To Investigate Different Parameters of Economic Sliding Based Seismic Isolation System

ABSTRACT Masonry structures are frequently constructed in developing countries because of their affordability and ease of construction, even though they are vulnerable and perform poorly during earthquakes. Experimental and numerical studies were carried out to determine the optimum thickness of the isolation layer for a recently developed, low-cost seismic isolation system known as the Reinforced-Cut-Wall (RCW). The parameters of the isolation layer were optimized during the experimental and numerical investigations using hollow concrete blocks. Quasi-static cyclic tests were performed on a 1:3 reduced-scale brick masonry wall incorporating the RCW isolation scheme, validating the better performance of the proposed isolation technique.


Introduction
An earthquake of magnitude M7.6 hit northern Pakistan and Azad Jammu and Kashmir (AJK) on 8 th October 2005 at local time 8:50 am.The epicenter of this earthquake was about 9 km away from Muzaffarabad city (Capital of AJK) in the northeast direction (Arya 2000;Bayraktar, Coşkun, and Yalçin 2007;Bilham, Gaur, and Molnar 2001;Durrani et al. 2005;ERRA 2006).According to estimates, the number of damaged or completely destroyed buildings was above 780,000.The majority of these damaged buildings were declared unusable for living in the future.Most of the completely destroyed buildings were non-engineered unreinforced masonry (URM) buildings, which included approximately 170,000 schools and major hospitals.The most common mode of construction in that region was either single-or double-story masonry buildings consisting of reinforced concrete floors and unreinforced solid concrete block, solid brick, or stone masonry-bearing walls (ERRA 2006).
Surveys have revealed that among all the buildings destroyed in the 2005 earthquake, more than 95% were non-engineered structures lacking aseismic features.Artisans, engineers, and technicians did not follow aseismic construction.The reason may be the negligence of the area as an earthquakeprone region, regardless of a clear history of several earthquakes (Bilham, Gaur, and Molnar 2001).
Masonry buildings are the most commonly used type of construction worldwide, especially in developing countries like Pakistan, India, Iran, Nepal, Turkey, etc.Some important features of masonry like the economy, involvement of least engineering, no requirement of skilled labor, ease of availability of materials, and ease of repair make this building type the most preferred one.Poor seismic resistance is the only problem associated with this building type and is a major limitation for such buildings to be constructed in earthquake-active regions (Arya 2000).According to previous research conducted on past earthquakes, it was found that the only building type that underwent maximum damage was masonry.All masonry structures had very poor seismic performance and resulted in tremendous loss of human lives (Battaglia, Ferreira, and Lourenço 2021;Faiella, Calderoni, and Mele 2022;Qin et al. 2021).Damages caused to masonry buildings by the recent earthquakes in Pakistan have necessitated the development of an effective low-cost seismic isolation system for masonry buildings.Earthquakes are one of the most challenging natural hazards for civil engineers because they bring enormous loss of human lives as well as damage or even collapse of the entire infrastructure.So, in order to deal with such challenges, many techniques have been developed, such as energy-dissipating devices, damping devices (active, semi-active, and passive dampers), and seismic base isolation.The only hurdle in adopting the previously developed energy-dissipating devices is their higher cost, especially in developing countries like Pakistan, India, Nepal, etc.
Despite the fact that masonry structures do not exhibit high seismic performance, they are still the most constructed type of buildings in all developing countries.Experimental testing and analytical investigations are efficiently being conducted all over the world to understand the seismic behavior of masonry buildings and to find an economical solution for seismic hazard mitigation.Masonry structures possess highly complex constitutive properties in terms of structural materials.The geometrical and physical behaviors of masonry are nonlinear during dynamic testing.All these problems associated with masonry structures result in complexities of numerical modeling of their seismic behavior (Bayraktar, Coşkun, and Yalçin 2007).Therefore, there is a dire need to study such base isolation systems that are economical as well as efficient in making masonry structures resistant to earthquakes.A sliding-based isolation system using the reinforced cut wall (RCW) technique could be an efficient solution in contrast to the traditional strengthening techniques proposed previously (Zhang, Ali, and Sun 2021).
The main objective of seismic base isolation is to introduce such components at the base of a building, which is horizontally flexible but vertically stiff.This setup uncouples the super-structure from highfrequency shaking earthquakes.Seismic base isolation increases the natural time period of the structure that causes a reduction in force demand on the structure due to a reduction in acceleration.An efficient isolation system is one that reduces the amount of energy that is transferred to the structure by dissipating it through energy dissipation mechanisms within the system.A seismic isolation system shifts the displacement demand from the superstructure to the isolation system.Another phenomenon associated with seismic isolation is damping which is the reduction in the amplitude of oscillation (Zhang and Ali 2021).
Several seismic base isolation techniques have been explored and are still under consideration.The most widely used passive isolation systems include sliding isolation and elastomeric bearings.The sliding-based isolator, which utilizes surface friction to dissipate seismic energy, has an advantage over the elastomeric bearings as it could be utilized under a variety of excitations having varying frequencies.Among them, the Friction Pendulum system, initially proposed by (Zayas, Low, and Mahin 1990) and explored by many researchers (Barone, Calvi, and Pavese 2019), is one of the popular sliding isolation systems that has already reached the implementation stage in developed countries.
The isolation systems developed previously have some problems associated with them, one such problem faced by most of the previously developed isolators is their life span being less than the service life of the building.Furthermore, the implementation of isolators developed previously is limited to very few structures because of their uneconomic fabrication, design complexities, heavyweight, and difficulties in material availability, and is therefore not the optimum solution for developing countries.Out of all these systems, the one that best suits the masonry mode of construction is the Pure Sliding Friction (P-S-F) base isolation system.This system consists of introducing a sliding layer at the plinth level of masonry buildings.The superstructure rests on the sliding layer and is free to slide away from the frictional resistance.
The optimum friction between the sliding interfaces plays a vital role in the overall performance of the isolator.It was previously observed that small friction between the sliding interface (0.05-0.15) could be obtained by the utilization of Teflon sheets and stainless steel (Constantinou, Mokha, and Reinhorn 1991).However, the utilization of continuous bonding of steel sheets would result in higher costs and heavyweight and complex construction.On the other hand, various researchers have suggested various low-cost alternatives, such as screened sand, stone pebble, graphite, marble, grease, and rubber soil mixture, and these materials are also not suitable for long-term use as the sand can get crushed and spill out of the isolation layer after the seismic shock, the grease could be polluted with dust, graphite, and marble could be affected by environmental effects (Bibi et al. 2020) and usage of rubber soil mixture or stone pebbles as an isolation layer below the foundation could produce rocking phenomena in the building.
Pure sliding friction isolation being much more economical is best suited for developing countries where the majority of buildings are low-cost low-rise masonry structures.Except for the complications in the re-centering mechanism, this type of system is the simplest among all base isolation techniques, and it does not involve any design complexities (Lee 1984).The p-S-F system is based on a simple phenomenon of introducing a sliding interface between the super-structure and foundation, most preferably at the plinth level, both of them move together and behave as an elastic structure until and unless the frictional resistance is not exceeded.At the time of building construction, such an isolation system can be uniformly distributed throughout the base of the structure in the form of a layer at the plinth level (Calvi and Calvi 2018).
In India, earthquakes during the early 1930s made researchers notice a better performance of sliding base structures as compared to corresponding fixed-base structures.The same observation led to the validation of the concept of introducing a sliding layer between the foundation and superstructure, and this exploration was validated by many researchers later on, such as (Arya 1984;Furinghetti et al. 2019;Jalali, Cardone, and Narjabadifam 2011;Li et al. 2021;Mostaghel and Tanbakuchi 1983;Westermo and Udwadia 1983;Younis and Tadjbakhsh 1984).
The seismic isolation of masonry buildings got the attention of researchers in China from 1960 to 1976.Earthquakes during this interval brought colossal damage to small masonry structures.It was observed that those buildings that were constructed on sandy soils and those that had some slippage at their bases performed better than fixed base structures.Such observations pulled attention toward testing and analyzing friction isolation systems (Scacco et al. 2020).According to (Lee 1984), the seismic performance of the buildings can be improved by introducing an artificial horizontal sliding interface at the plinth level.This sliding interface helped to dissipate seismic energy during the event of severe earthquake shaking.The proposed sliding interface was composed of loose sand grains enclosed between terrazzo plates.This setup helped to isolate the structure from the destructive horizontal earthquake ground motions.The efficiency of this isolation layer was validated by installing it in the actual 3D structural model, which was then tested on the shake table.Among other passive isolation systems, the flat sliding isolation system is preferred because of its better performance at multiple seismic intensities (Wei et al. 2021).
The low-cost seismic isolation proposed by (Lee 1984) was investigated only at a small scale, requires two thick slabs above and below the isolation system, and was based on specially screened sand, which necessitates a sieving procedure before its use (Tsiavos et al. 2020(Tsiavos et al. , 2021)), presented a holistic numerical and large-scale experimental campaign that demonstrated the efficiency of the use of a thin sand layer, sandwiched between two PVC surfaces, for the low-cost sliding seismic isolation of structures.This low-cost seismic isolation, defined as a PVC sandwich, utilizes the mechanics of rolling (not pre-screened) sand grains with various diameters for the initiation of sliding and requires only one slab above the isolation layer and a 50 mm thick unreinforced blinding layer below the isolation layer (Tsiavos et al. 2020), quantified experimentally at a large scale the density of sandwiched sand that is required between the two PVC surfaces to optimize the sliding behavior of the seismic isolation system, thus protecting the structure from seismic damage.(Lou, Wang, and Su 1992) proposed the use of graphite as an economical and durable material to be placed under the load-bearing walls for seismic isolation.Graphite is a very stable material under varying sliding cycles and is not affected by humidity (Yuan et al. 2019).Established numerical isolation models using the finite element analysis software ABAQUS, for traditional masonry, sand cushion seismic isolation, and composite isolation.During the dynamic analysis, the composite isolation system was found best among all others because it dissipated maximum seismic energy.(Barbat and Bozzo 1997) discussed the outcomes of different numerical analysis methods used for the seismic isolation of elastic and inelastic structures.(Saitta et al. 2018) studied the effect of the coefficient of friction on the behavior of the isolation interface.According to them, a sufficient value of the coefficient of static friction is very important in maintaining the structural integrity during wind loads or minor earthquakes.While during strong motion incidents, the same static coefficient of friction makes a structure perform normally.
The seismic design specifications and seismic design codes especially emphasize the provision of a proper recentering mechanism while designing an isolation system.(Sassu 2006) implemented a low-cost friction-based isolation system at the base of a multi-story building.The isolator was made up of a 50 mm thick weak mortar layer enclosed between two elastomeric membranes.The isolator was composed of a weak mortar layer so that it could help in energy dissipation through its cracking during strong earthquakes.(Zongjin, Rossow, and Shah 1989) was among the very first researchers to install friction-based seismic isolation systems in actual 3D building models and test them on the shake table.One such full-scale masonry model was tested on a shake table and it performed well during extreme shaking in both experimental and numerical studies.It was found that introducing a sliding interface at the foundation level improved the performance of the structure during strong seismic events.(Tsang 2008;Tsang et al. 2012) proposed the use of locally available materials for seismic isolation.They concluded that if a mixture of rubber and sand is used beneath the foundation, it would not only reduce the vertical acceleration response but would also reduce the horizontal vibrations.The proposed system was considered an aid towards the development of sustainable construction as it incorporates scrap rubber tire aggregates.The concept of using scrap tires as a medium of energy dissipation was experimentally tested by (Xiong and Li 2013).A different percentage of rubber was added to the soil and installed as an isolation medium; however, this change in the percentage of rubber did not show any significant effect on the seismic performance of the isolation system.(Ahmad, Ghani, and Raghib Adil 2009) incorporated recycled mortar instead of cement in constructing a 1:4 scale reduced brick masonry structure.The isolated model contained coarse dry sand as an isolation layer material at the plinth level.The model was tested on a shake table to check the acceleration and displacement responses of fixed and isolated base structures.They concluded that up to 20% of cement by weight could be replaced by recycled mortar in order to give better numerical and analytical results.The isolated model dissipated 70% more seismic energy as compared to the fixed base model.During shake-table testing, the isolated model did not undergo any visible cracking up to a shake-table force of 4.5 kN (1300 rpm).
In order to introduce an economical seismic energy dissipation system for developing countries, various researchers investigated the effectiveness of composite isolation layer materials (Cao and Li 2019;Pitilakis, Karapetrou, and Tsagdi 2015;Sun et al. 2021;Tsiavos et al. 2019).Various other researchers have also worked on introducing economic base isolation systems so far (Tsang and Pitilakis 2019).(Banović, Radnić, and Grgić 2019) used stone pebbles as an isolation layer material in isolating the superstructure from the base and tested the models experimentally to validate the usefulness of the isolation system.They also investigated the effect of repeated excitations on the mechanical behavior of the pebble isolation layer.For low-rise residential houses in the high seismic region of Turkey (Cilsalar andConstantinou 2019, 2020), developed a full-size single concave roller isolator made of urethane and reinforced with steel cores (which can help in load-bearing capacity, even if urethane is damaged).The structure isolated by an isolated system has acceptable collapse performance and fulfilled the displacement capacity requirement as per standard ASCE (2007).(Zhang, Ali, and Sun 2021) experimentally verified the effect of the value of the coefficient of friction on seismic energy dissipation.In their experiment, they tested different locally available materials for their ability to dissipate seismic energy.Before incorporating the materials in experimental models, all the materials were tested for the values of static and dynamic coefficient of friction.During their experimental study, they constructed small prototype models consisting of a top and bottom hollow concrete block representing the top and bottom masses of the structure.The isolation layer was 50 mm high and sandwiched between two Teflon sheets.The locally available low-cost materials used in different isolation layers include weak mortar (1:6) having normal sand (NS), coarse sand (CS) mortar (1:6), rubberized (RA) mortar (1:6) with 20% rubber aggregate by weight of mortar, steel balls (BB) of 9 mm diameter enclosed in the mortar (1:6) and asphalt cushion (AC) with 8% bitumen content.Cyclic loading tests were performed on all the samples, and the behavior was verified numerically by using the finite element software Abaqus.They concluded that among all the materials tested in their research, rubberized mortar dissipated maximum seismic energy.
Various other features that affect the behavior of seismic isolation systems include axial pressure, temperature, sliding velocity, base shear, and vertical excitation.All of these factors must be kept in mind while designing a friction-based seismic isolation system.The most suitable value of the coefficient of friction of an isolator is considered to lie in the range u = 0.15 to 0.20.The systems having higher values of the coefficient of friction are considered inadequate to dissipate seismic energy in the case of small to medium ground shaking (Constantinou, Mokha, and Reinhorn 1991).
The friction-based isolation systems are independent of the frequency variation of the ground motion, so they do not undergo resonance during earthquakes.These systems transfer such an amount of acceleration to the superstructure that is almost equal in magnitude to the maximum limiting friction (Brito et al. 2020;Hsu and Chang 2021;Jangid 1997;Kitayama and Constantinou 2021).During the 1985 Mexico City earthquake, more than 40% of the asymmetric buildings were adversely affected by torsional buckling.This is the major flaw of friction-based isolation systems (Tena-Colunga and Escamilla-Cruz 2007).The same problem was observed by many researchers during their studies in the past (Fallahian, Paytam, and Khoshnoudian 2016;Jangid 2000;Lee 1980;Nagarajaiah, Reinhorn, and Constantinou 1993).According to (Jangid 2000), there exists a direct relationship between the mass of the superstructure and base friction.Thus, there is no eccentricity between the center of mass and the center of stiffness of the superstructure in the case of friction-based isolation systems.The seismic behavior of structures is also greatly affected by the long-term behavior and service life of the isolation system.An effective sliding isolation system should have adequate lateral flexibility, it must undergo limited residual displacements and it must be able to stay stable during minor earthquakes and should develop inelastic behavior to dissipate energy during intensive shaking (Mazza 2019).(Ali et al. 2022) also proposed a Reinforced-Cut-Wall isolation system that is not only economical but also eco-friendly and easy to install and maintain.The experimental results are validated numerically using Abaqus.They modeled both isolated and fixed base structures in Abaqus and subjected them to five different ground motions to check the dynamic responses of the structures at various levels of shaking.Their models were provided with recentering bars at a distance of 150 mm, and these recentering bars were penetrated out from the foundation and passed through the isolation layer and vertically passed through a brick masonry wall up to a height of 304.8 mm in the 1:3 scaled model.
The purpose of this study is to investigate the effect of isolation layer thickness on the overall seismic performance of structures isolated with the reinforced cut wall (RCW) and to study the performance of the proposed RCW system in actual structures.The RCW system consists of an isolation layer made of weak mortar sandwiched between two sand layers and is reinforced with a series of vertical rebars distributed throughout the base.The isolation layer is a crucial component of the proposed friction-based isolator in terms of energy dissipation and enhancement of overall seismic performance, and thus needs to be designed carefully.The results of experimental and numerical studies conducted to validate the optimum value of isolation layer thickness are presented in this article along with the results obtained through cyclic loading tests conducted on RCW base-isolated and fixed base unconfined brick masonry walls.In this research, the RCW isolation layer material consists of a weak coarse-sand mortar that is prepared by using coarse aggregates obtained from Nizam Purr sand quarries.This weak mortar layer is sandwiched between two layers of loose coarse sand to facilitate the interface sliding.Different prototype hollow block samples with varying isolation layer thickness and a 1:3 reduced scale brick masonry wall with an isolation layer thickness of 75 mm (2.5 inches) and recentering rebars spacing of 152.4 mm (6 inches) were fabricated and experimentally tested by using a cyclic loading test.Similar specimens were also modeled in Abaqus to compare the results with the experimental one.

Material Properties
The 1:6 weak sand mortar with a 28-day compressive strength of 8.701 MPa (1262 Psi) used in this study is prepared by using locally available Ordinary Portland Cement and coarse sand obtained from Nizam-Pur sand quarry located in Nowshera, Khyber-Pakhtunkhwa, Pakistan.The fineness modulus of the sand is 2.8 and is calculated using the ASTM C136 standard testing procedure.The RCW prototype models were fabricated by using two types of hollow concrete blocks of 13.1 MPa compressive strength available in the local market.One set of hollow concrete blocks has 3 circular tubing while the other set has two rectangular tubing.Three rectangular tubing blocks and two circular tubing blocks were used to prepare each set of RCW samples with 50 mm, 65 mm, and 75 mm thicknesses.The recentering was achieved by providing vertical steel bars having a diameter of 6.44 mm based on a previous study (Zhang, Ali, and Sun 2021), penetrating through the bottom block, the RCW, and the top block.The 1:3 reduced scale RCW base-isolated brick masonry wall was made up of standard clay bricks reduced to 1:3 scale along all the sides and had a mass density of 1723 kg/m 3 .

Design of the RCW Isolated Coupon and Structural Models
The coupon RCW models used in the current study are composed of hollow concrete blocks having a compressive strength of 13.1 MPa (1900 Psi) and 203:2mm � 203:2mm � 406:4mm dimension.The RCW isolation layer is sandwiched between the top and bottom concrete blocks.This isolation layer is made up of 1:6 weak sand mortar made up of coarse sand and is sandwiched between a 3 mm layer of loose coarse sand.This RCW isolation layer is reinforced by using vertical steel bars having a diameter of 6.44 mm and penetrating through both top and bottom blocks.Figure 1a shows the detailed 3D sketch of the proposed isolation system that is implemented in an actual reduced-scale masonry model for the cyclic loading test.Similarly, Fig. 1b indicates the specimens used for the cyclic loading test to study the effect of the isolation layer thickness.These reinforcing bars provide a recentering mechanism for the whole RCW isolated assembly, prevent the residual displacement after the shaking is stopped, and also keep the structure intact against vertical loads.In order to identify the best isolation layer thickness, the prototype models were prepared to have an isolation layer thickness of 50 mm, 65 mm, and 75 mm, respectively.Two types of concrete blocks having hollow and rectangular tubing were used in these models to observe the effect of this type of block on the behavior of the whole assembly.Three sets of RCW models were prepared for each value of the isolation layer thickness in the case of rectangular tubing, while two in the case of blocks with circular tubing.This weak mortar layer is supposed to remain intact in case of mild shaking (minor earthquakes or wind-induced vibrations), while in the case of severe shaking (major earthquakes), this layer is expected to undergo cracking to dissipate energy when the frictional resistance is not sufficient to dissipate all the seismic energy.The loose sand layers provided above and below the isolation layer would allow the sliding of the substructure and superstructure and would also help to dissipate seismic energy through this sliding friction.The same design strategy is adopted in the actual 1:3 reduced scale unconfined brick masonry wall model.

Sample Preparation
Two types of block samples were prepared, one using hollow concrete blocks with rectangular tubing and the other with circular tubing having a compressive strength of 13.1 MPa (1900 Psi).The dimensions of the hollow block's sample were 406.4 mm in length, 203.2 mm in width, and 475 mm in height.In the first stage of sample preparation, two 6.35 mm bars are anchored in the hollow portion of the blocks and then 1:3:6 concrete is filled in the lower blocks.After proper curing, the next stage is the construction of the RCW layer.First of all, the top surface of the bottom block is properly dried and then a 3 mm layer of loose sand is laid on its surface.The next step is to lay a layer of weak mortar.Three sets of RCW prototype samples were prepared on the basis of differences in the height of the RCW isolation layer from the blocks having rectangular tubing and two from the ones having circular tubing.One set of blocks had a 50 mm isolation layer, the second set had a 65 mm and the third 75 mm thick weak mortar layer.Once the weak mortar layer is dried, another layer of loose sand is laid on its top in order to sandwich it between two layers of 3 mm loose sand.The last step is to place the top block and then pour concrete into its hollow portion.For each value of isolation layer thickness, three blocks were prepared with rectangular and two with circular tubing in order to maintain the accuracy of cyclic loading test results.After the completion of the final fabrication stage, all the samples were cured for 28 days to gain maximum strength.Different stages of prototype sample preparation are shown in Fig. 2.
The unconfined brick masonry wall isolated with the proposed RCW isolation system had a dimension of 1200 mm in length, 1200 mm in height and 76.2 mm in thickness.This wall was isolated by installing the reinforced cut wall at its plinth level.The reinforced cut wall isolation system was composed of a 65 mm thick layer of weak sand mortar sandwiched between a 3 mm layer of loose coarse sand.Once prepared, the brick wall was properly cured before experimentally testing it for the practical investigation of the proposed isolation system.The recentering rebars were penetrating vertically through the foundation and plinth beam up to the first bricklayer in the wall.The recentering rebars were installed at a lateral center-to-center distance of 152.4 mm (6 inches) as validated (Zhang, Ali, and Sun 2021).There were two concrete beams in the wall.One is the plinth beam placed on the top of the isolation layer, while the other is the top beam provided as a platform for the application of lateral and vertical gravity loads.Different stages of RCW isolated brick wall construction are mentioned in Fig. 3.

Instrumentation and Loading Protocol for the Block Samples
The specimens were subjected to cyclic loading with displacement increments of 0.5 mm, 0.75 mm, 1 mm, 2 mm, 3 mm, 4 mm, 6 mm up to 26 mm, and the load was applied until the isolation layer was cracked to such an extent that it was no more able to take the horizontal load.The cyclic loading history adopted during the test is shown in Fig. 4.
The effect of isolation layer thickness on the energy dissipation capacity was studied by a cyclic loading test.A prefabricated steel box was used to keep the samples laterally stable and to restrict the horizontal motion of the bottom block.This assembly also helped to minimize the risk of outof-plane failure during the cyclic loading test.This assembly was then clamped to the steel girder using nuts and bolts.A vertical hydraulic jack with a loading capacity of 50 tons (491 kN) was used to apply a constant vertical load of 1.2 tons (12.25 kN).A hydraulic jack of 50 tons (491 kN) loading capacity was used to apply a lateral cyclic load.The displacement response of the top block and RCW was monitored with the help of 3 LVDTs (60 mm displacement capacity) attached on both sides of the sample to monitor any out-of-plane behavior of the upper block.A 32-channel  data acquisition system compatible with the LVDTs was used to capture the force deformation response of the models.The compression of the isolation layer under vertical load was monitored with the help of a 25 mm capacity dial gauge connected vertically to the RCW.The testing setup is shown in Fig. 5.

Cyclic Loading Test Setup and Loading Protocols for Reduced Scale Masonry wall
Displacement-controlled cyclic loading tests were performed in order to investigate the seismic response of the reduced scale RCW isolated and fixed brick masonry wall.The 3 rd scale masonry wall is designed according to the similitude requirements and scaling factors mentioned in Table 1.A typical full-scale masonry wall used locally is 3200 mm in height, 3660 mm in width, and 230 mm in thickness (12 0 � 12 0 � 9 00 ).The testing setup and loading protocols are mentioned in Fig. 6.Hydraulic actuators, with a capacity of 50 tons, were used to apply constant vertical and varying horizontal loads.The specimen was bolted to the steel beam on both sides to ensure no out-of-plane displacements.Displacements in different directions were recorded with the help of Linear Velocity Displacement Transducers (LVDTs).Dial gauges were also connected at various locations to measure the compression of the isolation layer under a constant vertical load of 12.25 kN. Figure 7 indicates a detailed sketch of the instrumentation layout.

Numerical Verification
The proposed RCW system was also verified numerically using the finite element analysis software Abaqus.Initially, the concrete block specimens having the isolation layer of various thicknesses were modeled and subjected to the same cyclic loading history to compare the results with the one obtained during the experimental study.Figure 8a,b show the mesh model and boundary conditions adopted for the numerical study, respectively.The concrete block and isolation layer were modeled as a 3D solid element having 8 nodes (C3D8R) and the re-centering rebars were modeled as a truss element (T3D2), which is a linear 2-node linear 3D element.All the elements, such as concrete blocks, isolation layer, loose sand, and re-centering rebars, were modeled separately and then combined into a single model as shown in Fig. 8.The interface friction between the sliding surfaces was modeled as surface-to -surface interaction using penalty as friction among the sliding interfaces.The interface friction of 0.253 was used in the previously conducted experimental study (Ali et al. 2022;Zhang, Ali, and Sun 2021).The boundary conditions applied are shown in Fig. 8b, where the bottom block is fixed, and the top block is allowed to slide in the horizontal direction under a given cyclic amplitude.A constant vertical load of 12.25 kN (1.2 tons) was also applied to the center of the top concrete block, which is evenly distributed on the upper surface of the concrete block through the MPC constraint available in Abaqus.

CPD Model for Concrete Blocks and RCW
To simulate the non-linear response of the concrete blocks and the isolation layer, the widely used concrete damage plasticity model (CDP) was also adopted here (Aghaeipoor and Alembagheri 2022;Deng and Yang 2020;Habieb, Valente, and Milani 2019;Li et al. 2022;Zeng et al. 2022).The physical properties of concrete, isolation layer, and rebars are  mentioned in Table 2.The compression and tension behavior of concrete and RCW weak mortar used in developing the CPD model are given in Tables 3 and 4, respectively.Various important properties were obtained from the numerical study, such as hysteresis response, stress-strain behavior, and damage propagation under compression and tensile loading.The uniaxial compressive and tension stress-strain curve of the CDP model used in the numerical study is illustrated in Fig. 9.

Fineness Modulus of Sand
The ASTM C136 standard testing procedure was used to calculate the fineness modulus of the sand used in both RCW weak mortar and the 3 mm loose sand layer.The sand was coarse sand with a fineness modulus of 2.8 as shown in Fig. 10 and was suitable to provide the sliding friction between the different layers of RCW.Table 5 shows details of the properties of the sand used, and its fineness modulus test results.

Coefficient of Friction
ASTM 513 D1984-14 standard testing procedure (Committee 2014) was used to determine the coefficients of friction (static and dynamic).The static coefficient of friction between loose sand and weak-sand mortar was 0.362, while the dynamic coefficient of friction was 0.253.The values of the coefficients of static and dynamic friction lie in the recommended range (u = 0.15 to u = 0.40) as suggested by the researchers (Quaglini et al. 2022).

Visual Observation and Overall Response of the Isolator
At the start of the cyclic loading test, all the samples behaved as a fixed base system ensuring hard contact among the sliding interfaces to preserve the structure's integrity during minor seismic events.As the proposed isolation system is not meant to be activated by wind or other minor earthquake loadings.However, after a few cycles, when the friction resistance between the sliding interfaces was overcome by the lateral load, then the isolation layer was activated and the sliding of the isolation layer between two layers of loose sand was observed clearly.When the displacement increment was increased, minor diagonal cracks were observed in the isolation layer.When the lateral displacement was above 20 mm then wide cracks and spalling were observed in the RCW and finally the test was stopped when slippage of major portions of RCW was observed.The activation shear is initially controlled by the friction capacity of the sand layer, as the RCW isolator is activated when the externally applied load overcomes the frictional resistance.With the increase in lateral load, the shear capacity of the mortar layer will then dissipate the energy in the form of cracks.Similarly, during the elastic stage, the recentering rebars have the potential to recenter the model both in the case of rocking and shear mechanisms.Due to the rebars' low levels of inelasticity, their ability to control sliding is not jeopardized.However, when the externally applied load exceeds and results in permanent deformation in rebars, then they will not be able to recenter the model.Therefore, it is necessary to design the isolator re-centering mechanism in such a way that it remains elastic for the maximum considered earthquake in a particular region where the structure is located.

Analysis of Hysteretic Behavior of the Samples of the Block
The hysteretic performance for the block samples under cyclic loading is described with the help of force-deformation hysteresis loops.The force-deformation hysteresis curves are based on the cyclic loading test data, and the backbone curves are obtained by connecting the peaks of these hysteresis curves as mentioned in Figs.11-13.
All the samples have almost similar failure mechanisms under the effect of repeated cyclic loading.There are three stages of failure observed.These three stages are the elastic, inelastic, and failure stages.Figure 14 indicates the crack propagation during the cyclic loading test corresponding to various drift levels.The samples are uncracked during the elastic stage, while small vertical and diagonal cracks appeared in the inelastic stage, while the RCW isolation layer is completely cracked at the failure stage.During the elastic stage, the rebar has very small stress and strain and shows no action.With the increasing number of loading cycles and load amplitudes, cracks begin to develop and widen.When the samples enter the inelastic stage, the loose sand starts spalling from the edges and the isolation layer has already started sliding (as the frictional resistance was overcome) between two layers of loose sand.The final stage is the failure of the sample when there is a huge spalling of the mortar layer and the rebars are visible at the RCW portion.The loading is stopped beyond this point.Also, it is worth mentioning that the lateral load-carrying percentage for the initial cycles (when no sliding occurs) was relatively higher as compared to the next few cycles when the sliding started.Furthermore, when some major cracks appeared, the percent increase in the lateral load-carrying ability of the specimens was relatively lower as compared to the initial cycles.The specimen with the 50 mm thick isolation layer showed a rigid behavior up to a lateral displacement of 1.5 mm, without starting to slide where the lateral load was 4.21 kN.The sliding was observed in a 2.5 mm cycle and the specimen remains in the elastic range without any visible cracks up to a lateral displacement of 4 mm, where the corresponding load was 6.548 kN.At the 14 mm cycle, visible cracks appeared vertically in the isolation layer and the lateral load was 14.65 kN.The maximum lateral load observed for RCW-50 mm was 20.263 kN corresponding to a 22 mm cycle, where some major cracks have appeared in the sample.
In the case of the 65 mm isolation layer, the sliding was observed at 2.0 mm, before which the specimen showed a rigid behavior.The RCW-65 mm specimens took a maximum lateral load of 22.60 kN at a drift of 25 mm where some diagonal cracks appeared in the isolation layer.Similarly, for RCW-75 mm samples, the sliding was observed at a 3.00 mm cycle and the specimen took a maximum load of 15.055 kN corresponding to a 24 mm drift.
The skeleton curves mentioned in Fig. 15 are obtained by connecting the points of peak displacements and loads with each other.However, it is difficult to analyze the exact material response from these skeleton curves because of damage accumulation with the increase in loading cycles.Therefore, the simplified and interpolated skeleton curves are plotted as shown in Fig. 15.These simplified skeleton curves provide a better representation of the degradation of RCW during a seismic event.In Fig. 16 section OA represents the elastic stage, AB is the elastic-plastic stage, while BC shows the plastic declination stage where all the isolation material is crushed, and steel bars are visible.A similar trend was observed during the test as mentioned in Fig. 15.It can also be observed that the specimen with a 65 mm isolation layer has a vast elastoplastic range compared to the others.
The combined skeleton curves obtained from the experimental and numerical study shown in Fig. 15a,b indicate the maximum lateral load and the corresponding displacement.The plastic descent stage of the samples during the experimental study with 50 mm and 75 mm thick RCW samples starts at a load of 20.263 kN and 15.055 kN, respectively, while on the other hand, the samples that have 65 mm thick RCW remain in the elastoplastic stage up to 22 mm lateral displacement, where the corresponding lateral load was around 21 kN beyond which cracks appeared in the samples.However, these specimens still have the capacity to take an additional lateral load, peaking at 22.6   kN corresponding to a 26 mm cycle.Similarly, during the numerical study, the plastic descent stage for RCW-65 mm starts when the lateral load reaches 23.725 kN, corresponding to 24 mm drift.The energy dissipated by the samples appears in the form of the cracking of RCW layers, while in the case of the samples having a 65 mm thick RCW layer, the energy dissipated by the sample is mainly consumed to facilitate the sliding of the RCW layer between two layers of loose sand and to overcome the sliding friction.

Behavior of RCW Layer Under Vertical Compressive Loading
In order to check the vertical load-bearing capacity of the RCW and to study the effect of the isolation layer on vertical load-carrying capacity, all the specimens were subjected to an axial vertical load with a maximum amplitude of 12 kN.The same load was sustained during the cyclic loading tests of the samples, and the value is based on the compressive strength of the RCW.The behavior of samples with different thicknesses of RCW layer under vertical compressive loads is represented in Fig. 17.The samples with 50 mm isolation layer thickness showed minimum compression under vertical load and were compressed up to 2 mm and after that the compression was constant.In the case of samples having a 65 mm thick isolation layer, vertical compression was around 2.2 mm at a vertical load of 10 kN, and the compression was constant up to a vertical compressive load of 12 kN.The samples with 75 mm isolation layer  thickness underwent maximum compression under vertical load with a compression of 2.8 mm at a load of 12 kN.All the specimens had enough vertical strength to withstand the self-weight of the structure.

Effect of RCW Thickness on Equivalent Damping and Energy Dissipation
The equivalent damping ratio is an important parameter that can be used to assess the energy dissipation capacity of the samples.Figure 18 indicates the variation of damping corresponding to the percentage drift.As mentioned, the hysteresis damping for all specimens continues to increase beyond a drift of 0.5%.For the 50 mm RCW specimen, the hysteresis damping was 17.4% corresponding to a peak lateral drift of 5.81% in the case of the experimental study and 18.52% corresponding to a peak lateral drift of 5.66% in the case numerical study.Similarly, for the 65 mm specimen, the damping was 24.52% corresponding to 5.82% drift, and 23.5% corresponding to 5.54% drift in the case of experimental and numerical study, respectively.Also, for the 75 mm specimen, the maximum hysteresis damping was 21.05% and 19.8% corresponding to a peak drift of 5.29% and 5.40% in the case of experimental and numerical study, respectively.Even though the variation of thickness of the isolation layer does not considerably influence the hysteresis damping variation corresponding to drift.Despite this, the RCW sample with a 65 mm thick isolation layer indicated higher hysteresis damping as observed in Fig. 18 which results in higher energy dissipation, and the results were found in agreement with the high energy dissipation ability of low-cost seismic isolation consisting of a sandwiched sand layer, demonstrated experimentally during the large scale shaking table tests performed by (Tsiavos et al. 2020).
The seismic dissipation capacity of a structure is the maximum amount of energy dissipated by a structure and it results in slight cracking or sudden collapse of a structure.During the cyclic loading test, the seismic energy dissipated by the sample during a specific cycle is represented by the area under the complete hysteresis loop for that cycle.The accumulative energy dissipated by the sample is calculated as the sum of areas under all the load-displacement hysteretic curves (Abbas, Adil, and Ali 2021;Javidan, Ali, and Kim 2022;Nasab et al. 2021).Figure 19 shows the accumulative seismic energy dissipated by each specimen during the cyclic loading.It is observed that specimens with RCW thickness of 65 mm show maximum energy dissipation capacity as compared to the other two types of specimens.
Stiffness is a very important feature in the seismic analysis of structures.There is an inverse relationship between drift and stiffness, thus it decreases with the increase in drift (Bybordiani and Arici 2019).The secant or normalized stiffness is represented as the ratio between stiffness during a particular cycle and initial stiffness k 0 .The relationship between the percentage drift ratio and normalized stiffness is presented in Fig. 20.The samples with a 50 mm thick RCW layer show the  least initial stiffness and undergo less stiffness degradation as compared to the other two specimens.Even though the samples with a 75 mm thick RCW layer have maximum initial stiffness, it undergoes maximum stiffness drop with displacement.The sample with a 65 mm thick RCW layer has enough initial stiffness and sustains its value better with displacement as compared to the other two samples.

Hysteresis Response
The effect of the isolation layer thickness of the proposed RCW isolator is also investigated numerically by modeling the same assembly in Abaqus, which was used during the experimental study as shown in Fig. 8.The hysteresis response obtained from the numerical investigation matches fairly well with the one obtained from the experimental study for all the models as shown in Figs.21-23.The force-deformation response is then analyzed to obtain various important factors, such as backbone  curves, energy dissipation, damping variation, and stiffness degradation.All these parameters are then compared with the results obtained from the experimental results as mentioned above.

Stress-Strain Response and Damage Propagation
Figures 24 and 25 indicate the Mises stress and maximum strains developed under cyclic loading.The maximum stresses are concentrated at the isolation layer for all types of specimens as shown in Fig. 24 indicating the suitability of the proposed isolation system, as maximum damages would be concentrated at the isolator.Similarly, the maximum principal logarithmic strain is also concentrated at the isolation layer as mentioned in Fig. 25.This also points out that maximum damage would be observed by the isolation layer due to its low lateral stiffness as compared to the substructure and superstructure.Damage propagation is also an important parameter for the seismic evaluation of the structures.Figure 26 indicates the compressive and tensile damage propagation of the prototype sample under cyclic loading.It can be observed that the tensile and compressive damages are also concentrated at the isolation layer and none of the top and bottom blocks got damaged.This indicates that the maximum seismic energy would be dissipated within the isolation layer instead of transferred to the superstructure.
The re-centering rebars bring back the superstructure to its original position after the seismic event has passed.It also has some adverse effects on the isolator, as it introduces punching shear in the  isolation layer when subjected to horizontal loading.As mentioned in Fig. 25, the maximum inelastic strain was generated in the location of re-centering rebars.The re-centering rebars need to be designed in such a way that they remain in the elastic region for earthquake design.The under-design of the recentering rebars would lead to the yielding of rebars before reaching the design limits.On the other hand, if the re-centering rebars are over designed then it would increase the lateral stiffness of the isolator and practically no sliding would occur.

Hysteretic Response of the RCW Base-Isolated Wall Model
The results presented here obtained from the experimental study of the brick masonry walls are for the reduced scale model as the similitude requirement mentioned in Table 1 was followed to obtain various outputs.The load-displacement hysteretic loops are used to study and compare the seismic response parameters of the reduced-scale RCW isolated wall with the one with the fixed base.  Figure 27.indicates the damage evolution in both models.As the primary purpose of the proposed isolation system is to achieve the collapse prevention performance level, therefore, the test was also extended up to the complete collapse of either the isolation system or the masonry wall.The isolated sample took a peak lateral load of 74.8 kN corresponding to the lateral displacement of 12 mm, where some major cracks have appeared in the isolation layer.However, the isolated sample was subjected to a peak lateral displacement of 16 mm where the corresponding lateral load was reduced to 67.5 kN, beyond which an out-of-plane behavior was also observed in the sample.On the other hand, in the case of a fixed base model, the peak lateral load was 48.1 kN corresponding to 6 mm lateral displacement and the maximum lateral displacement to which the fixed model was subjected was 10 mm where the corresponding lateral load was 42.4 kN.
The dark green line plotted in Fig. 28 is the backbone curve and is obtained by connecting the peaks of all the individual hysteretic loops.The hysteretic response is divided into four stages of failure that include the uncracked stage called the elastic stage, the minor cracking called the inelastic stage, major crack propagation called the ultimate stage, and the last stage of complete failure.The first stage is the elastic stage where all the stresses in steel and mortar are reversible  with no sign of cracking.The second stage is the startup of minor cracking which is caused by an incremental increase in displacements and lateral load.This stage is seen as the pinching of hysteresis loops.After a further increase in the lateral load, the isolation layers start sliding past each other with the formation of further thin diagonal cracks in the masonry.Further, an increase in load beyond this stage slows down the rate of crack propagation, which is depicted by the slow bulging up of the hysteretic loops and the pinching phenomenon starts disappearing.Up to a lateral displacement of 6 mm, the rate of crack propagation is slow and beyond this point, horizontal cracks start developing in the bottom area of the masonry wall and the isolation layers.The last stage of failure is a major cracking stage in which wide vertical cracks appear in the isolation layer breaking up into larger parts that ultimately leads to spalling of the isolation layer out of the wall.This is the stage of major energy dissipation through the cracking of the isolation layer.The lateral load application is stopped beyond this point as the loadbearing capacity of the system gets reduced by 85% of the peak lateral load.
In order to compare the load-displacement response characteristics of the RCW isolated wall with the normal fix-based structure, a fix-based wall of the same size and scale was also tested using a cyclic loading test and the hysteresis loops for this wall are given in Fig. 29.The fix-based wall was unable to take lateral load beyond a lateral displacement of 10 mm, and a very small amount of energy was dissipated through cracking up of this wall as compared to the RCW isolated wall.
The energy dissipated during the cyclic loading tests by the fix-based wall is almost half of the energy dissipated by the RCW isolated wall, thus reflecting the efficiency of the proposed isolation system in absorbing the input energy without undergoing complete failure as shown in Fig. 30.The energy dissipated by the isolated wall is around 3600 kN-mm while, on the other hand, the energy absorbed by the fix-based wall is around 1600 kN-mm.
Considering the response of the reduced scale model, a similar response is expected for the fullscale model as the response of the reduced scale model was obtained using the similitude requirements.

Conclusions and Recommendations
In this study, the effects of the isolation layer thickness on the seismic performance of a previously developed low-cost base isolation technique called Reinforced-Cut-Wall (RCW) is studied, and the same is verified by installing the same base isolation system in a reduced-scale brick masonry wall model.The proposed isolation system is not only economical and efficient but can also be practically implemented in low-rise masonry buildings to enhance their seismic performance.In the current study, small prototype block samples were prepared which consist of a seismic isolation system sandwiched between a top and bottom block.Three different thicknesses of the isolation layer were used i-e 50 mm, 65 mm, and 75 mm.All the block samples and the masonry wall were subjected to displacement-controlled cyclic loading tests and the cracking pattern, failure modes, and sliding of the isolation layer were observed for different thicknesses of the isolation layer.This paper also presents a detailed discussion of key parameters obtained from the results of the cyclic loading test including stiffness degradation, displacement ductility, equivalent viscous damping, and seismic energy dissipation.From the experimental and numerical studies conducted during this work, the following conclusions were drawn: (1) Three different types of failures were observed in all the samples during cyclic loading testing.
Horizontal cracking of isolation layer, diagonal cracking, and slippage, and spalling of isolation layer under extreme loading event.Both flexure and shear failures were observed in the isolation system, and none of the top and bottom blocks was damaged.This indicates that the maximum damages would be limited to the isolation layer without transferring it to the upper structures.Similarly, steel bars used for the re-centering mechanism did not yield and were visible after spalling and severe slippage of the isolation layer.
(2) The hysteresis loops showed a clear and symmetric pinching phenomenon in all the samples.
The maximum lateral resistance in the case of the samples with a 65 mm thick isolation layer was 22.6 kN, which is more than the ones with 50 mm and 75 mm thick isolation layers where it was 20.2 kN and 15.0 kN, respectively.There was a gentler descent and a wider plateau around the maximum lateral resistance observed in hysteresis curves of the samples with 65 mm thick RCW as compared to the other two types of samples.This indicates that these samples have better ductility and greater seismic energy dissipation capacity as compared to the rest of the samples.
(3) It can also be observed from experimental and numerical results that the energy dissipation capacity of the samples is considerably affected by the variations in the thickness of the isolation layer.The equivalent viscous damping also keeps on increasing and becomes constant at a much later stage in the case of samples with 65 mm as compared to the samples with 50 mm and 75 mm thick RCW.(4) The coefficient of static and dynamic friction of sand used in sliding layers are 0.362 and 0.253, respectively, which ensure the initial resistance to lateral loads during minor shaking and facilitation of sliding of the isolation layer in case of severe seismic shaking without undergoing uncontrolled slippage due to re-centering rebars.The loose sand layer behaved rigidly and monolithically at the start of lateral loading cycles and with the increase in lateral load, this layer became activated, and the isolation layer kept sliding between two sand layers.Thus, it utilizes the seismic energy through sliding friction and preventing its transfer to the top block.(5) No cracking of top and bottom concrete blocks was observed at all the stages of cyclic loading tests, which indicates a better seismic dissipation capacity of the isolation layer; however, this could not be the final judgment drawn from these results as these are the prototype models and to observe the performance of this isolation system in actual buildings, such models need to be constructed which are actual representation of masonry buildings.To understand the behavior of the proposed isolation system and the effects of various parameters, it is necessary to install this isolation layer in actual building models and test it experimentally and numerically.Further research is needed to be carried out in this field to optimize the size and spacing of recentering rebars, their effectiveness in the case of extreme events and the possibilities of retrofitting the proposed isolator after the seismic event has passed.(6) From the hysteresis response of the 1:3 reduced scale RCW brick masonry wall, it has been observed that the energy dissipation capacity of the RCW base isolated system increases up to almost double as compared to a fixed-based structure of the same size.(7) The design properties of the recentering rebars depend on the size of the structure (scale effect) and the dynamic effect of the excitation and cannot be accurately obtained through the presented small scale, static-cyclic tests.Therefore, large scale shaking table tests are needed to verify experimentally the ability of the rebars to recenter the structure back to its original position.In addition, due to the presence of vertical rebars, the dissipator is also efficient under vertical ground shaking where the rocking phenomena are prominent.

Disclosure Statement
No potential conflict of interest was reported by the author(s).

Figure 1 .
Figure 1.(a) 3D sketch of proposed isolation system; (b) Specimen details for the experimental and numerical study.

Figure 3 .
Figure 3. Different stages of RCW isolated brick wall construction.

Figure 4 .
Figure 4. Cyclic loading history used for the experimental and numerical study.

Figure 5 .
Figure 5. (a) Schematic diagram of the cyclic loading test setup; (b) Cyclic loading test assembly.

Figure 7 .
Figure 7. Sketch of instrumentation layout used for cyclic loading test.

Figure 9 .
Figure 9.The concrete damage plasticity model used; (a) the uniaxial compression behavior of concrete, (b) the uniaxial tension behavior of concrete, (c) the uniaxial compression behavior of mortar, and (b) the uniaxial tension behavior of the mortar.

Figure 11 .
Figure 11.Hysteresis response of RCW specimen with 50 mm isolation layer.

Figure 12 .
Figure 12.Hysteresis response of RCW specimen with 65 mm isolation layer.

Figure 13 .
Figure 13.Hysteresis response of RCW specimen with 75 mm isolation layer.

Figure 15 .
Figure 15.Interpolated skeleton curves for different isolation layer thicknesses: (a) from the experimental study; (b) from the numerical study.

Figure 17 .
Figure 17.Compression of isolation layers under vertical loading.

Figure 18 .
Figure 18.Variation of damping with change in % drift.

Figure 20 .
Figure 20.Degradation in normalized stiffness with percentage drift.

Figure 21 .
Figure 21.Hysteresis response comparison of the experimental and numerical case for RCW-50 mm.

Figure 22 .
Figure 22.Hysteresis response comparison of the experimental and numerical case for RCW-65 mm.

Figure 23 .
Figure 23.Hysteresis response comparison of the experimental and numerical case for RCW-75 mm.

Figure 26 .
Figure 26.Observed (a) Compressive and (b) Tensile damages during the numerical study.

Figure 27 .
Figure 27.Damage pattern of the reduced scale masonry wall models.(a) the fixed base model, and (b) the Isolated model.

Figure 28 .
Figure 28.Hysteresis loops of reduced scale RCW isolated wall.

Figure 29 .
Figure 29.The force-displacement hysteresis data of fix-based wall.

Figure 30 .
Figure 30.Energy dissipated by RCW isolated and fix-based wall.

Table 1 .
Similitude requirement for reduced scale brick masonry model.
a ¼ a p =a m

Table 2 .
Physical properties of prototype material.
b0 ¼ bidirectional compressive strength, f c 0 ¼ unidirectional compressive strengthK = ratio of second stress invariant on the tensile meridian, e is the eccentricity.

Table 3 .
Tensile and compressive properties of concrete utilized in the CPD model.

Table 4 .
Tensile and compressive features of RCW mortar used in the CPD model.

Table 5 .
Sieve analysis of coarse sand.