway Administration under the contract number DTFH61-13-C-00020.

ORCID iD

Faramarz Gordaninejad

References

American Association of State Highway and Transportation Officials (AASHTO) (2012) Bridge design specifications, customary U.S. units, with 2012 and 2013 interim revisions. Available at: https://app.knovel.com/hotlink/toc/id:kpAASHTO32/aashto-lrfd-bridge-design/aashto-lrfd-bridge-design

American Association of State Highway and Transportation Officials (AASHTO) (2015) Bridge construction specifications with 2016 interim revisions. Available at: https://app.knovel.com/hotlink/toc/id:kpAASHTO4A/aashto-lrfd-bridge-construction/aashto-lrfd-bridge-construction

American Association of State Highway and Transportation Officials (AASHTO) (2016) Standard specifications for transportation materials and methods of sampling and testing. Available at: https://app.knovel.com/web/toc.v/cid:kpSSTMMS65/viewerType:toc//root_slug:standard-specifications/url_slug:standard-specifications

AmiriHormozaki

(2013) Analytical fragility curves for horizontally curved steel girder highway bridges. Doctoral Dissertation, University of Nevada, Reno, Reno, NV.

ANSYS Maxwell Computer Software (version 11.1) (2016) ANSOFT/Maxwell3D FEA software package. Available at: http://www.ansys.com/products/electronics/ansys-maxwell (accessed 25 November 2016).

American Society of Civil Engineers (ASCE) (2016) Minimum Design Loads for Buildings and Other Structures. Reston, VA: ASCE.

Behrooz

Wang

Gordaninejad

(2013a) Modeling of a new magnetorheological elastomer-based isolator. In: Proceedings of SPIE, San Diego, CA, 10–14 March, vol. 8688, p. 86880Z. Bellingham, WA: SPIE.

Behrooz

Wang

Gordaninejad

(2013b) Seismic control of base isolated structures using novel magnetorheological elastomeric bearings. In: Proceedings of ASME, Snowbird, UT, 16–18 September, p. V001T03A027. New York: ASME.

Behrooza

Yarra

Marc

et al . (2016) A self-sensing magnetorheological elastomer-based adaptive bridge bearing with a wireless data monitoring system. In: Proceedings of SPIE, Las Vegas, NV, 20–24 March, vol. 9803, p. 98030D-1. Bellingham, WA: SPIE.

10.

Chen

Gong

(2008) Effect of carbon black on the mechanical performances of magnetorheological elastomers. Polymer Testing 27(3): 340–345.

11.

Chen

Wang

Zhang

et al . (2016) Optimal design of laminated-MRE bearings with multi-scale model. Smart Materials and Structures 25(10): 105037.

12.

Computers and Structures, Inc. (2015) CSiBridge (computer software version 15). Available at: https://www.csiamerica.com/products/csibridge (accessed 25 May 2014).

13.

Davis

(1999) Model of magnetorheological elastomers. Journal of Applied Physics 85(6): 3348–3351.

14.

Deng

Gong

(2008) Application of magnetorheological elastomer to vibration absorber. Communications in Nonlinear Science and Numerical Simulation 13(9): 1938–1947.

15.

Deng

Gong

Wang

(2006) Development of an adaptive tuned vibration absorber with magnetorheological elastomer. Smart Materials and Structures 15(5): N111.

16.

Dong

et al . (2012) The pressure-dependent MR effect of magnetorheological elastomers. Smart Materials and Structures 21(7): 075014.

17.

Zhang

(2011) Semi-active variable stiffness vibration control of vehicle seat suspension using an MR elastomer isolator. Smart Materials and Structures 20(10): 105003.

18.

Zheng

et al . (2013) A new magnetorheological elastomer isolator in shear—compression mixed mode. In: 2013 IEEE/ASME international conference on advanced intelligent mechatronics (AIM), Wollongong, NSW, Australia, 9–12 July, pp. 1702–1706. New York: IEEE.

19.

Ginder

Nichols

Elie

et al . (1999) Magnetorheological elastomers: properties and applications. In: Smart structures and materials, Newport Beach, CA, 1 March, vol. 3675, pp. 131–139. Bellingham, WA: SPIE.

20.

Ginder

Nichols

Elie

et al . (2000) Controllable-stiffness components based on magnetorheological elastomers. In: Proceedings of SPIE—The International Society for Optical Engineering, Newport Beach, CA, 6–9 March 2000, pp. 418–425. Bellingham, WA: SPIE.

21.

Gong

Chen

(2007) Study of utilizable magnetorheological elastomers. International Journal of Modern Physics B 21(28–29): 4875–4882.

22.

Hoang

Zhang

(2010) An adaptive tunable vibration absorber using a new magnetorheological elastomer for vehicular powertrain transient vibration reduction. Smart Materials and Structures 20(1): 015019.

23.

Jolly

Carlson

Muñoz

et al . (1996) The magneto viscoelastic response of elastomer composites consisting of ferrous particles embedded in a polymer matrix. Journal of Intelligent Material Systems and Structures 7(6): 613–622.

24.

Jung

Lee

Jang

et al . (2009) Dynamic characterization of magneto-rheological elastomers in shear mode. IEEE Transactions on Magnetics 45(10): 3930–3933.

25.

Kashima

Miyasaka

Hirata

(2012) Novel soft actuator using magnetorheological elastomer. IEEE Transactions on Magnetics 48(4): 1649–1652.

26.

(2014) Base isolator with variable stiffness and damping: design, experimental testing and modelling. In: 23rd Australasian conference on the mechanics of structures and materials (ACMSM23) (ed. ST

Smith

), Byron Bay, NSW, Australia, 9–12 December, vol. II, pp. 913–918. Lismore, NSW, Australia: Southern Cross University.

27.

et al . (2013) Development and characterization of a magnetorheological elastomer based adaptive seismic isolator. Smart Materials and Structures 22(3): 035005.

28.

Liu

Zou

(2017) Development of a semi-active dynamic vibration absorber for longitudinal vibration of propulsion shaft system based on magnetorheological elastomer. Smart Materials and Structures 26: 075009.

29.

The MathWorks Inc . (2014) MATLAB R2014a (version 8.3). Natick, MA: The MathWorks Inc.

30.

Nguyen

Ramanujan

(2010) Novel coiling behavior in magnet-polymer composites. Macromolecular Chemistry and Physics 211(6): 618–626.

31.

Stepanov

Abramchuk

Grishin

et al . (2007) Effect of a homogeneous magnetic field on the viscoelastic behavior of magnetic elastomers. Polymer 48(2): 488–495.

32.

Mei

et al . (2014) Research on new type MRE isolator and its mechanical model. Materials Research Innovations 18(Suppl. 2): S2-552–S2-558.

33.

Wahab

NAA

Mazlan

Kamaruddin

et al . (2016) Fabrication and investigation on field-dependent properties of natural rubber based magneto-rheological elastomer isolator. Smart Materials and Structures 25(10): 107002.

34.

Xing

et al . (2015) A laminated magnetorheological elastomer bearing prototype for seismic mitigation of bridge superstructures. Journal of Intelligent Material Systems and Structures 26(14): 1818–1825.

35.

Yang

et al . (2015) A new magnetorheological elastomer isolator in shear–compression mixed mode. Journal of Intelligent Material Systems and Structures 26(10): 1290–1300.

36.

Yang

Sun

et al . (2014) A novel magnetorheological elastomer isolator with negative changing stiffness for vibration reduction. Smart Materials and Structures 23(10): 105023.

37.

Yarra

Behrooz

Pekcan

et al . (2017a) A large-scale adaptive magnetorheological elastomer-based bridge bearing. In: SPIE smart structures and materials + nondestructive evaluation and health monitoring, Portland, OR, 25–29 March, p. 1016425. Bellingham, WA: SPIE.

38.

Yarra

Pekcan

Behrooz

(2017b) Characterization of carbon black-filled natural rubber and silicone magnetorheological elastomers under pure shear loading. In: 8th ECCOMAS thematic conference on smart structures and materials, Madrid, 6 June 2017, pp. 860–870. Barcelona, Spain: International Center for Numerical Methods in Engineering (CIMNE).

39.

Zhao

et al . (2017) A miniature MRE isolator for lateral vibration suppression of bridge monitoring equipment: design and verification. Smart Materials and Structures 26(4): 047001.