Enhancing seismic resilience of buildings using prefabricated foamed concrete infill walls in earthquake-prone Indonesia

Abstract

Indonesia is highly exposed to seismic hazards, requiring building systems that can improve structural safety while remaining feasible for widespread application. This study investigates the effectiveness of prefabricated foamed concrete infill walls in enhancing the seismic resilience of buildings in earthquake-prone regions of Indonesia. The research integrates material characterization, structural performance assessment, and seismic response analysis to evaluate the role of lightweight prefabricated infill systems in reducing seismic demand on structural frames. Experimental testing and numerical simulations were conducted to analyze load–displacement behavior, energy dissipation capacity, and damage patterns under earthquake loading scenarios representative of Indonesian seismic conditions. The results demonstrate that the use of prefabricated foamed concrete infill walls significantly improves lateral stiffness and energy dissipation while reducing structural damage and overall seismic vulnerability. Compared to conventional infill systems, the proposed approach offers improved constructability, reduced structural mass, and enhanced post-earthquake functionality. This study provides evidence that integrating prefabricated foamed concrete infill walls into building design can serve as an effective and scalable strategy to enhance seismic resilience in developing, high-risk seismic regions, supporting safer and more sustainable built environments in Indonesia.

1 Introduction

Indonesia is located in one of the most seismically active regions in the world due to its position along the Pacific Ring of Fire, exposing its built environment to frequent earthquakes (Reid and Mooney, 2023). The catastrophic impacts of past seismic events, including the 2004 Aceh earthquake and the 2018 Lombok earthquake, have highlighted the vulnerability of conventional building systems in Indonesia, particularly in terms of structural damage and human casualties (Setyonugroho and Maki, 2024). Traditional reinforced concrete frames with masonry infill walls often fail under lateral seismic forces, primarily due to brittle behaviour and poor energy dissipation capacity (Rahman and Setiawan, 2018). As a result, there is a pressing need for innovative structural solutions that combine safety, constructability, and affordability for widespread implementation in earthquake-prone regions.

Prefabricated building components have emerged as a promising approach to enhance seismic resilience while addressing construction efficiency and quality control (Song et al., 2025). Among these, foamed concrete has gained attention for its lightweight properties, ease of installation, and favourable energy dissipation characteristics under dynamic loads (Mansyur and Tjaronge, 2025; Parung et al., 2023a; Parung et al., 2023b). Previous studies have demonstrated that incorporating lightweight infill walls can reduce seismic demand on structural frames and improve overall building performance (Ali et al., 2025; Zhang et al., 2024; Zheng et al., 2024). However, most research has focused on conventional concrete or small-scale experimental setups, with limited exploration of prefabricated foamed concrete systems in realistic earthquake scenarios relevant to Indonesia.

Indonesia is situated along numerous active earthquake faults, making seismic events a recurrent hazard across various regions of the country. In the construction of residential buildings, masonry infill reinforced concrete (RC) frames are widely employed for both exterior walls and interior partitions. Despite the prevalence of this structural system, a significant number of houses and low-rise buildings are built without adherence to established construction standards, proper detailing, or sound structural design, categorizing them as non-engineered structures. Burnt clay bricks are the most commonly used infill material for RC frames in residential buildings ranging from one to three stories, as well as in schools, offices, and other public buildings. Recent seismic events in Indonesia, such as the Palu Earthquake in 2019 and the Mamuju-Majene Earthquake in 2021, have demonstrated the high vulnerability of RC frame buildings with burnt clay brick infills under seismic loading, as illustrated in Figure 1. These observations have prompted increased attention and efforts toward mitigating the use of burnt clay bricks as infill walls to enhance earthquake resilience.

FIGURE 1

Recent numerical and experimental studies indicate that foamed concrete infill walls can significantly enhance lateral stiffness and ductility while mitigating damage accumulation in reinforced concrete frames (Zhang et al., 2023; Xu and Chen, 2023; Yang et al., 2025). These findings suggest potential benefits for seismic-prone regions, yet few studies have directly addressed the integration of prefabricated foamed concrete infill walls into medium-to high-rise buildings in developing countries. Moreover, the impact of such systems on post-earthquake functionality, construction time, and cost-effectiveness remains underexplored (Peng et al., 2026). Consequently, translating these promising laboratory results into practical design strategies requires comprehensive evaluation through both material characterization and structural performance assessment.

In Indonesia, the adoption of prefabricated seismic-resistant systems has been limited, partly due to regulatory gaps, lack of local expertise, and insufficient experimental evidence supporting their effectiveness (Pribadi et al., 2023). Previous efforts in retrofitting and upgrading existing buildings have shown improvements in seismic performance but often involve labour-intensive and time-consuming processes (Couto et al., 2024; Laghi et al., 2017). Therefore, lightweight prefabricated solutions such as foamed concrete infill walls present an opportunity to enhance safety while maintaining rapid deploy ability, which is critical in densely populated urban areas and disaster-prone regions.

The current study aims to fill this research gap by systematically investigating the seismic performance of prefabricated foamed concrete infill walls in reinforced concrete frames under earthquake loading scenarios representative of Indonesian conditions. Material testing, numerical simulation, and structural analysis are integrated to evaluate load–displacement behaviour, energy dissipation, and damage patterns. The novelty of this research lies in assessing a scalable, lightweight, and prefabricated infill system that not only enhances lateral stiffness but also supports improved post-earthquake functionality compared to conventional masonry or cast-in-place solutions.

By comparing the proposed system with conventional infill walls, this research provides insights into potential improvements in seismic resilience, including reductions in structural mass, construction time, and overall vulnerability. Furthermore, the study contextualizes findings within existing literature on lightweight and prefabricated seismic solutions, highlighting both consistencies and divergences with previous experimental and analytical investigations. This comparative approach ensures that the recommendations are grounded in current scientific knowledge while addressing the specific challenges of earthquake-prone regions in Indonesia.

Ultimately, the outcomes of this research are expected to inform building design codes, guide policy-making for disaster-resilient construction, and provide evidence-based strategies for integrating prefabricated foamed concrete infill walls into new and existing buildings. By demonstrating the effectiveness, scalability, and practicality of this system, the study contributes to safer, more sustainable, and resilient urban environments in seismic regions, aligning with global objectives of disaster risk reduction and sustainable development (UNDRR, 2021).

2 Materials and Methods

2.1 Summary of the experimental program

The experimental program was designed to systematically investigate the seismic performance of reinforced concrete (RC) frame buildings with prefabricated foamed concrete (FC) infill walls. This study addresses a critical research gap by evaluating the behavior of FC infill under earthquake loading conditions representative of Indonesia’s seismic environment. The methodology integrates comprehensive material characterization, full-scale lateral cyclic structural testing, and advanced numerical simulations to assess both material-level properties and overall structural response. Full-scale specimens were designed to reflect typical low-to mid-rise residential buildings while incorporating the innovative FC infill system. Laboratory experiments, standardized test protocols, and computational modeling were combined to ensure reliability, reproducibility, and practical applicability. The overall approach provides a robust framework for understanding the potential of FC infill walls as a safer, lighter, and more sustainable alternative to conventional masonry in earthquake-prone regions.

2.2 Materials characterization

2.2.1 Foamed concrete infill

Prefabricated foamed concrete (FC) panels were produced using a mixture of cement, fine aggregates, water, and a foaming agent. Visual observation of the fresh concrete revealed a cohesive and adhesive mix with no signs of segregation or bleeding, and the measured slump was 22 cm, indicating good workability for casting. Hardened FC specimens were cured under controlled laboratory conditions and tested at 3, 7, and 28 days to evaluate changes in physical and mechanical properties over time. The unit weights measured at these ages were 1508.3, 1476.9, and 1457.7 kg/m³, showing a gradual decrease with curing duration. Based on SNI 03–3449–2002 and ACI 213R-03, these values classify the panels as structural lightweight concrete, within the acceptable range of 800–1850 kg/m³.

The produced foamed concrete exhibited unit weights ranging from 850 to 1,200 kg/m³, depending on the mix proportion and curing stage. In comparison, conventional red clay brick typically has a unit weight of approximately 1,600–1,900 kg/m³, while lightweight brick (such as AAC or cellular lightweight blocks) generally ranges between 500 and 900 kg/m³. These values indicate that the foamed concrete used in this study falls within the lightweight material category and is substantially lighter than traditional fired clay brick. The reduced unit weight contributes to lowering the overall seismic mass of the structure, which is particularly advantageous in earthquake-prone regions such as Indonesia. Compared to conventional masonry infill, the use of foamed concrete panels therefore offers a favorable balance between structural contribution and mass reduction, potentially improving global seismic response.

Compressive strength testing showed progressive increases with age, reaching 5.32, 6.16, and 7.38 MPa at 3, 7, and 28 days, respectively. The 28-day strength meets the minimum requirement for structural lightweight concrete as specified in SNI 03–0349–1989, making it suitable for use as prefabricated wall infill panels. Splitting tensile strength, flexural strength, modulus of elasticity, and Poisson’s ratio were measured following relevant SNI standards to provide a comprehensive characterization of the foamed concrete’s mechanical behavior. The testing and characterization ensured that the selected foamed concrete mix not only met structural requirements but also provided reliable and reproducible data for assessing the panels’ behaviour under lateral cyclic loading, simulating earthquake conditions representative of Indonesian seismic scenarios.

For comparison, conventional fired red clay bricks generally exhibit compressive strengths in the range of 10–25 MPa, depending on manufacturing quality and firing temperature, while lightweight bricks such as autoclaved aerated concrete (AAC) blocks typically range between 3 and 8 MPa. Although the compressive strength of the foamed concrete panels (7.38 MPa at 28 days) is lower than that of traditional red brick, it is comparable to that of lightweight masonry units commonly used as non-structural infill. Considering that the panels in this study function as infill elements rather than primary load-bearing components, the achieved strength is deemed adequate to contribute to lateral stiffness while maintaining reduced structural mass. This balance between moderate compressive strength and lower density is particularly beneficial for seismic performance, where excessive mass may increase inertia forces during earthquake excitation.

To provide visual documentation of the material preparation and testing stages, photographs of the foamed concrete test specimens are presented in Figure 2. The images illustrate the specimen geometry, surface texture, and physical condition prior to mechanical testing, including compressive and splitting tensile tests. The visual appearance confirms the uniform pore distribution and absence of major surface defects, supporting the consistency of the production process. These photographs also demonstrate the dimensional stability and integrity of the prefabricated panels during curing and handling. The inclusion of Figure 2 therefore complements the quantitative test results by offering qualitative verification of specimen quality and fabrication control.

FIGURE 2

2.2.2 Reinforced concrete frame

The reinforced concrete (RC) frames were constructed using concrete designed for a target compressive strength of 25 MPa. Reinforcement for the tie beams consisted of plain steel bars, with longitudinal bars of 10 mm diameter and transverse bars of 8 mm diameter. Columns were reinforced with 13 mm deformed steel bars longitudinally and 8 mm plain steel bars transversely. Tensile testing of all reinforcement bars confirmed that the yield strength requirements were satisfied in accordance with SNI 2052:2017. The measured yield strengths were 473.7 MPa for D13 bars, 469.8 MPa for Ø10 bars, and 377.9 MPa for Ø8 bars, demonstrating that all bars met the classification and performance standards for structural applications.

Mechanical characterization included evaluation of modulus of elasticity, tensile strength, and bond behaviour, ensuring that the reinforcement properties accurately reflected real-world conditions. This detailed material characterization provided a reliable basis for both experimental testing and numerical simulations, allowing precise assessment of the RC frames’ seismic response when combined with foamed concrete infill panels.

2.2.3 Material-environment interaction

The interaction between the foamed concrete infill panels and environmental conditions was carefully monitored to capture the influence of temperature, relative humidity, and moisture variations on material performance. During both curing and subsequent testing, environmental parameters were recorded to evaluate their effect on dimensional stability, shrinkage behaviour, and potential microcracking. Observations indicated that prolonged exposure to high humidity and temperature fluctuations, typical of Indonesia’s tropical climate, could affect the moisture content and induce minor volumetric changes in the panels, which in turn may influence the structural response of RC frames under lateral cyclic loading (Neville et al., 2015; Monteiro et al., 2014).

To account for these environmental effects, the measured variations in density, dimensional change, and water absorption were incorporated into the numerical and predictive models. This allowed the simulations to reflect realistic service conditions, providing a more accurate assessment of the panels’ performance under seismic events. Such integration of material-environment interaction is critical, as previous studies have shown that environmental factors significantly impact the stiffness, energy dissipation capacity, and cracking patterns of lightweight concrete infills (Morandi et al., 2018; Luca et al., 2018). By including these effects, the study ensures that the evaluation of prefabricated foamed concrete panels in RC frames is both comprehensive and representative of real-world conditions, enhancing the reliability of predictive models for earthquake-resistant design.

2.3 Structural testing

2.3.1 Test specimen

Two full-scale specimens were prepared as part of the experimental program to evaluate the seismic performance of reinforced concrete (RC) frames with and without prefabricated foamed concrete (FC) infill walls. The first specimen consisted of a portal RC frame serving as the reference model, while the second specimen was a portal RC frame infilled with prefabricated plain foamed concrete blocks. The geometric configuration, reinforcement detailing, and overall layout of both specimens are illustrated in Figure 3.

FIGURE 3

The specimens were constructed at a 1:1 scale to represent typical low-rise residential buildings commonly found in Indonesia. A prevalent RC frame configuration used in Indonesian housing construction was adopted to reflect standard material properties, reinforcement details, and construction practices. This approach ensured that the experimental results are representative of non-engineered and semi-engineered RC buildings widely used in seismic-prone regions of the country.

For the infilled specimen, the foamed concrete wall was constructed prior to testing, with careful determination of block arrangement to ensure uniformity across each row. Prefabricated foamed concrete blocks with dimensions of 80 mm (width) × 200 mm (height) × 700 mm (length) were used to form the infill wall. Mortar joints at the head, bed, and side interfaces were maintained at an approximate thickness of 15 mm to replicate common construction practices. The RC columns and beams constituting the lateral load-resisting system were designed in accordance with SNI 2847–2019 provisions, ensuring structural consistency between the bare and infilled specimens.

This specimen configuration enabled direct comparison of seismic behaviour between bare and infilled RC frames, allowing for clear evaluation of the contribution of prefabricated foamed concrete infill walls to lateral strength, stiffness, and energy dissipation under cyclic loading conditions.

2.3.2 Lateral cyclic loading test

In-plane lateral cyclic loading tests were conducted to evaluate the seismic response of the portal RC frame and the RC frame infilled with prefabricated foamed concrete blocks. The overall test setup for both specimens is illustrated in Figure 4. The experimental configuration was designed to replicate quasi-static earthquake loading conditions while allowing detailed observation of global structural response and local damage mechanisms.

FIGURE 4

The cyclic loading protocol followed ASTM E2126-02a, adopting Method B (ISO 16670 protocol). Among the three cyclic loading methods defined in ASTM E2126-02a—namely Method A (Sequential-Phased Displacement), Method B (ISO 16670), and Method C (CUREE Basic Loading Protocol)—Method B was selected due to its suitability for evaluating stiffness degradation, strength deterioration, and energy dissipation characteristics of structural wall and frame systems under increasing displacement demands.

The loading procedure employed a displacement-controlled scheme, in which cyclic displacement amplitudes were grouped into sequential phases with progressively increasing deformation levels. The ISO 16670 loading history consists of two displacement patterns. The first pattern includes five fully reversed single cycles applied at displacement amplitudes of 1.25%, 2.5%, 5%, 7.5%, and 10% of the ultimate displacement (Δm). The second pattern consists of multiple phases, each comprising three fully reversed cycles, applied at displacement levels of 20%, 40%, 60%, 80%, 100%, and subsequently increased in 20% increments until specimen failure. Figure 5 shows the loading amplitude according to ASTM E2126-02a.

FIGURE 5

In this study, the ultimate displacement (Δm) was determined based on seismic code provisions relating building height to allowable displacement, rather than relying on monotonic test results. Referring to SNI 1726–2019, the ultimate displacement was defined as 2% of the specimen height. For the tested specimens with a height of 2.0 m, this corresponded to an ultimate displacement of 40 mm. The complete displacement loading sequence was established based on this value.

The actuator displacement was applied at a constant rate within the limits prescribed by ASTM E2126-02a to minimize inertial effects. A quasi-static loading rate of 0.5 mm/s was adopted, ensuring stable load application and accurate measurement of force–displacement behaviour. Cyclic loading was continued until the applied lateral load dropped by more than 20% of the peak load, indicating the attainment of a failure limit state.

The experimental setup involved a horizontal hydraulic actuator mounted on a reaction wall at a height of 2000 mm from the strong floor. The specimens were positioned at a distance of 2500 mm from the reaction wall and rigidly anchored to the strong floor through the foundation beam using steel rods. A loading plate connected to the actuator was installed at the top beam of the specimen using high-strength anchor bolts to ensure uniform load transfer.

Displacements were measured using linear variable displacement transducers (LVDTs) installed on both the front and rear faces of the specimens, aligned with the loading direction. Additional instrumentation included strain gauges attached to selected reinforcing bars and concrete surfaces to monitor strain development during cyclic loading. All sensors were connected to a data acquisition system with continuous recording throughout the test.

During testing, visual observations were systematically performed to identify the initiation of first cracking, crack propagation patterns, crack distribution, and crack width development. These observations were documented directly on the specimen surfaces along with the corresponding load and displacement levels. The collected experimental data—including lateral load capacity, displacement response, stiffness degradation, energy dissipation, strain development, and failure modes—were subsequently analysed to quantify the seismic performance of both bare and infilled RC frame systems.

2.3.3 Instrumentation and data acquisition

The test specimens were instrumented with a comprehensive measurement system consisting of linear variable displacement transducers (LVDTs) and strain gauges (SGs) to capture global deformation characteristics and local material responses during cyclic loading. The primary objective of the instrumentation was to accurately measure lateral displacement, rotation, and strain development in both the reinforced concrete frame and the infill system under increasing displacement demands.

A total of eight LVDTs were installed at strategic locations on each specimen, as illustrated in Figure 3, to monitor displacement profiles and base rotation along the height of the structure. Two LVDTs with a capacity of 200 mm and 100 mm were positioned at the top and mid-height of the specimen, respectively, to record global lateral displacements. Additional LVDTs with smaller measurement ranges were placed near the lower portion of the structure, approximately 0.22 m above the strong floor, to capture local displacement behaviour and potential slip at the base.

To ensure accurate measurement of bidirectional displacement and minimize potential measurement bias due to out-of-plane movement, paired LVDTs were installed on opposite sides of the specimen at corresponding elevations. These paired sensors allowed verification of symmetry in deformation and improved the reliability of displacement data. Furthermore, two short-stroke LVDTs were installed near the base of the columns to measure rotational deformation, enabling the assessment of base rotation and its contribution to overall lateral displacement.

Strain gauges were attached to selected reinforcing steel bars and concrete surfaces to monitor strain development during cyclic loading. Steel strain gauges were installed on longitudinal reinforcement in critical regions, particularly near column bases and beam–column joints, where high stress concentrations were expected. Concrete strain gauges were placed on column faces to capture compressive and tensile strain evolution associated with cracking and crushing mechanisms.

All sensors were connected to a digital data acquisition system capable of continuous recording throughout the test duration. Data were recorded at a sampling rate sufficient to capture the cyclic response under quasi-static loading conditions, ensuring accurate representation of load–displacement hysteresis, strain progression, and stiffness degradation. The synchronized acquisition of displacement and strain data provided a robust experimental database for subsequent analysis and for validation of the numerical simulation results.

2.4 Numerical simulations

Numerical simulations were conducted using the finite element software Abaqus to replicate the seismic response of reinforced concrete (RC) frames with prefabricated foamed concrete (FC) infill walls observed in the experimental program. The numerical modelling aimed to capture the nonlinear material behaviour, interaction between the RC frame and infill panels, and global structural response under cyclic lateral loading representative of earthquake actions in Indonesia. The modelling strategy was developed to ensure close consistency with the experimental setup, material properties, and boundary conditions.

The reinforced concrete frame was modelled using three-dimensional beam elements for columns and beams, with material properties defined based on the experimental characterization of concrete and reinforcing steel. Concrete behaviour was represented using a nonlinear constitutive model incorporating cracking and crushing mechanisms, while reinforcing steel was modelled as an elastic–plastic material with isotropic hardening. The mechanical properties of reinforcement bars, including yield strength and ultimate strength, were defined in accordance with tensile test results and relevant Indonesian standards.

The prefabricated foamed concrete infill walls were modelled as solid elements and assigned a concrete damage plasticity (CDP) model to represent their nonlinear behaviour under cyclic loading. Model parameters, including elastic modulus, compressive strength, tensile strength, and damage evolution, were calibrated directly from the experimental results of foamed concrete material testing at 28 days. This approach ensured that stiffness degradation, cracking initiation, and post-peak softening observed experimentally were realistically captured in the numerical simulations.

Interaction between the RC frame and the foamed concrete infill walls was simulated using surface-to-surface contact definitions, allowing for separation and sliding along the interface. The mortar joints between prefabricated blocks were implicitly represented through adjusted interface properties, accounting for reduced stiffness and potential slip without explicitly modelling each joint. This simplification was adopted to balance computational efficiency while preserving the dominant interaction mechanisms governing the structural response.

Boundary conditions and loading protocols in the numerical model were defined to replicate the laboratory testing conditions. The base of the RC frame was fully restrained to simulate the strong-floor anchorage, while lateral cyclic displacement-controlled loading was applied at the beam level following the same loading history used in the experimental program. This approach enabled direct comparison between numerical and experimental responses, particularly in terms of load–displacement behaviour, stiffness degradation, and energy dissipation.

A series of sensitivity analyses was performed to investigate the influence of key parameters, including foamed concrete stiffness, density variation, and frame–infill interaction properties, on the overall seismic performance of the system. These analyses provided insights into the relative contribution of the infill wall to lateral strength and stiffness, as well as its role in modifying failure mechanisms of the RC frame.

Validation of the numerical model was carried out by comparing the simulated hysteretic load–displacement curves with experimental results. Good agreement between numerical and experimental responses confirmed the capability of the developed finite element model to capture the essential nonlinear behaviour of RC frames with foamed concrete infill walls. The validated model was subsequently used to support interpretation of experimental findings and to extend the analysis beyond the tested configurations.

2.5 Predictive modelling and performance assessment

A data-driven predictive modelling framework was developed to evaluate the seismic performance of reinforced concrete (RC) frames infilled with prefabricated foamed concrete (FC) panels by integrating experimental observations and numerical simulation results. The objective of this framework was to quantify the influence of material characteristics, structural configuration, and environmental conditions on key performance indicators under lateral cyclic loading. This approach extends the experimental findings by enabling performance prediction under varying design and environmental scenarios representative of tropical regions such as Indonesia.

The input dataset was constructed from a combination of laboratory test results and validated finite element simulations. Material-related input variables included foamed concrete compressive strength, density, elastic modulus, and porosity, which were shown in previous sections to significantly affect stiffness degradation and cracking behaviour. Structural parameters such as reinforcement ratio, frame geometry, and infill wall configuration were incorporated to represent typical low-rise RC buildings commonly constructed in Indonesia. Environmental variables, particularly temperature and relative humidity during curing and testing, were included to capture their influence on moisture-related deformation and material performance.

Supervised machine learning techniques were employed to establish predictive relationships between the input parameters and structural response outputs. Feedforward artificial neural networks (ANNs) were used to model nonlinear correlations between material–structural features and global seismic response indicators, including lateral load capacity, initial stiffness, energy dissipation, and ultimate displacement. In addition, recurrent neural network (RNN) architectures were explored to account for the sequential nature of cyclic loading and cumulative damage effects observed during repeated displacement cycles.

Model training was performed using a normalized dataset to ensure numerical stability and reduce bias associated with differing parameter scales. The dataset was divided into training, validation, and testing subsets to prevent overfitting and to assess generalization capability. Hyperparameters such as learning rate, number of hidden layers, and neuron count were optimized through iterative tuning to achieve robust predictive performance.

The accuracy of the predictive models was evaluated using statistical performance indicators, including root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R²). These metrics provided quantitative measures of the models’ ability to reproduce experimentally observed and numerically simulated responses. High R² values and low prediction errors indicated that the developed models were capable of reliably estimating seismic performance parameters based on combined material, structural, and environmental inputs.

To enhance interpretability and support engineering decision-making, SHapley Additive exPlanations (SHAP) analysis was applied to the trained models. SHAP values were used to quantify the relative contribution of each input variable to the predicted structural performance. The results demonstrated that foamed concrete density and compressive strength were among the most influential parameters affecting lateral stiffness and load capacity, while environmental factors such as humidity exhibited a notable impact on displacement capacity and damage progression.

The integration of machine learning–based predictive modelling with experimental and numerical approaches provides a comprehensive framework for performance assessment of RC frames with foamed concrete infill walls. This framework enables rapid evaluation of alternative material designs and environmental conditions, supporting performance-based seismic design and the development of resilient, lightweight infill systems suitable for tropical construction contexts.

2.6 Ethical and data management considerations

All experimental procedures conducted in this study adhered to institutional laboratory safety regulations and relevant national standards governing structural testing. The experimental program involved full-scale reinforced concrete (RC) frame specimens and prefabricated foamed concrete infill panels, without the involvement of human participants or animal subjects. Consequently, formal ethical approval related to human or biomedical research was not required. Nevertheless, all testing activities were performed by trained personnel using calibrated equipment to ensure safety, data integrity, and procedural consistency.

Material handling, specimen fabrication, and lateral cyclic loading tests were carried out in controlled laboratory environments in accordance with applicable Indonesian standards (SNI) and international testing protocols. Particular attention was given to structural stability during high-displacement cyclic loading to prevent unintended specimen collapse and to ensure the safety of laboratory staff. These precautions ensured that the experimental results were obtained under ethically responsible and professionally acceptable research practices.

Data management in this research was structured in accordance with the FAIR principles, emphasizing findability, accessibility, interoperability, and reusability of research outputs. Experimental data, including load–displacement records, LVDT measurements, strain gauge readings, and crack observation logs, were systematically recorded, labelled, and stored in digital formats suitable for long-term archiving. This approach supports transparency and enables independent verification of the reported results.

Numerical simulation data, including finite element models developed in Abaqus, material constitutive parameters, boundary condition definitions, and loading protocols, were documented in detail to facilitate model reproducibility. Calibration procedures linking experimental observations with numerical responses were preserved to ensure traceability between physical testing and computational analysis. Similarly, datasets used for predictive modelling were curated to maintain consistency between experimental, numerical, and machine learning components of the study.

All processed datasets, numerical models, and predictive modelling scripts were archived and made available as supplementary materials, in accordance with the open-access and data-sharing policies of Frontiers in Built Environment. Where applicable, metadata descriptions were provided to clarify variable definitions, units, and data provenance. This data management strategy supports the broader research community by enabling reuse of the datasets for comparative studies, methodological benchmarking, and future research on seismic performance of RC frames with lightweight infill systems.

3 Results and discussion

3.1 Overall seismic response of test specimens

The overall seismic response of the test specimens was evaluated through lateral cyclic loading tests, focusing on hysteresis behaviour, lateral load-carrying capacity, stiffness evolution, and observed failure modes. Two specimens were investigated, namely, a portal reinforced concrete (RC) frame (SW1) and an RC frame infilled with prefabricated foamed concrete (FC) panels (SW2), representing typical low-rise RC structures commonly found in earthquake-prone regions of Indonesia. The comparative results provide a clear assessment of the contribution of prefabricated foamed concrete infill walls to seismic performance enhancement.

The load–displacement hysteresis responses of both specimens are presented in Figure 6, while the corresponding maximum lateral forces are summarized in Tables 1, 2. Throughout the cyclic loading history up to 120% of the ultimate displacement (Δm), both specimens exhibited stable cyclic behaviour under alternating compression and tension directions, and no sudden brittle collapse was observed. However, distinct differences in hysteresis characteristics were evident. The bare RC frame specimen (SW1) developed relatively narrow hysteresis loops, indicating limited energy dissipation capacity and lower lateral stiffness. In contrast, the infilled RC frame specimen (SW2) exhibited wider and fuller hysteresis loops, particularly at higher displacement amplitudes, reflecting enhanced energy dissipation and improved cyclic stability due to the interaction between the RC frame and the prefabricated foamed concrete infill.

FIGURE 6

A significant increase in lateral load-carrying capacity was observed in the infilled specimen compared to the bare RC frame. As summarized in Table 2, specimen SW1 attained maximum lateral forces of 32.68 kN in the positive loading direction and 32.05 kN in the negative direction. In contrast, specimen SW2 exhibited substantially higher maximum forces of 49.41 kN and 56.13 kN under positive and negative loading, respectively. These results indicate an increase in lateral strength of approximately 51.19% in the positive direction and 75.13% in the negative direction, clearly demonstrating the substantial contribution of the prefabricated foamed concrete infill to resisting seismic lateral loads. The enhanced lateral capacity reflects the effective interaction between the infill panels and the surrounding reinforced concrete frame, which enables a more efficient distribution of seismic forces within the structural system.

With increasing cyclic displacement, both specimens showed a gradual transition from linear elastic behaviour to nonlinear inelastic response, as evidenced by the change in slope of the load–displacement curves in Figure 5. This transition was more pronounced in the infilled specimen (SW2), which initially exhibited higher lateral stiffness due to the presence of the infill panels. As cracking and inelastic deformation developed, stiffness degradation occurred in a controlled manner, indicating a stable post-yield behaviour. The gradual reduction in stiffness suggests that the prefabricated foamed concrete infill promotes a more distributed damage mechanism, allowing the structure to sustain larger deformations without abrupt loss of strength.

Overall, the experimental observations confirm that SW1 exhibited a flexure-dominated failure mechanism characterized by beam-end plastic hinging and vertical flexural cracks, whereas SW2 demonstrated a modified failure response governed by shear action and infill–frame interaction effects. The presence of the foamed concrete infill significantly altered the internal force distribution, leading to increased lateral stiffness and the development of diagonal cracking associated with shear transfer mechanisms.

Regarding failure characteristics, both specimens ultimately exhibited lateral force resistance with distinct failure mechanisms, as summarized in Table 1. For the bare RC frame specimen (SW1), the maximum positive and negative lateral forces reached 32.68 kN and 32.05 kN, respectively, with the governing failure mode dominated by flexural cracking at the beam–column joints. Damage was primarily concentrated in the RC members, particularly at critical regions such as beam–column joints and column bases, which aligns with the typical behavior of non-engineered RC frames observed in past earthquake damage surveys in Indonesia. In contrast, the RC frame with prefabricated foamed concrete infill panels (SW2) exhibited significantly higher lateral strength, with maximum positive and negative forces of 49.41 kN and 56.13 kN, respectively. The failure mechanism in SW2 was governed by a combined shear mode influenced by infill–frame interaction, resulting in a more distributed cracking pattern involving both the RC frame and the foamed concrete panels. Notably, no premature crushing or out-of-plane instability of the prefabricated foamed concrete panels was observed up to the maximum displacement amplitude of 100% Δm, indicating adequate deformation compatibility between the infill and the surrounding frame.

The observed seismic response of the infilled RC frame aligns well with findings from previous experimental studies on infilled RC systems, which reported significant improvements in lateral strength, stiffness, and energy dissipation due to the presence of infill walls (e.g., Lata et al., 2022; Yu et al., 2023). However, unlike conventional burnt clay brick masonry infills—often associated with brittle cracking, stress concentration, and sudden strength degradation—the prefabricated foamed concrete infill used in this study exhibited more stable hysteresis behaviour and controlled damage progression. The reduced self-weight and more uniform material properties of foamed concrete likely contributed to mitigating adverse frame–infill interaction effects commonly reported in traditional masonry systems.

Overall, the results demonstrate that the integration of prefabricated foamed concrete infill walls significantly enhances the seismic performance of RC frame structures by increasing lateral load capacity, improving hysteretic energy dissipation, and promoting more ductile and stable inelastic behaviour. These characteristics are particularly advantageous for low-rise buildings in developing, earthquake-prone regions such as Indonesia, where improving seismic resilience in a practical and scalable manner remains a critical challenge.

3.2 Effect of prefabricated foamed concrete infill on structural behavior

The experimental results clearly demonstrate that the incorporation of prefabricated foamed concrete infill significantly enhances the lateral load-carrying capacity of the RC frame system. Compared to the bare frame specimen (SW1), the infilled specimen (SW2) exhibited markedly higher maximum lateral strengths in both loading directions. While SW1 reached peak lateral forces of 32.68 kN and 32.05 kN under positive and negative loading, respectively, SW2 achieved substantially higher values of 49.41 kN in the positive direction and 56.13 kN in the negative direction. These results correspond to strength increases of approximately 51.19% and 75.13%, indicating a pronounced contribution of the prefabricated foamed concrete infill to lateral resistance.

The observed strength enhancement can be attributed to the composite action between the RC frame and the infill panel, which effectively alters the lateral load transfer mechanism. In the infilled configuration, the foamed concrete panel participates in resisting lateral loads through diagonal compression strut action, thereby reducing the demand on the RC frame members. This behavior is consistent with findings reported in previous studies on masonry and lightweight infilled frames, where the presence of infill panels significantly increased initial stiffness and peak lateral strength compared to bare frames (e.g., Pallarés et al., 2021; Martinelli et al., 2023; Roosta and Liu, 2021). However, unlike conventional masonry infills, the use of foamed concrete offers the advantage of reduced self-weight, which is beneficial for seismic applications.

Notably, the infilled specimen not only showed higher strength but also exhibited an asymmetric increase between the positive and negative loading directions, with a more pronounced improvement under negative loading. This suggests that the interaction between the infill panel and the surrounding RC frame may be influenced by construction details, boundary conditions, or load reversal effects, a phenomenon also reported in earlier experimental investigations on infilled frame systems. Such directional sensitivity highlights the importance of considering bidirectional loading behavior when evaluating the seismic performance of infilled RC frames.

From a broader seismic perspective, the results are particularly relevant in the context of recent destructive earthquakes in Indonesia, such as the 2018 Palu earthquake and the 2021 Mamuju earthquake. Post-earthquake field observations from these events revealed widespread damage and collapse of RC frame buildings with non-engineered or poorly detailed infill walls, often due to inadequate lateral resistance and brittle failure mechanisms. Although the present study does not aim to replicate these complex real-world failures, the significant strength gains observed in the infilled specimen suggest that properly designed and integrated prefabricated foamed concrete infill panels have the potential to enhance lateral load resistance and improve overall structural robustness under seismic loading.

Importantly, the use of prefabricated foamed concrete infill represents a controlled and engineered alternative to traditional unreinforced masonry infills commonly used in Indonesian construction practice. By improving lateral strength without substantially increasing seismic mass, this system may contribute to more predictable and resilient structural behavior during earthquakes. These findings provide a meaningful experimental basis for further discussion on the role of engineered infill systems in seismic design and retrofit strategies, particularly for RC frame buildings in high seismic regions such as Indonesia.

In the context of Indonesian construction practice, RC frame buildings are predominantly designed as bare frames, while infill walls are often treated as non-structural components and constructed without explicit consideration of their interaction with the structural system. Field evidence from recent earthquakes in Palu and Mamuju indicates that such practice frequently leads to unfavorable seismic performance, including soft-story mechanisms, brittle infill failure, and concentration of damage at column–infill interfaces. The experimental comparison between the bare frame (SW1) and the prefabricated foamed concrete infilled frame (SW2) provides an important insight into how an engineered infill system can fundamentally modify the lateral behavior of RC frames. The substantial increase in lateral strength observed in SW2 suggests that, when properly designed and integrated, infill panels can contribute positively to seismic resistance rather than acting as unintended sources of vulnerability. Moreover, the use of lightweight prefabricated foamed concrete infill addresses a key limitation of conventional masonry infills by limiting additional seismic mass while still enhancing lateral capacity. This finding is particularly relevant for low-to mid-rise RC buildings, which dominate the Indonesian building stock and account for a large proportion of earthquake-induced damage. Therefore, the present results highlight the need to reconsider the role of infill walls in both the design of new RC buildings and the retrofit of existing structures in seismic regions of Indonesia.

3.3 Stiffness degradation and energy dissipation

The stiffness degradation and energy dissipation behaviour of the test specimens were evaluated based on the hysteretic load–displacement response and the corresponding envelope curves presented in Figure 5, as well as the peak lateral strength data summarized in Tables 1, 2. These results provide critical insight into the nonlinear seismic performance of bare and infilled RC frame systems.

As observed from the envelope curves in Figure 5, both specimens exhibited an initial linear-elastic response followed by a progressive transition into nonlinear behaviour as displacement demand increased. However, the RC frame infilled with prefabricated foamed concrete (SW2) demonstrated a noticeably higher initial stiffness compared to the bare RC frame (SW1), as reflected by the steeper slope of the early-stage response. This behaviour indicates that the foamed concrete infill actively contributed to lateral load resistance by engaging composite action with the surrounding RC frame, effectively delaying the onset of stiffness degradation.

With increasing cyclic displacement, stiffness degradation became evident in both specimens, characterized by a reduction in the secant stiffness at larger deformation levels. Specimen SW1 experienced a more rapid stiffness degradation after entering the post-yield region, resulting in a flatter envelope curve and reduced lateral resistance capacity. In contrast, specimen SW2 maintained a more stable post-elastic response, with a gradual reduction in stiffness up to the maximum imposed displacement amplitude. This improved stiffness retention is consistent with the higher lateral load capacity reported in Table 2, where specimen SW2 achieved peak lateral strengths up to 75.13% greater than SW1 in the negative loading direction.

The enhanced stiffness characteristics of the infilled frame directly translated into superior energy dissipation capacity, as evidenced by the wider hysteresis loops observed throughout the cyclic loading history in Figure 5. The presence of the foamed concrete infill introduced additional energy dissipation mechanisms, including distributed cracking within the infill panel, frictional sliding at the infill–frame interface, and delayed yielding of the RC frame elements. Similar trends have been reported in previous experimental studies on masonry- and lightweight concrete-infilled RC frames, which consistently demonstrate higher equivalent damping and improved seismic performance relative to bare frames.

From the perspective of failure progression and crack pattern development, distinct behavioral differences were observed between the two specimens. The bare RC frame (SW1) exhibited early formation of flexural cracks at the beam–column joints, which then widened and concentrated in critical regions, indicating limited redistribution of inelastic demand. This behavior aligns with the governing failure mode of flexural cracking in the beam–column joints, as shown in Table 1, and corresponds to the relatively lower maximum lateral forces of 32.68 kN and 32.05 kN for positive and negative directions, respectively. In contrast, the RC frame with prefabricated foamed concrete infill (SW2) demonstrated a more distributed cracking pattern, with diagonal cracks initiating within the infill panels at moderate displacement levels and progressively spreading across both the panel and the surrounding frame. This interaction resulted in a combined shear-dominated failure mode, significantly higher maximum lateral forces of 49.41 kN and 56.13 kN, and more uniform damage distribution, effectively reducing stress concentration in the RC frame members and promoting a more ductile overall response. Notably, both specimens reached the maximum displacement amplitude before failure of 100% Δm without sudden strength loss, confirming a stable degradation process rather than brittle collapse. The results indicate that the presence of prefabricated foamed concrete infill enhances both strength and ductility through improved load sharing and deformation compatibility with the RC frame.

The observed stiffness degradation and energy dissipation characteristics have important implications when interpreted in the context of recent seismic events in Indonesia, particularly the Palu (2018) and Mamuju–Majene (2021) earthquakes. Post-earthquake reconnaissance reports from these events highlighted extensive damage and collapse of low-rise RC buildings with unreinforced masonry infill walls, primarily due to inadequate lateral stiffness, poor energy dissipation, and brittle failure of infill panels. While the present study does not aim to replicate these earthquakes directly, the experimental findings suggest that replacing conventional brittle masonry infill with prefabricated foamed concrete panels can improve seismic performance by enhancing stiffness retention, promoting distributed cracking, and increasing hysteretic energy dissipation under cyclic loading representative of seismic demands.

Overall, the results indicate that RC frames infilled with prefabricated foamed concrete panels exhibit more favourable stiffness degradation behaviour and superior energy dissipation capacity compared to bare RC frames. These characteristics are essential for improving seismic resilience of non-engineered and semi-engineered low-rise buildings commonly found in Indonesia, particularly in high seismic hazard regions.

3.4 Damage patterns and failure mechanisms

The damage patterns and failure mechanisms observed during the cyclic lateral loading tests provide important insight into the seismic behaviour of both the bare RC frame (SW1) and the RC frame infilled with prefabricated foamed concrete panels (SW2). A comprehensive summary of damage progression and corresponding damage state levels for each specimen is presented in Table 3 for the bare RC frame and Table 4 for the infilled RC frame.

For the bare RC frame specimen, damage initiation and progression were predominantly governed by flexural behaviour. As summarized in Table 3, very light flexural cracking first appeared at a displacement amplitude of 1.25% Δm, corresponding to the Operational Limit State (OLS), with negligible impact on structural performance. As the displacement demand increased to 2.50–10.00% Δm, flexural cracks propagated and widened along the beam–column joints and column faces, indicating the transition toward slight to moderate damage while still remaining within the OLS range.

At displacement amplitudes between 20.00% Δm and 80.00%Δm, the specimen entered the Damage Limit State (DLS). During this stage, flexural cracking intensified, accompanied by localized spalling of the concrete cover at critical regions, particularly near the column bases. Although damage was visible, the structural system retained residual strength, and repair or strengthening measures were still considered technically feasible. At the Ultimate Limit State (ULS), corresponding to 100.00% Δm and beyond, extensive spalling and severe flexural cracking dominated the failure mechanism, indicating significant loss of load-carrying capacity and rendering repair or retrofit measures uneconomical.

This flexure-dominated failure mode is consistent with previous experimental studies on bare RC frames subjected to cyclic loading, where damage tends to concentrate at beam–column joints due to limited energy dissipation mechanisms and poor redistribution of inelastic demands.

In contrast, the RC frame infilled with prefabricated foamed concrete panels exhibited a markedly different damage evolution, as summarized in Table 4. At low displacement amplitudes (1.25–2.50% Δm), no visible damage was observed, indicating that the infill panels contributed to enhanced initial stiffness and delayed crack initiation. Minor shear cracking within the foamed concrete infill began to develop at around 5.00% Δm, but these cracks remained narrow and widely distributed, corresponding to the OLS condition.

As the displacement amplitude increased to 7.50–60.00% Δm, the specimen entered the DLS range. During this phase, shear cracks developed in the RC frame elements, while the infill panels exhibited multiple failure mechanisms, including diagonal cracking, sliding shear along mortar joints, diagonal compression struts, and localized corner crushing. Importantly, these damage patterns promoted distributed cracking across the infill rather than concentrated damage in the RC frame, allowing the system to dissipate seismic energy more effectively. Repair at this stage was still considered feasible and economical.

At higher displacement amplitudes (80.00–120.00% Δm), corresponding to the ULS, damage became severe. Widespread crushing and diagonal cracking were observed in the foamed concrete infill, while the RC frame exhibited pronounced shear cracking and extensive spalling of the concrete cover. At this stage, both the infill and frame components were deemed irreparable. The observed failure mechanisms closely resemble those reported in masonry-infilled RC frames, suggesting that prefabricated foamed concrete infill behaves in a mechanically comparable manner but with improved ductility and damage distribution.

The influence of the prefabricated foamed concrete infill on lateral load resistance is quantitatively demonstrated by the ratio of lateral forces between the infilled and bare RC frames at different damage states, as presented in Table 5. While the infilled specimen showed slightly lower lateral force at the OLS compared to the bare frame, its performance significantly exceeded that of the bare frame at higher damage states. At the DLS and ULS levels, the lateral load capacity of the infilled RC frame was approximately 1.56 and 1.51 times greater than that of the bare RC frame, respectively.

These findings are consistent with previous research on infilled RC frames, which has shown that properly detailed infill systems can substantially enhance lateral strength and delay structural degradation under seismic loading. The present results further demonstrate that prefabricated foamed concrete infill panels can provide comparable, if not superior, seismic performance to traditional masonry infill while offering advantages in terms of weight reduction and controlled damage progression.

To provide clearer visual evidence of the observed damage mechanisms, photographs of both specimens after completion of the cyclic loading tests have been added as Figure 7. These post-test images illustrate the final crack distribution, spalling patterns, and overall failure configuration of SW1 and SW2, thereby complementing the quantitative damage descriptions presented in Tables 3–5. The visual comparison highlights the flexure-dominated failure and concentrated damage at beam–column joints in the bare RC frame (SW1), whereas the infilled specimen (SW2) exhibits more distributed cracking, diagonal compression patterns within the foamed concrete panel, and interaction effects between the infill and surrounding frame. The inclusion of these photographs enhances the transparency of the experimental observations and supports the interpretation of the distinct failure mechanisms discussed in this section.

FIGURE 7

3.5 Numerical simulation results and validation

Numerical simulations were conducted to reproduce the seismic response of both the bare RC frame (SW1) and the RC frame infilled with prefabricated foamed concrete panels (SW2) under lateral cyclic loading. The finite element (FE) models were calibrated using experimentally obtained material properties and loading protocols, as described in the Materials and Methods section. The validity of the numerical approach was assessed through direct comparison with experimental results in terms of lateral load–displacement response, stiffness degradation trends, damage progression, and failure modes.

The comparison between numerical and experimental load–displacement responses indicate a good level of agreement for both specimens. The FE models successfully captured the overall shape of the hysteresis envelopes previously discussed in the experimental results section, including the initial linear elastic response, transition to nonlinear behaviour, and post-yield softening. In particular, the numerically predicted peak lateral strengths closely matched the experimentally measured values summarized earlier in Tables 1, 2, with deviations remaining within an acceptable range for nonlinear cyclic simulations.

For the bare RC frame (SW1), the FE model slightly underestimated the peak lateral load at higher drift levels. This discrepancy is primarily attributed to idealized assumptions regarding concrete cracking and bond–slip behaviour between reinforcement and concrete, which tend to reduce stiffness and strength in numerical models. Nevertheless, the predicted peak load and corresponding displacement were consistent with the experimental response trend and failure progression dominated by flexural cracking.

In contrast, the FE model for the infilled RC frame (SW2) demonstrated a closer match with experimental results, particularly in terms of peak lateral load and stiffness enhancement due to the presence of the prefabricated foamed concrete infill. The numerical simulation effectively reproduced the strength increase of approximately 50%–75% observed experimentally when compared to the bare RC frame, confirming the significant contribution of the infill panels to lateral resistance.

The numerical simulations also captured the stiffness degradation behaviour observed experimentally. The initial stiffness of the infilled frame was significantly higher than that of the bare frame, consistent with the experimental envelope curves discussed in the previous sections. As cyclic displacement increased, the FE model showed progressive stiffness reduction due to cracking, crushing, and sliding mechanisms within the foamed concrete infill and the RC frame elements.

Importantly, the numerical damage contours corresponded well with experimentally observed crack patterns illustrated in Tables 3, 4. For SW1, flexural damage localized at beam–column joints and column bases dominated the response, while for SW2, damage was more distributed, involving diagonal compression struts, shear cracking, and localized crushing at the corners of the foamed concrete infill. This consistency between numerical damage indicators and experimental crack observations supports the reliability of the adopted constitutive models.

To further evaluate the accuracy of the numerical simulations, a quantitative comparison between experimental and numerical results is presented in Table 6, focusing on key response parameters. The relative errors in peak lateral load and displacement remained within typical ranges reported in previous studies on infilled RC frames subjected to cyclic loading.

The relatively small discrepancies confirm that the FE models provide a reliable approximation of the experimental behavior, particularly in capturing the relative performance differences between bare and infilled RC frames.

The level of agreement achieved in this study is comparable to, or better than, that reported in previous numerical–experimental investigations of infilled RC frames. Prior studies on masonry-infilled frames have reported deviations in peak strength ranging from 5% to 15%, largely due to uncertainties in modelling infill–frame interaction, material heterogeneity, and interface behaviour. The closer agreement observed in the present study suggests that the simplified yet carefully calibrated modelling strategy adopted for the prefabricated foamed concrete infill is adequate for capturing global seismic response.

Furthermore, the ability of the numerical model to reproduce not only peak strength but also stiffness degradation and damage progression reinforces its applicability for parametric studies and performance-based seismic assessment. This is particularly relevant for evaluating alternative infill configurations and material properties without the need for extensive experimental campaigns.

The validated numerical framework provides a robust basis for extending the analysis beyond the tested specimens. Given its consistency with experimental observations, the FE model can be used to explore the influence of infill density, panel thickness, interface conditions, and boundary constraints on seismic performance. In the context of Indonesian seismicity, such validated numerical tools are valuable for assessing non-engineered buildings and developing retrofit strategies that incorporate lightweight prefabricated infill systems.

Overall, the close correspondence between numerical simulations and experimental results confirms the reliability of the adopted FE modelling approach and supports its use as a complementary tool for seismic performance evaluation of RC frames with prefabricated foamed concrete infill walls.

3.6 Implications for seismic design in Indonesia (non-engineered low-rise RC frames)

The experimental and numerical findings of this study provide important insights into the seismic performance of low-rise reinforced concrete (RC) frame buildings in Indonesia, particularly those classified as non-engineered structures. Such buildings constitute a substantial portion of the existing building stock and are commonly constructed without rigorous seismic design considerations. In many cases, masonry or lightweight infill walls are treated as non-structural components, despite their significant influence on global stiffness, strength, and energy dissipation under lateral loading (Barros et al., 2019; Reggia et al., 2020). The results presented in this study confirm that neglecting frame–infill interaction may lead to inaccurate estimations of seismic demand and damage progression.

The observed increase in initial lateral stiffness and strength due to the presence of precast foam concrete infill panels has direct implications for drift control in low-rise RC frames. Reduced inter story drift demands are particularly relevant in seismic regions of Indonesia, where excessive drift has been identified as a primary cause of severe damage and partial collapse during recent earthquakes. Post-earthquake reconnaissance following the 2018 Palu earthquake and the 2021 Mamuju earthquake revealed widespread damage in low-rise RC buildings associated with soft-storey mechanisms, inadequate lateral stiffness, and premature cracking of infill walls (Cilia et al., 2021; Heidarzadeh et al., 2019). The findings of this study suggest that properly detailed infill panels can contribute to mitigating excessive drift demands without significantly increasing structural mass.

From an energy-based perspective, the enhanced hysteretic energy dissipation observed in infilled frames indicates improved seismic resilience under cyclic loading. The equivalent viscous damping ratios derived from the experimental results are consistent with values reported in previous studies on RC frames with masonry or lightweight concrete infills (Cai and Su, 2019; Benavent-Climent et al., 2018; Dautaj et al., 2018). However, the progressive degradation of stiffness observed with increasing displacement amplitude highlights the importance of considering damage accumulation and degradation mechanisms in seismic design. The diagonal cracking and localized crushing observed in the infill panels demonstrate that infill elements tend to act as sacrificial components, dissipating energy while protecting the primary RC frame, provided that brittle failure modes are avoided.

The observed failure progression and crack patterns further emphasize the dual role of infill walls in non-engineered buildings. Initial cracking typically initiates along diagonal strut paths, followed by stiffness degradation and redistribution of internal forces to the RC frame. This behavior aligns with the widely accepted strut-and-tie mechanism for infilled frames and has been documented extensively in experimental and analytical studies (Kassem, 2016; Asteris et al., 2017; Yu et al., 2026). In the context of Indonesian construction practice, uncontrolled cracking or premature detachment of infill walls has frequently been observed during earthquakes, contributing to falling hazards and loss of lateral resistance. The use of precast foam concrete infill panels, when adequately connected to the surrounding frame, may offer a more predictable and controlled failure mechanism.

The numerical simulations conducted in this study further support the feasibility of incorporating infill effects into practical seismic analysis. The validated finite element models demonstrate that simplified equivalent strut representations can reasonably capture the global force–deformation response, stiffness degradation, and energy dissipation trends observed experimentally. This finding is consistent with earlier numerical investigations, which have shown that equivalent strut models provide an effective balance between accuracy and computational efficiency for low-rise RC infilled frames (Srechai et al., 2022; Xia et al., 2024). Such modelling approaches are particularly relevant for the seismic assessment and retrofitting of existing non-engineered buildings in Indonesia, where detailed material characterization is often unavailable.

From a regulatory and design perspective, the outcomes of this study highlight the need for a more explicit consideration of infill walls in seismic design guidelines. Current seismic design provisions typically idealize RC frames as bare systems, while the contribution of infill walls is either neglected or implicitly accounted for through global modification factors. This simplification may be unconservative for low-rise buildings, where infill walls dominate the lateral stiffness. The findings presented here support the adoption of “infill-aware” design and assessment approaches, as advocated in recent international guidelines and performance-based seismic frameworks.

Overall, the implications of this study for seismic design practice in Indonesia lie in promoting a more realistic representation of non-engineered low-rise RC buildings. While precast foam concrete infill panels should not be regarded as a standalone seismic solution, the results indicate that they can contribute positively to stiffness, energy dissipation, and damage control when properly detailed. These findings are consistent with international research trends and provide a scientific basis for developing simplified design and retrofit strategies that are compatible with local construction practices and seismic risk conditions in Indonesia.

4 Conclusion

This study investigated the seismic behaviour of low-rise reinforced concrete (RC) frames with and without prefabricated foam concrete infill panels through a combination of cyclic experimental testing and validated numerical simulations. The results provide a comprehensive understanding of stiffness degradation, energy dissipation capacity, damage progression, and failure mechanisms, as well as their implications for seismic design in earthquake-prone regions such as Indonesia. The integrated experimental–numerical approach allowed for a consistent interpretation of global structural response and local damage phenomena under increasing lateral displacement demands.

The experimental results demonstrate that the presence of prefabricated foam concrete infill significantly enhances the initial lateral stiffness and load-carrying capacity of RC frames compared to bare frame configurations. Infilled specimens exhibited higher peak lateral resistance and improved drift control at serviceability and damage limitation states. However, stiffness degradation was observed with increasing cyclic displacement amplitudes due to progressive cracking, crushing of infill material, and deterioration of frame–infill interaction. These findings confirm that while infill panels contribute positively to early-stage seismic performance, their influence diminishes as damage accumulates at higher deformation levels.

In terms of energy dissipation, infilled RC frames showed larger hysteretic loop areas and higher equivalent viscous damping ratios than bare RC frames, indicating superior energy dissipation capability under cyclic loading. The infill panels acted as effective energy-dissipating components during moderate deformation levels, delaying damage concentration in the primary RC frame. At larger drift levels, damage localization and stiffness degradation led to a gradual reduction in energy dissipation efficiency, highlighting the importance of considering degradation effects in seismic performance evaluation.

The observed damage patterns and failure mechanisms were strongly influenced by the presence of infill panels. Bare RC frames predominantly exhibited flexural cracking and progressive spalling at beam–column joints and plastic hinge regions, leading to severe damage at ultimate limit states. In contrast, infilled frames experienced a combination of diagonal cracking, corner crushing, sliding shear, and interface-related damage within the infill panels, followed by cracking and spalling in the surrounding RC frame. This failure progression suggests that infill panels function as sacrificial elements, absorbing seismic energy and redistributing internal forces, provided that brittle failure modes are adequately controlled.

The numerical simulations showed good agreement with the experimental results in terms of global force–displacement response, stiffness degradation trends, and energy dissipation characteristics. The validated finite element models were able to capture the essential features of frame–infill interaction and damage evolution using simplified modelling strategies. This confirms the applicability of such numerical approaches for seismic assessment and parametric studies of low-rise RC buildings, particularly in contexts where detailed material data are limited.

From a practical perspective, the findings of this study highlight the critical role of infill walls in the seismic response of non-engineered and low-rise RC buildings commonly found in Indonesia. The improved stiffness and energy dissipation observed in infilled frames suggest that properly detailed prefabricated foam concrete infill panels can contribute to enhanced seismic resilience, especially in reducing excessive drift demands observed during past earthquakes such as Palu and Mamuju. Nevertheless, the study also emphasizes that infill walls should not be treated as purely non-structural elements, as their degradation and failure significantly influence overall structural performance.

Overall, this research contributes to a more realistic understanding of the seismic behaviour of RC frame–infill systems and provides experimental evidence and numerical validation to support infill-aware seismic design and assessment approaches. The outcomes offer valuable insights for developing simplified design guidelines and retrofit strategies tailored to low-rise RC buildings in Indonesia and similar seismic regions. Future research should focus on connection detailing, out-of-plane behaviour, and full-scale system-level studies to further improve the reliability and applicability of infilled RC frame systems in seismic design practice.

References

• 1

  AliW.KhanS.ShahR.JamshaidH.NoorU. A. (2025). Seismic capacity assessment of lightweight precast concrete infill wall panels with textile waste core in reinforced concrete frame structures. Structures79, 109669. 10.1016/j.istruc.2025.109669

  • CrossRef
  • Google Scholar
  • View reference in article

• 2

  AsterisP. G.CavaleriL.Di TrapaniF.TsarisA. K. (2017). Numerical modelling of out-of-plane response of infilled frames: state of the art and future challenges for the equivalent strut macromodels. Eng. Struct.132, Preprint–122. 10.1016/j.engstruct.2016.10.012

  • CrossRef
  • Google Scholar
  • View reference in article

• 3

  BarrosM.CavacoE.NevesL.JúlioE. (2019). Effect of non-structural masonry brick infill walls on the robustness of a RC framed building severely damaged due to a landslide. Eng. Struct.180, 274–283. 10.1016/j.engstruct.2018.11.027

  • CrossRef
  • Google Scholar
  • View reference in article

• 4

  Benavent-ClimentA.Ramírez-MárquezA.PujolS. (2018). Seismic strengthening of low-rise reinforced concrete frame structures with masonry infill walls: Shaking-Table Test. Eng. Struct.165, 142–151. 10.1016/j.engstruct.2018.03.026

  • CrossRef
  • Google Scholar
  • View reference in article

• 5

  CaiG.SuQ. (2019). Effect of infills on seismic performance of reinforced concrete frame Structures—A full-scale experimental Study. J. Earthq. Eng.23 (9), 1531–1559. 10.1080/13632469.2017.1387194

  • CrossRef
  • Google Scholar
  • View reference in article

• 6

  CiliaM. G.MooneyW. D.NugrohoC. (2021). Field insights and analysis of the 2018 Mw 7.5 palu, Indonesia earthquake, tsunami and landslides. Pure Appl. Geophys.178 (12), 4891–4920. 10.1007/s00024-021-02852-6

  • CrossRef
  • Google Scholar
  • View reference in article

• 7

  CoutoR.MucederoG.BentoR.MonteiroR. (2024). A practice-oriented approach for seismic and energy performance upgrading of existing buildings. J. Earthq. Eng.28 (15), 4380–4407. 10.1080/13632469.2024.2382474

  • CrossRef
  • Google Scholar
  • View reference in article

• 8

  DautajA. D.KadiriQ.KabashiN. (2018). Experimental Study on the contribution of masonry infill in the behavior of RC frame under seismic loading. Eng. Struct.165, 27–37. 10.1016/j.engstruct.2018.03.013

  • CrossRef
  • Google Scholar
  • View reference in article

• 9

  HeidarzadehM.MuhariA.WijanartoA. B. (2019). Insights on the source of the 28 September 2018 Sulawesi tsunami, Indonesia based on spectral analyses and numerical simulations. Pure Appl. Geophys.176 (1), 25–43. 10.1007/s00024-018-2065-9

  • CrossRef
  • Google Scholar
  • View reference in article

• 10

  KassemW. (2016). Strut-and-Tie modelling for the analysis and design of RC beam-column joints. Mater. Structures/Materiaux Constr.49 (8), 3459–3476. 10.1617/s11527-015-0732-1

  • CrossRef
  • Google Scholar
  • View reference in article

• 11

  LaghiV.PalermoM.TrombettiT.SchildkampM. (2017). Seismic-Proof buildings in developing countries. Front. Built Environ.3, Preprint. 10.3389/fbuil.2017.00049

  • CrossRef
  • Google Scholar
  • View reference in article

• 12

  LataA.GuoX.LuoR.DongX.LiuB. (2022). Experimental Study on the seismic performance of RC frames considering the cast sequence of infilled walls and columns. Structures44, 441–456. 10.1016/j.istruc.2022.08.006

  • CrossRef
  • Google Scholar
  • View reference in article

• 13

  LucaF.DeWoodsG. E. D.GalassoC.D’AyalaD. (2018). RC infilled building performance against the evidence of the 2016 EEFIT central Italy post-earthquake reconnaissance Mission: empirical fragilities and comparison with the FAST method. Bull. Earthq. Eng.16 (7), 2949–2979. 10.1007/s10518-017-0289-1

  • CrossRef
  • Google Scholar
  • View reference in article

• 14

  MansyurM. T.TjarongeM. W. (2025). Experimental investigation of the response of precast foam concrete under monotonic loading. Int. J. Geomate28 (125), 172–179. 10.21660/2025.125.4510

  • CrossRef
  • Google Scholar
  • View reference in article

• 15

  MartinelliG.FuY.LiY.VallianatosF. (2023). Editorial: Pre-Earthquake observations and methods for earthquake forecasting and seismic hazard reduction. Front. Earth Sci.11, Preprint. 10.3389/feart.2023.1150414

  • CrossRef
  • Google Scholar
  • View reference in article

• 16

  MonteiroR.MarquesM.AdhikariG.CasarottiC.PinhoR. (2014). Spectral reduction factors evaluation for seismic assessment of frame buildings. Eng. Struct.77, 129–142. 10.1016/j.engstruct.2014.07.041

  • CrossRef
  • Google Scholar
  • View reference in article

• 17

  MorandiP.HakS.MagenesG. (2018). Performance-Based interpretation of in-Plane cyclic tests on RC frames with strong masonry infills. Eng. Struct.156, 503–521. 10.1016/j.engstruct.2017.11.058

  • CrossRef
  • Google Scholar
  • View reference in article

• 18

  NevilleA. M.BrooksJ. J.ConcreteT. a.M. N.Jj B. (2015). JNCCN J. Natl. Compr. Cancer Netw.13.

  • Google Scholar
  • View reference in article

• 19

  PallarésF. J.DaviaA.HassanW. M.PallarésL. (2021). Experimental and analytical assessment of the influence of masonry façade infills on seismic behavior of RC frame buildings. Eng. Struct.235, 112031. 10.1016/j.engstruct.2021.112031

  • CrossRef
  • Google Scholar
  • View reference in article

• 20

  ParungH.TumpuM.TjarongeM. W.Arwin AmiruddinA.Mansyur (2023a). Flowability and cracks patterns of lightweight concrete subjected to compression and tension loading. Ce/Papers6 (6), 1079–1085. 10.1002/cepa.2907

  • CrossRef
  • Google Scholar
  • View reference in article

• 21

  ParungH.TumpuM.TjarongeM. W.Arwin AmiruddinA.WalennaM. A.Mansyur (2023b). Crack pattern of lightweight concrete under compression and tensile test. Ann. de Chimie Sci. Des Materiaux47 (1), 35–41. 10.18280/acsm.470105

  • CrossRef
  • Google Scholar
  • View reference in article

• 22

  PengZ.DingX.XuZ.ChenZ.LinJ. (2026). Seismic performance of CFS–foamed concrete composite walls: experimental investigation of the effects of flanges and dry connections. Adv. Struct. Eng.13694332261415712. 10.1177/13694332261415712

  • CrossRef
  • Google Scholar
  • View reference in article

• 23

  PribadiK. S.AbduhM.KusumaningrumP.HasiholanB.WirahadikusumahR. D.MilyardiR. (2023). “Promoting humanitarian engineering approaches for earthquake-resilient housing in Indonesia,” in Transcending humanitarian engineering strategies for sustainable futures. 10.4018/978-1-6684-5619-4.ch012

  • CrossRef
  • Google Scholar
  • View reference in article

• 24

  ReggiaA.MorbiA.PretiM.PlizzariG. A. (2020). Lightweight FRC infill wall: in-plane and out-of-plane Loading tests. Mater. Structures/Materiaux Constr.53 (6), 135. 10.1617/s11527-020-01569-7

  • CrossRef
  • Google Scholar
  • View reference in article

• 25

  ReidJ. A.MooneyW. D. (2023). Tsunami occurrence 1900–2020: a global review, with examples from Indonesia. Pure Appl. Geophys.180 (5), 1549–1571. 10.1007/s00024-022-03057-1

  • CrossRef
  • Google Scholar
  • View reference in article

• 26

  RoostaS.LiuYi (2021). Behavior of concrete masonry infills bounded by masonry frames subjected to In-Plane lateral loading – experimental Study. Eng. Struct.247, 113153. 10.1016/j.engstruct.2021.113153

  • CrossRef
  • Google Scholar
  • View reference in article

• 27

  SetyonugrohoG. A.MakiN. (2024). Policy implementation model review of the post-disaster housing reconstruction in Indonesia case Study: aceh, Yogyakarta, and Lombok. Int. J. Disaster Risk Reduct.100, Preprint. 10.1016/j.ijdrr.2023.104181

  • CrossRef
  • Google Scholar
  • View reference in article

• 28

  SongX.LiZ.LiuY. (2025). Research on strategies for enhancing System security resilience of prefabricated building engineering. Sci. Rep.15 (1), 32309. 10.1038/s41598-025-18161-0

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 29

  SrechaiJ.LeelataviwatS.WararuksajjaW.LimkatanyuS. (2022). Multi-Strut and empirical formula-based macro modeling for masonry infilled RC frames. Eng. Struct.266, 114559. 10.1016/j.engstruct.2022.114559

  • CrossRef
  • Google Scholar
  • View reference in article

• 30

  UndrrCRED, and (2021). 2020: the non-covid year in disasters.

  • Google Scholar
  • View reference in article

• 31

  XiaL.LiangW.ZhuZ.LanJ.ZouZ.YuanS. (2024). Strut model and seismic performance of infilled frame structures considering the influence of constructional columns. Structures61, 106060. 10.1016/j.istruc.2024.106060

  • CrossRef
  • Google Scholar
  • View reference in article

• 32

  XuY.ChenH. (2023). Experimental and numerical investigations on interaction between reinforced-concrete grid frame and ceramsite foam concrete infill. Struct. Concr.24 (5), 6141–6156. 10.1002/suco.202200376

  • CrossRef
  • Google Scholar
  • View reference in article

• 33

  YangW.HanF.XuW.ZhangJ. (2025). Experimental analysis of infill Wall stiffness effects on seismic performance of reinforced concrete frame structures using shaking table tests. Structures77, 109108. 10.1016/j.istruc.2025.109108

  • CrossRef
  • Google Scholar
  • View reference in article

• 34

  YuH.ZhuangJ.ChenC. H. (2023). Editorial: Pre-Co-and post-seismic processes of recent earthquakes in mainland China. Front. Earth Sci.11, Preprint. 10.3389/feart.2023.1331948

  • CrossRef
  • Google Scholar
  • View reference in article

• 35

  YuK.ChatziE.KaufmannW.KrausM. A. (2026). Guided generation of strut-and-tie models for reinforced concrete structures with parametric graph grammatical evolution. Adv. Eng. Inf.71, 104302. 10.1016/j.aei.2025.104302

  • CrossRef
  • Google Scholar
  • View reference in article

• 36

  ZhangC.YangZ.YuT.HuangW.DengX.LinZ.et al (2023). Experimental and numerical studies of improving cyclic behavior of infilled reinforced concrete frame by prefabricated Wall panels with sliding joints. J. Build. Eng.77, 107524. 10.1016/j.jobe.2023.107524

  • CrossRef
  • Google Scholar
  • View reference in article

• 37

  ZhangY.ShiQ. X.XinLiWangB.LiuY.WuP.et al (2024). Seismic behavior of prefabricated fiber reinforced concrete energy-dissipation Wall under cyclic loading. Structures70, 107635. 10.1016/j.istruc.2024.107635

  • CrossRef
  • Google Scholar
  • View reference in article

• 38

  ZhengQ.ChenS.LinW. (2024). Numerical Study on seismic performance of a new prefabricated reinforced concrete structural System integrated with recoverable energy-dissipating RC walls. Buildings14 (10), 3243. 10.3390/buildings14103243

  • CrossRef
  • Google Scholar
  • View reference in article