Behaviour of hybrid precast reinforced concrete beam-column connections with steel sections under cyclic loading

ABSTRACT

This research presents an experimental and analytical investigation of an interior precast beam-column connection using a hybrid steel connector. The beam-column connections were introduced with different positive and negative reinforcement alignments to achieve behaviour comparable to a monolithic reinforced concrete connection. The primary objective of this study is to examine key parameters such as load-carrying capacity, mode of failure, crack pattern, ductility, stiffness, and energy dissipation capacity in beam-column connections. Three half-scale beam-column connection specimens were tested under displacement-controlled lateral cyclic loading combined with constant axial loading. One cast-in-place control specimen and two precast specimens were tested. The control specimen was designed based on the strong-column–weak-beam concept to meet code requirements for strength, while the precast specimens were detailed to replicate this design approach. To enhance performance, the precast specimens (PBC1 and PBC2) were fabricated with different configurations, including variations in reinforcement alignment and using hybrid steel connectors of varying sizes. Considering the test variables, the above-mentioned key parameters for the control specimen were compared with those of the precast specimens (PBC1 and PBC2). The results showed that the precast specimens demonstrated enhanced strength and energy dissipation compared to the monolithic specimen, based on single-specimen test results. This clarification addresses the limited statistical representation due to using only one specimen per configuration. Using hybrid steel connectors with asymmetric detailing and varied connector heights represents a novel approach to improving seismic behaviour in interior joints. Seismic performance was further evaluated using damage indices and validated through Finite Element Analysis (FEA) using ABAQUS software. The FEA results closely correlated with the experimental results. Additionally, the analytical results help to predict the experimental results at the peak point.

Keywords:
Precast members; Beam-column connectors; Energy dissipation; Finite element analysis; Hybrid steel connectors

1. INTRODUCTION

Earthquakes are devastating natural disasters that damage infrastructure and threaten human lives. India is rapidly developing in the construction sector using various technologies, and the construction sector must keep pace with the growing population. At the same time, ensuring buildings can withstand seismic events, especially in regions prone to earthquakes [¹]. A key solution is using prefabricated systems in construction, which offer reduced construction time, enhanced quality control, and lower overall costs. Although the initial cost of prefabrication may be higher than traditional methods, it becomes a competitive and efficient alternative when considering factors like material waste, labour costs, and maintenance [²]. However, a critical aspect of these systems is the beam-column connection, which must resist lateral and axial forces during earthquakes. Historically, precast concrete structures, even when designed for seismic conditions, have faced challenges such as loss of shear strength, buckling of longitudinal reinforcement, reduced lateral load resistance, and excessive drifts. These failures, especially in beam-column joints, have resulted in significant structural damage and collapses. Moment-resisting precast beam-column connections are typically categorized as wet or dry. In wet connections, reinforcement bars are extended using mechanical couplers or splicing through the joint, providing continuity.

In contrast, dry connections rely on bolted, welded, or post-tensioned elements for assembly. Hybrid connections, a combination of wet and dry methods, have gained popularity due to their balance of strength and flexibility. Recent research emphasizes the importance of ductile detailing in these connections to prevent shear damage and enable inelastic deformation, adopting the strong-column–weak-beam concept to improve seismic performance. This philosophy is widely incorporated into design codes such as IS 13920-2016 [³], ACI 318-19 [⁴], and EN 1998-1:2004 [⁵], all of which provide guidelines for reinforcement detailing and seismic performance evaluation of Reinforced Concrete (RC) members. Several studies have explored various connection types to enhance seismic resilience [⁶].

The precast beam-column connections with T-section steel inserts under cyclic loading showed that monolithic specimens exhibited ductile behaviour, while precast specimens with double steel inserts performed better [⁷]. Similarly, a U-profile steel corbel in beam-to-column connections demonstrates sufficient torsional strength during the assembly phase [⁸]. Additionally, the role of embedded channel-shaped steel (ECSS) in precast concrete beam-ends was explored, and it was found that increasing ECSS length and densifying stirrup spacing significantly improved bearing and deformation capacity [⁹]. The hybrid steel-concrete connections demonstrated that precast specimens with embedded steel sections provided excellent ductility and strength under cyclic loading. It aligns with the drive to enhance joint performance [¹⁰]. The novel connection detail incorporates steel connectors and Engineered Cementitious Composites (ECC), significantly improving beam-column joints’ ductility and seismic resilience [¹¹]. The strides were developing interlocking mechanical joints for precast concrete, offering rapid construction and eliminating the need for concrete pour forms, making them a promising alternative to conventional methods [¹²]. The seismic behaviour of precast hybrid beam-column connections with welded components, focusing on the welding area between the beam-end plates and the bottom longitudinal rebars. Their results highlight the potential of hybrid connections to enhance the seismic resilience of precast concrete structures [¹³]. Another study has developed a Hybrid Prefabricated Concrete (HPC) connection combining wet (mechanical couplers for reinforcement bars) and dry techniques (bolted steel angles for concrete elements). Their results showed significant improvements in strength, energy dissipation, and crack control, enhancing the seismic resilience of precast concrete structures [¹⁴]. The precast beam-column connections use small-diameter high-strength bars to replace ordinary steel bars at the bottom of the prefabricated beam. Their study showed that these precast joints exhibited superior strength, and high-strength bars effectively shifted the plastic hinge zone away from the column surface, enhancing the overall structural behaviour [¹⁵].

Furthermore, reviewed the advantages of hybrid beam-column connections in precast concrete structures, emphasizing their ability to strengthen seismic resilience through improved strength and ductility. The authors highlighted hybrid connections as a promising solution for addressing the limitations of traditional methods, offering benefits like increased construction speed and material durability. Some studies also investigate alternative methods of precast connections [¹⁶, ¹⁷]. Similarly, the seismic performance of precast beam-column connections with embedded bolt connectors was validated, highlighting that their performance is comparable to cast-in-place connections [¹⁸].

Despite these advancements, significant gaps remain in current research. Most studies focus on specific connection types, without fully exploring the potential of hybrid connectors that combine positive and negative reinforcement. Challenges persist in accommodating beams on all sides of columns in interior joint configurations, particularly under seismic loads. Additionally, there has been limited exploration of grout concrete depth limits in beam-column joints; exceeding these limits can result in incomplete filling and reduced structural integrity. This underscores the need for further research into optimal grout application techniques for deeper joints. Practical limitations, such as obstructions caused by corbels or bolt extensions, and the reliance on temporary props during installation, also present opportunities for developing more efficient connection designs. This study investigates the use of hybrid steel connectors in beam-column joints, focusing on overcoming seismic limitations in interior joints and improving performance through varying reinforcement alignments. It also addresses practical constraints in current construction practices. Although previous studies have investigated precast joints with mechanical connectors, most focused on symmetrical reinforcement layouts or full-depth steel connections, especially in exterior joints. Researchers have explored hybrid approaches, such as T-section steel inserts, embedded channel-shaped steel, and steel–ECC composite joints. However, these systems rarely address interior joint detailing or construction-related challenges like formwork and grouting.

In contrast, the present study introduces a hybrid steel connector system for interior beam-column joints, featuring a cross-shaped steel core with anchor bolts, bolted side plates, and asymmetrical reinforcement. Two variations in connector embedment depth (115 mm and 230 mm) were tested to evaluate their influence on seismic performance. The system eliminates the need for temporary props during assembly and does not require post-inserted grout, as the connector’s thickness ensures a tight fit and full engagement. Only minimal shuttering is needed to cast the top zone of the joint. These innovations represent a novel and practical approach for enhancing constructability and seismic energy dissipation, validated through experimental and numerical analysis. Compared to existing hybrid systems, the proposed connector integrates cross-shaped anchorage with bolted assembly and asymmetrical reinforcement, designed specifically for interior joints. The system reduces on-site labor by eliminating grouting and the need for temporary props, requiring only minimal formwork during final concrete pouring. This approach offers a practical and novel solution for seismic-resistant precast construction. The research methodology of the present study is illustrated in Figure 1.

Figure 1
Research methodology of the research.

2. MATERIAL PROPERTIES

2.1. Concrete

The test specimens were produced using Ordinary Portland Cement (OPC) 53 grade and M₃₀ grade concrete with a water-cement ratio of 0.495. The fine aggregate had a fineness modulus of 4.66, and the coarse aggregate had a fineness modulus of 7.65. The concrete cubes were cast and tested for compressive strength at 28 days, with an average result of 38.81 N/mm². The tensile strength was assessed using three cylindrical specimens (150 mm in diameter × 300 mm in height), yielding a recorded tensile strength of 3.85 N/mm². Additionally, three beam specimens (100 mm × 100 mm × 500 mm) were cast and tested for flexural strength, with an average result of 4.633 N/mm². The mixed proportions of the present study are reported in Table 1.

Table 1
Mix proportions of concrete kg/m³.

CEMENT	FA	CA	WATER CONTENT	WATER CEMENT RATIO
445	877	973	220	0.495
Ratio – 1:1.99:2.19

2.2. Reinforcing steel bars and steel plate

Fe500-grade deformed steel bars were used for longitudinal and transverse reinforcement in monolithic and precast concrete specimens. Additionally, steel plates and anchor bars were utilized to form the Hybrid steel connector at the joint of the precast concrete test specimens. The mechanical properties of the steel were determined through uniaxial tensile tests conducted in the laboratory. The steel rebars and sections used in the monolithic and precast specimens, and the mechanical properties of steel rebars and sections, are presented in Table 2.

Table 2
Mechanical properties of steel rebar and section.

REBARS/SECTION	YIELD STRENGTH (MPa)	ULTIMATE STRENGTH (MPa)
6 mm rebar	544.72	633.77
8 mm rebar	541.25	629.71
Steel plate	290.37	360.02
8 mm Anchor bar	517.18	589.54

2.3. Prototype building design for interior precast beam-column joints

A 4-story reinforced concrete office building prototype with plan dimensions of 12 m × 12 m was considered in this study. Each story had a height of 3.00 m, and the bay width was 3.00 m for all floors. The beam dimensions were selected as 230 mm (width) × 250 mm (depth), and the column dimensions as 300 mm × 300 mm, under IS 456-2000 [¹⁹]. Figure 2 illustrates the plan and elevation of the four-story reinforced concrete frame, highlighting the selected interior beam–column joint (A) for testing. Adjacent joints B, C, D, E, F and G are also marked to provide context within the structural system. The analysis focused on joint A in the middle of the first floor, where shear forces, bending moments, and axial forces were calculated under various seismic load combinations. The seismic analysis was done using SAP2000 software, following the equivalent static force method per IS 1893-2016 [²⁰]. The design and detailing of beams, columns, and joints were based on IS 456-2000 and IS 13920-2016, consistent with global design philosophies adopted in ACI 318–19 and Eurocode 8, which emphasise strong-column–weak-beam behaviour and ductility enhancement in seismic zones. The Strong-Column–Weak-Beam (SCWB) concept was applied to ensure that beams would yield and dissipate energy during seismic events. At the same time, the columns remained undamaged, preventing catastrophic collapse and facilitating repair. This study tested three interior beam–column concrete specimens: one monolithic and two precast specimens. The precast specimens incorporated hybrid steel connectors embedded within the beam–column joint to enhance energy dissipation under lateral cyclic loading.

Figure 2
The four-story office building frame’s plan, elevation and member details.

Before specimen fabrication, SAP2000 analysis of the full-scale frame yielded peak design moments of 40.76 kN.m for the beams and 57.81 kN.m for the columns, resulting in a moment capacity ratio ∑Mc ≥ 1.4 ∑Mb ≈ 1.42, thereby satisfying the SCWB criterion ∑Mc ≥ 1.4 ∑Mb as per IS 13920:2016. The specimens were scaled to half the size of the prototype beam–column frame elements, using Cauchy’s similitude law [²¹], which ensures dynamic and geometric similarity. A geometric scale factor (λ) of 1:2 was adopted, meaning all lengths (beam depth, column width, etc.) were reduced by half. This results in a scale factor of λ² = 1/4 for cross-sectional areas, and λ³ = 1/8 for volumes and self-weight. Moments and stiffness values were indirectly affected by these reductions. However, the material properties (concrete strength, steel grade) were not scaled to maintain realistic behaviour, a common practice in structural engineering. Reinforcement detailing followed the same percentage of steel area (A_s) relative to cross-section, ensuring similar stress-strain response. This approach allowed us to capture the nonlinear cyclic behaviour under scaled lateral displacements with sufficient accuracy. The scaled beam section measured 115 mm × 125 mm (width × depth), and the column section was 150 mm × 150 mm. The length of both the beam and column test elements was 750 mm, simulating the mid-span region of 1.5 m-long full-scale members, where moments are minimal under lateral seismic loads. These dimensions were derived by applying a 1:2 geometric scaling factor to the original prototype beam section of 230 mm × 250 mm and the column section of 300 mm × 300 mm, ensuring the scaled specimens reflected the behaviour of the full-size design. Monolithic and precast specimens used the exact dimensions and reinforcement details to provide comparable strength. Figure 3 shows the isometric view of the monolithic specimen (MBC) and the precast specimens (PBC1 and PBC2).

Figure 3
Isometric view of (a) MBC, (b) PBC1 and (c) PBC2.

2.4. Monolithic connection

The monolithic reinforced concrete specimen was designed and detailed based on structural analysis results obtained from SAP2000. The analysis of the prototype frame provided the internal forces such as moments, shears, and axial loads at the selected beam–column joint. Using these extracted forces, the beam and column sections were designed by IS 456-2000 and detailed for seismic performance per IS 13920-2016, suitable for moderate seismic zones. The concrete beam had flexural reinforcement consisting of three ϕ6 mm bars at the top and two ϕ6 mm bars at the bottom, resulting in longitudinal reinforcement ratios of 0.41% and 0.26%, respectively. Shear reinforcement consisted of two-legged ϕ6 mm stirrups, spaced at 26.75 mm near the beam ends (within 2d distance) and 53.50 mm in the mid-span region. The column had four longitudinal ϕ8 mm bars (0.8% steel ratio), with two-legged ϕ6 mm stirrups spaced at 37.50 mm near the ends (distance l_o) and 75 mm in the middle. Additional confined reinforcement was provided within the joint region. Figure 4 illustrates the reinforcement configuration and concrete cross-sections for the monolithic specimen.

Figure 4
Cross-section and detailing of monolithic specimen.

2.5. Precast concrete specimens

The precast specimens were constructed using precast beams and columns connected through a hybrid steel connector system. Each precast beam incorporated a 3 mm thick, 60 mm wide, and 310 mm long steel plate pre-fitted with shear bolts, embedded centrally within its cross-section during casting. Of this length, 260 mm was cast into the beam, while the remaining 50 mm projected outward into the joint core. The central connecting element was a cross-shaped mild steel section, initially anchored to the lower precast column using four 8 mm diameter, 300 mm long anchor bolts made of ribbed reinforcing bars (rebars). The ends of these bars were threaded to allow mechanical fastening with the cross-shaped steel section. During tensile testing of the anchor bars, direct fixing with nuts led to slippage. Therefore, two 8 mm TMT rebars were threaded at both ends using a lathe machine, with partial removal of ribs to facilitate nut engagement. A mechanical coupler connected the threaded segments, and the specimen was successfully mounted in the testing machine. The average yield and ultimate strengths were 376.67 N/mm² and 429.35 N/mm², respectively, indicating that the threading process did not significantly compromise the mechanical performance. Although the anchor bolts were embedded in concrete during cyclic testing, no signs of distress or slip were observed, confirming the adequacy of the threaded connections.

The 6 mm diameter shear bolts and stud bolts used in the side plate connections and steel side plates were commercial mild steel bolts, typically conforming to Property Class 4.6 as per IS 1367 [or ISO 898-1], with an estimated double shear capacity of 12.36 kN per bolt. The bolts were manually tightened using nuts on both sides without pretensioning. During assembly of the precast specimens, two precast beams were brought in from opposite sides and aligned with the cross-shaped steel section. Steel side plates (60 mm × 43 mm × 3 mm) were then bolted through both the projecting beam plates and the face of the cross-core section, forming a rigid mechanical interlock. The 6 mm diameter shear bolts, arranged in three rows across a 70 mm width of the embedded beam plate, and the 6 mm stud bolts used for the side plate connections were commercial mild steel bolts, typically conforming to Property Class 4.6 as per IS 1367 (or ISO 898-1), with an estimated double shear capacity of 12.36 kN per bolt. All bolts were manually tightened using nuts on both sides without pretensioning. The number, diameter, and spacing of bolts were selected based on the anticipated shear force demand at the joint interface, ensuring sufficient capacity under lateral cyclic loading. Shielded Metal Arc Welding (SMAW) with 6 mm fillet welds was used to fabricate the cross-shaped steel section. Welds were applied along all connector contact surfaces using IS 816 procedures, ensuring adequate ductility and preventing premature failure under reversed cyclic loading. After the beams were fastened, the upper column, also precast, was aligned and bolted on top of the cross-section using a similar anchor arrangement. This complete configuration, comprising embedded steel plates, bolted side plates, anchor bolts, and the welded cross-shaped steel section, is collectively referred to as the hybrid steel connector, as illustrated in Figure 5. The detailing of this system followed the recommendations of IS: 800-2007 [²²]. The beam-column joints were designed with plastic hinge formation expected within the beam span, away from the column face, as recommended in IS 13920:2016. The detailing avoided stress concentrations near the column by extending the hybrid connector into the joint core and providing adequate anchorage and confinement. The shear connectors comprising a cross-shaped steel core, anchor bolts, and bolted plates strengthened the joint core region, forcing yielding and damaging the adjacent beam region. This behaviour was confirmed through the observed cracking pattern and damage concentration during testing.

Figure 5
(a) Cross-shaped steel section, (b) anchor bolt, (c) steel plate with shear bolts, (d) plate connector, (e) final assembly of hybrid steel connector.

To enable proper reinforcement continuity, a gap was created at the top face of each beam, allowing the placement of top bars across the joint. Once the reinforcement cage was fully fixed and all connections secured, cast-in-place concrete was poured into the joint core and the beam’s upper zone as depicted in Figures 6(a) and (b). This filled the voids and completed the monolithic behaviour of the joint. The actual construction sequence is illustrated in Figure 7(a–f), showing: (a) Precast column, (b) Precast beam, (c) Hybrid connector for PBC1, (d) Hybrid connector for PBC2, (e) Connector installation, and (f) Final alignment before concreting.

Figure 6
Precast beam-column joint assemblies - (a) PBC1 and (b) PBC2.

Figure 7
Construction sequence of the precast beam-column joint using the hybrid steel connector.

2.5.1. Precast concrete connection 1 (PBC1)

In the precast concrete specimen (PBC1), the Hybrid steel connector was positioned at the beam-column joint, with a gap left for top reinforcements to facilitate connections. The cross-shaped steel section in PBC1 had a height of 115 mm. The longitudinal reinforcement ratios matched those of the Monolithic connection (MBC) (0.414% – 3# – ϕ6). Bottom reinforcements were installed up to the beam face near the beam-column joint and secured with a 90° upward-bent bar, 55 mm long. The longitudinal reinforcement ratios were consistent with those of MBC (0.262% – 2# – ϕ6). In the column section, all longitudinal bars protruded from the bottom and were anchored sideways with a 90° hook, following the MBC ratios (0.8% – 4# – ϕ8) depicted in Figure 8(a). After installation, cast-in-place concrete was poured into the joint’s core and the upper region of the beam.

Figure 8
Reinforcement details of precast beam-column connections: (a) PBC1, (b) PBC2.

2.5.2. Precast concrete connection 2 (PBC2)

In PBC2, the joint steel support connection and detailing were similar to PBC1, with the key difference being the height of the cross-shaped steel section and bottom reinforcement of the beam. The cross-shaped steel section in PBC2 was increased to a height of 230 mm, double that of PBC1. Additionally, bottom reinforcements were anchored downward with a 90° hook, and the bent bar length was extended to 110 mm, connecting to the base plate of the steel joint support, as depicted in Figure 8(b). The theoretical moment capacities of the monolithic specimen were computed to be 2.66 kN·m for the beam and 9.94 kN·m for the column, yielding a SCWB ratio of 3.74, which indicates a highly conservative design. For the precast specimens (PBC1 and PBC2), the hybrid steel connector was analytically evaluated: the moment resistance of each side plate (60 mm wide, 3 mm thick, 290.3 MPa yield strength) was calculated as 0.039 kN·m, totalling 0.078 kN·m for both plates. Additionally, each 6 mm bolt (with a double shear capacity of 12.36 kN) placed across a 60 mm wide connector provided 0.371 kN·m of moment. With six bolts arranged symmetrically, the total bolt-contributed moment reached 2.23 kN·m. Consequently, the overall moment capacity of the hybrid connector was estimated at 2.31 kN·m, which is approximately 87% of the monolithic beam’s capacity. While not entirely equivalent to monolithic continuity, the hybrid system was intentionally designed to balance seismic strength, ductility, and constructability performance attributes confirmed through experimental and numerical studies.

2.6. Experimental setup

The boundary conditions of the test specimens were designed to represent the mid-span behaviour of full-scale structural elements under seismic loading. In buildings, beams’ mid-span and columns’ mid-height typically experience near-zero moments during lateral sway. Only half the span length (1.5 m of a 3 m member) was modelled and tested to simulate this condition, which was further scaled to 0.75 m in the physical specimens. Mechanical hinge supports were provided at the ends of the beam and at the base of the lower column to represent the approximate zero-moment boundary conditions observed at mid-span in real structures.

Two key restraint mechanisms were incorporated to prevent any out-of-plane rotation or instability. First, the hinge supports were designed to allow only in-plane rotation of the beam and column, i.e., rotation about the horizontal axis within the loading plane, ensuring that no unintended out-of-plane freedom was introduced. Second, two lateral raking steel supports were fixed at the base of the column in the direction perpendicular to the applied lateral load. These acted as stiff boundary braces to prevent any twisting or falling of the specimen in the out-of-plane direction. In addition, a constant axial load of approximately 87.3 kN, equivalent to 0.10 times the effective cross-sectional area of the column multiplied by the characteristic compressive strength of concrete (0.10f_ck·A_g), was applied at the top of the column using a hydraulic jack. The jack was stabilized using four high-strength steel rods anchored to the column base to ensure vertical alignment and to prevent uplift or eccentric movement during cyclic loading. This setup promoted beam yielding before any potential column failure under extreme loading conditions. Together, these measures provided full restraint against out-of-plane motion and torsional buckling. The specimen remained stable and deformed purely in-plane throughout testing, validating the boundary conditions assumed in the numerical simulations. Monolithic and precast concrete specimens were subjected to lateral displacement-controlled cyclic loading and a constant axial load. The axial load was manually monitored throughout the test by closely observing the oil pressure gauge at each displacement step (e.g., 1 mm, 1.5 mm, and 2 mm). This ensured the applied vertical force remained within the target range of 87.3 kN during the entire loading history.

All specimens were tested until failure under the 2000 kN capacity of the loading frame, as shown in Figure 9. LVDTs were installed to measure the deflection of the test specimen during the cyclic load test. Four dial indicators were positioned 100 mm from the face of the column, as shown in Figure 10(a). These dial indicators were placed on the beam and the column to accurately measure localised deformations during loading. The upper column was subjected to cyclic forward and backwards movements, simulating half-cycles to represent seismic loading. This method allowed the column to return to its initial position naturally, without the external force pulling it back, thereby capturing the realistic restoration behaviour of structures during seismic events. Figure 10(b) and Table 3 detail the loading history, including displacement cycles. The lateral displacement of the test specimens was expressed in terms of the story drift ratio (∆/h), where ∆ is the lateral displacement and h is the height of the test specimen.

Figure 9
Experimental setup of the specimen.

Figure 10
(a) Schematic experimental test setup, (b) Loading history.

Table 3
Displacement-based loading sequence.

STEP(S)	STORY DRIFT RATIO	DISPLACEMENT (MM)
1	0.000625	1
2	0.000937	1.5
3	0.001250	2
4	0.001870	3
5	0.002500	4
6	0.003700	6
7	0.005000	8
8	0.006870	11
9	0.010000	16
10	0.014370	23
11	0.020620	33
12	0.029370	47
13	0.041250	66

The loading began in the linear elastic range, starting at a drift ratio of 0.00063%. Subsequent drift ratios increased incrementally between 1 and 1.5 times the previous value, progressing through 0.00094%, 0.00125%, 0.00187%, and so on, until a maximum drift ratio of 0.035% was reached. At each stage, the peak amplitude was repeated three times to assess the stability of the hysteresis response, following the cyclic loading specifications of ACI T1.1-01 [²³]. Although this standard typically prescribes full reverse cyclic loading, the use of displacement-controlled half-cycles in this study effectively captured in-plane drift demands and degradation behaviour. This approach is appropriate for simulating realistic seismic effects, as structural elements in buildings are not externally restored after every drift excursion during actual earthquakes. The observed hysteresis loops and energy dissipation trends further validated the reliability of the half-cycle method for this investigation.

3. NUMERICAL MODELING

A total of three elements, such as monolithic specimen (MBC1), precast specimen (PBC1), and precast specimen (PBC2), were modelled as three-dimensional (3D) deformable elements in both concrete and steel reinforcement using ABAQUS finite element software [²⁴], as shown in Figures 11–12. The numerical specimens were analyzed using the same setup and loading history as their corresponding experimental specimens.

Figure 11
3D model of the element.

Figure 12
Reinforcement details of the element.

3.1. Material properties and element types

In numerical analysis, a compressive strength of 30 MPa was considered for the concrete. The Concrete Damage Plasticity (CDP) approach was used to simulate the non-linear behaviour of concrete [²⁵]. A sensitivity analysis was conducted on the monolithic specimen (MBC) by varying the dilation angle between 25° and 40°, and the viscosity parameter from 0.0005 to 0.01. Based on the best agreement with the experimental displacement response and stiffness degradation, a dilation angle of 31° and a viscosity parameter 0.001 were selected. A shape factor 0.667 was also applied to approximate the peak load observed experimentally. Sensitivity analyses were conducted to refine the CDP parameters, although some discrepancies remained between the numerical and experimental behaviours. After testing several values, a dilation angle of 31° was selected based on its overall influence on load-displacement response, and a shape factor of 0.667 was applied to approximate the peak load observed experimentally. A viscosity parameter of 0.001 was used to maintain stability and facilitate convergence. Additional properties for concrete included a density of 2,458 kg/m³, a Young’s modulus of 27,386 MPa, a Poisson’s ratio of 0.18, and an eccentricity of 0.01. The steel properties, including yield strengths and Young’s modulus, were based on laboratory tests, with high-tensile steel bars (8 mm, 6 mm) and steel plates having design yield strengths of 541.25 N/mm², 544.72 N/mm², and 290.37 N/mm², respectively. The density, Young’s modulus, and Poisson’s ratio for steel were 7,850 kg/m³, 200,000 MPa, and 0.3. Eight-node linear bricks with reduced integration (C3D8R) were used to model the concrete components of the beams and columns, while a two-node linear beam element (B31) was utilized to model the reinforcement cage.

3.2. Interaction properties and convergence criteria

The steel reinforcement and concrete interaction were treated as an embedded element. For the steel and concrete interaction, hard contact with separation was defined in the normal direction, and a friction coefficient of 0.3 was assigned in the tangential direction, a commonly used value in finite element modelling for such materials. Hinges were assigned at both ends of the beam and the bottom of the column using the connector section assignment manager. In the general static analysis, multiple steps were employed to define analysis intervals with the necessary time and increments to capture the non-linear behaviour of the specimens. Convergence criteria were applied to ensure stable results: Automatic Incrementation was enabled to optimise stability, with an initial increment size of 0.01, a minimum increment of 1e-25, and a maximum increment of 1, allowing ABAQUS to adjust increment sizes dynamically. Automatic stabilisation, with a specified dissipated energy fraction of 0.0002, was used to help maintain convergence and control non-linear behaviour.

3.3. Mesh and load module

The mesh for the numerical model was generated in ABAQUS using a global element size of 20 mm, which was applied to all parts of the specimen, including concrete, reinforcement bars, and steel connectors. The mesh size was selected based on geometric complexity and convergence stability, ensuring accurate stress resolution while maintaining computational efficiency. The mesh for the numerical models was automatically generated by ABAQUS, optimising the mesh size based on geometry and analysis requirements, as shown in Figure 13. This approach ensured an efficient balance between accuracy and computational cost. In the load module, a constant axial load was applied at the top of the column to simulate the loading conditions observed in the experimental setup. Boundary conditions were specified, with connectors assigned to model hinge behaviour similar to the corresponding experimental specimens.

Figure 13
Meshing of the element.

4. RESULTS AND DISCUSSION

4.1. Load carrying capacity

The yielding force and displacement approach was proposed and utilizing the criterion of reduced stiffness equivalent to an elasto-plastic yield in lateral load-displacement responses [²⁶]. The corresponding lateral load vs. displacement relationships for each specimen are illustrated in Figures 14–16. The maximum load-carrying capacities of MBC, PBC1, and PBC2 were recorded as 6.54 kN, 9.4 kN, and 11.2 kN in the experimental results, whereas the numerical analysis predicted 6.81 kN, 10 kN, and 11.22 kN, respectively. The comparison reveals that PBC2 exhibits the highest load-carrying capacity at peak displacement, surpassing MBC in both experimental and numerical analyses. The load vs. displacement curves indicate that PBC1 and PBC2 specimens are stronger than the MBC specimens. This increase in strength is attributed to incorporating the cross-shaped steel support with anchor bolts at the beam-column joint, enhancing load transfer efficiency by dissipating forces across the column and beam elements. At each drift level, the first cycle generally exhibited a slightly higher peak load than the subsequent cycles. This behaviour is attributed to the higher initial stiffness and intact bond conditions during the first loading at each drift. In precast specimens, additional effects such as bolt seating, frictional slip at steel interfaces, and closure of initial construction tolerances may also contribute to sharper initial peaks in the load-displacement response. These effects are typical in displacement-controlled cyclic tests involving hybrid or mechanically connected systems.

Figure 14
Comparison of the experimental and numerical load vs. displacement curves of the MBC specimen.

Figure 15
Comparison of the experimental and numerical load vs. displacement curves of PBC1 specimen.

Figure 16
Comparison of the experimental and numerical load vs. displacement curves of PBC2 specimen.

The overall load-displacement envelope is presented in Figure 17. The hysteresis response of the experimental specimen exhibits a loop shape that is initially narrow at lower displacements and gradually expands as displacement increases. In contrast, the numerically simulated specimen displays a more uniform loop shape with minimal variation in peak loads across cycles. The elastic region of the experimental response shows minor irregularities, whereas the numerical model maintains a smoother and more linear progression. Additionally, the experimental specimen demonstrates greater amplitude variation due to material degradation and microcracking at higher displacement levels, while the numerical model remains consistent throughout the cycles. A comparison of yield force and displacement further highlights these differences. The experimental yield forces for MBC, PBC1, and PBC2 were 5.58 kN, 7.56 kN, and 9.99 kN, while the numerical analysis yielded 5.7 kN, 8.55 kN, and 9.54 kN, respectively. Similarly, the experimental yield displacements were recorded as 25.32 mm, 19.65 mm, and 29.64 mm, whereas the numerical results showed 26.84 mm, 25.16 mm, and 24.25 mm, respectively. These differences indicate that the experimental results accurately capture real-world material behaviour, including degradation effects. In contrast, numerical results exhibit slightly higher peak load values due to idealized material properties and the assumption of perfect bonding between concrete and reinforcement. The comparative force and displacement values for the experimental and numerical specimens are presented in Figure 18.

Figure 17
Load-displacement envelopes of the specimens.

Figure 18
Experimental vs. numerical evaluation of strength and displacement in beam-column joints.

4.2. Energy dissipation capacity

The energy dissipation capacity of the specimens was evaluated by integrating the enclosed area of each displacement hysteresis loop, using the trapezoidal rule method. Cumulative energy dissipation curves for the monolithic (MBC) and precast specimens (PBC1 and PBC2) are shown in Figure 19, all computed up to a consistent target drift level of 4.13%.

Figure 19
Cumulative energy dissipation curves for MBC, PBC1, and PBC2 specimens.

The MBC specimen dissipated 372.12 kN·mm of energy. In comparison, PBC1 achieved 510.15 kN·mm, representing a 37% increase over MBC, while PBC2 reached 617.50 kN·mm, a 66% improvement. These enhancements are attributed to the hybrid steel connector, which contributed significantly to hysteretic damping and improved joint performance under cyclic loading. The steadily rising cumulative energy curves confirm that both precast specimens effectively absorbed seismic energy, with PBC2 demonstrating the highest capacity. A similar methodology was adopted, and energy dissipation was quantified by calculating the loop area under cyclic loading for precast beam–column joints. The agreement in approach reinforces the validity and applicability of this method for evaluating the seismic energy absorption characteristics of precast structural systems [²⁷, ²⁸, ²⁹, ³⁰, ³¹].

4.3. Equivalent viscous damping ratio

The equivalent viscous damping ratio (ζeq) provides a valuable measure of energy dissipation capacity during cyclic lateral loading. A higher damping ratio indicates improved energy dissipation, which helps reduce seismic vibrations and enhance structural resilience. In this study, (ζeq) was computed as the ratio of energy dissipated in a half cycle to the strain energy of an equivalent linear elastic system, divided by 2π. This approach is consistent with the method proposed, which has been widely adopted in cyclic testing of precast seismic connections [²⁷]. The strain energy was calculated for each drift level based on the actual peak force and displacement of that half cycle, allowing consistent normalization across variable drift amplitudes. Minor residual displacements were observed in later cycles due to plastic deformation, but their effect on energy and stiffness normalization was minimal. Displacement readings were adjusted cycle-by-cycle to ensure accurate area integration. Figure 20 illustrates the variation of the equivalent viscous damping ratio concerning the drift ratio for all specimens. The monolithic specimen (MBC) initially demonstrated better damping performance up to a drift ratio of 0.25%, with an equivalent viscous damping ratio of 6.261%. This can be attributed to its elastic behaviour. However, as inelastic deformations accumulated, the damping ratio gradually decreased, reaching its lowest value of 3.977% at a drift ratio of 0.68%. Beyond this point, due to the effects of hysteretic behaviour, the damping ratio exhibited fluctuations but remained relatively stable, reaching 5.036% at a drift ratio of 4.13%.

Figure 20
Equivalent viscous damping ratio curves for MBC, PBC1, and PBC2 specimens.

PBC1 initially exhibited a lower damping ratio for the precast specimens than the monolithic specimens, with a value of 3.575% at a drift ratio of 0.68%. However, its performance improved over higher drift levels, stabilizing at 6.425% beyond a drift ratio of 1.43% and maintaining a damping ratio of 5.705% at 4.13% drift. Conversely, PBC2 experienced an initial decline in damping, reaching 4.775% at a 0.68% drift ratio, followed by a sharper increase in damping performance. Eventually, PBC2 maintained a higher damping ratio beyond a drift ratio of 4.13%, reaching 6.065%. PBC2 consistently exhibited a higher damping ratio at larger drift levels than MBC and PBC1, indicating superior energy dissipation capacity. This behaviour can be attributed to the hybrid steel connector, which enhances hysteretic damping through controlled deformation, friction, and localized yielding. Unlike MBC, which relies solely on concrete cracking and rebar yielding, or PBC1, which may experience reduced joint engagement, PBC2’s hybrid connector maintains effective load transfer while permitting inelastic deformations. It increases energy dissipation, making PBC2 a more seismically resilient solution.

4.4 Secant stiffness degradation

Stiffness degradation is a critical parameter for evaluating the seismic performance of structural elements under repeated cyclic loading. This study used secant stiffness as a metric to quantify this degradation. It is defined as the slope of the line joining the origin to the peak point of each displacement cycle: Ksec = P_peak/Δ_peak. This method provides an average measure of global stiffness at each drift level, reflecting material nonlinearity and joint degradation [³²]. The variation of secant stiffness with increasing displacement is presented in Figure 21. For the Monolithic Specimen (MBC), the initial secant stiffness was 2.2 kN/mm at 1 mm displacement. A steep reduction occurred early in loading due to initial micro-cracking in the beam–column joint, reducing stiffness to 1.05 kN/mm at 2 mm. Thereafter, the decline was more gradual, reaching 0.1 kN/mm at 66 mm. The significant drop is attributed to flexural cracking, reinforcement yielding, and degradation of the joint core, particularly within the plastic hinge region adjacent to the column face.

Figure 21
Secant stiffness degradation curves for MBC, PBC1, and PBC2 specimens.

The Precast Specimen PBC1 showed a higher initial stiffness of 2.3 kN/mm, remaining greater than MBC up to 8 mm displacement (0.79 kN/mm). This improved early stiffness is due to the immediate engagement of its compact hybrid connector system, which facilitated efficient load transfer across the joint. However, the shorter embedded length and partial anchorage may have contributed to a steeper stiffness degradation beyond 8 mm, leading to a final value of 0.14 kN/mm at 66 mm. Numerical simulation results, shown in Figure 22(a) revealed significantly higher stress concentration across the hybrid connector region, confirming the immediate mobilization of the steel plates and bolts, which contributed to the increased initial stiffness response. In contrast, PBC2 exhibited a slightly higher initial stiffness of 2.5 kN/mm and a more gradual and sustained degradation path. Its larger connector embedment and full anchorage of bottom bars into the column core enabled better force distribution and delayed crack propagation. As a result, stiffness remained more stable under increasing drift, ending at 0.17 kN/mm at 66 mm, which is higher than MBC and PBC1. As shown in Figure 22(b), the stress concentration in PBC2 was more localized and lower, reflecting a reduced stiffness in early stages. However, this behaviour led to a more ductile performance with less stress accumulation, indicating improved cyclic stability over time. These results demonstrate that connector geometry, anchorage detailing, and reinforcement continuity significantly influence the stiffness degradation behaviour of precast joints. A more gradual stiffness decline indicates improved energy dissipation and ductile performance. Therefore, the superior stiffness retention observed in PBC2 highlights the effectiveness of its extended connectors and full-depth anchorage in enhancing seismic resilience under cyclic lateral loading.

Figure 22
Von mises stress and plastic strain contours for (a) PBC1 and (b) PBC2 at ultimate drift level.

4.5. Displacement ductility

Displacement ductility quantifies a structure’s ability to undergo plastic deformations beyond its initial yield point before failure. This study adopted the definition of equivalent yield and ultimate displacement based on the reduced stiffness equivalent elasto-plastic yield criteria in lateral load-displacement responses. The yield displacement of the equivalent elasto-plastic system was determined using the secant stiffness at 0.75 of the ultimate lateral load. Figure 23 presents the calculated displacement ductility factors for monolithic and precast beam-column connections. Among the specimens, PBC1 exhibited the highest displacement ductility factor (3.35), followed by MBC (2.60) and PBC2 (2.22), indicating that PBC1 had the most extraordinary capacity for plastic deformation before failure. However, this increased ductility came at the cost of significant stiffness degradation. PBC2, while having a lower ductility factor than PBC1, retained slightly better stiffness, suggesting that the hybrid steel connector in PBC2 contributed to improved stiffness retention. The presence of the hybrid steel connector facilitated a more efficient load transfer mechanism, enabling controlled plastic deformation and delaying stiffness degradation. Its adequate anchorage allowed forces to be distributed across multiple components, enhancing energy dissipation and ensuring a more stable cyclic response. The variations in ductility performance can be attributed to differences in joint detailing and the steel support provided in the precast specimens, which directly influenced their stiffness and deformation capacity. Among all specimens, MBC exhibited the least stiffness degradation, reinforcing the inherent rigidity of monolithic connections compared to precast alternatives.

Figure 23
Displacement ductility factor curves for MBC, PBC1, and PBC2 specimens.

4.6. Crack pattern and failure mode

Figures 24–26 illustrate the crack formations observed in all experimental and numerical test specimens under cyclic loading. As the load was applied at the top of the column from the right, the left side of the beam functioned as a tension zone at the bottom and a compression zone at the top. Conversely, on the right side, the beam experienced tension at the top and compression at the bottom due to both beam ends being hinge-supported. Initially, flexural cracks emerged in the potential plastic hinge zones of the beams in each specimen. As the drift level increased, these flexural cracks expanded, and additional cracks formed in the concrete columns adjacent to the joint. With further increases in drift, diagonal cracks appeared within the core of the beam-column joints.

For the monolithic specimen (MBC), shown in Figure 24, initial flexural cracks appeared in the beams at a drift level of 0.5%, extending about 50 mm from the column faces [³³]. The experimental image (Figure 24a) shows that cracks predominantly developed near the beam-column interface, with marked flexural and shear cracks. Multiple vertical flexural cracks were visible at a 1% drift level in the beam tension zones, spanning approximately 350 mm from the column faces. The numerical model (Figure 24b) captures a similar pattern, with red-highlighted regions indicating stress concentration zones, aligning with experimentally observed crack locations. When the drift level reached 1.43%, horizontal flexural cracks formed in the column below the joint. At a 2.0% drift level, existing flexural cracks near the column faces widened to about 2 mm, increasing further to 3 mm at a 3.0% drift level, with diagonal cracks forming in the beam-column joint. These crack widths were measured during testing using a handheld optical crack microscope. Although minor uncertainties existed due to surface roughness and lighting conditions, consistency was maintained by repeating measurements and verifying readings across both beam faces. Finally, at 4.0% drift, the concrete cover in the compression zone near the column faces began to degrade, with fragments detaching.

Figure 24
Crack formation: (a) Experimental specimen (MBC), (b) Numerical specimen (MBC).

Figure 25 illustrates the crack formations observed in the precast concrete specimen (PBC1) under cyclic loading. Due to a hanger bar extending along the beam length, only a few vertical flexural cracks were visible in the beam tension zone on the right side at a 0.375% drift level, extending about 50 mm to 200 mm from the column faces. On the left side, where the bottom bar was curtailed, cracks initiated at 50 mm and extended to 350 mm. At 1.43% drift, multiple vertical flexural cracks had spread to approximately 550 mm from the column faces. As seen in the experimental image (Figure 25a), the crack propagation followed a pattern similar to monolithic behaviour but with notable differences in crack distribution due to the precast connection. At 2.0% drift, the cracks near the column faces widened, reaching 2 mm on the right and 2.5 mm on the left sides. By 3.0% drift, further crack expansion was observed, reaching widths of 3 mm. At 4.0% drift, the concrete cover in the beam’s compression zone near the column faces degraded, leading to visible detachment of fragments [³⁴]. However, in contrast to the monolithic specimen, no significant vertical or horizontal flexural cracks were observed on the top or bottom of the columns, indicating different stress distribution characteristics in the precast system. The numerical model (Figure 25b) validates these findings, showing stress concentration zones (highlighted in red) in the beam-column joint and beam tension zones, correlating well with the experimentally observed cracks. This confirms that the hybrid steel connector in PBC1 influenced crack propagation and damage distribution differently than the monolithic specimen.

Figure 25
Crack formation: (a) Experimental specimen (PBC1), (b) Numerical specimen (PBC1).

Figure 26(a) illustrates the crack patterns in the precast beam-column specimen (PBC2) under cyclic loading. At an early drift level of 0.25%, a few vertical flexural cracks appeared in the beam tension zone on the left side, extending from the column faces to approximately 250 mm. This was due to the downward extension of the bottom reinforcement bar. However, no cracks were observed on the right side at this stage. At 0.5% drift, vertical flexural cracks started forming in the beam tension zone, extending up to 350 mm from the column faces. These cracks propagated further by a 2.0% drift, reaching approximately 450 mm from the column faces. At 3.0% drift, the pre-existing flexural cracks widened to about 1 mm on the right side and 1.5 mm on the left side of the beam. This indicates an asymmetric crack pattern, influenced by the reinforcement detailing. At 4.0% drift, significant degradation of the concrete cover in the compression zone near the column faces was observed, with visible detachment of concrete fragments.

Figure 26
Crack formation: (a) Experimental specimen (PBC2), (b) Numerical specimen (PBC2).

Additionally, vertical splitting cracks appeared at the top of the bottom column, likely due to the downward extension of the beam’s bottom reinforcement [³⁵]. The numerical simulation (Figure 26b) confirms these observations, where high-stress concentration areas (highlighted in red) align with the experimentally observed crack locations. Unlike PBC1, which exhibited more uniform crack distribution, PBC2 showed more pronounced cracks in the bottom column region, highlighting the influence of reinforcement layout and connection detailing on failure behaviour.

The study indicates that the primary failure mode for precast specimens PBC1 and PBC2 was flexural cracking in the potential plastic hinge regions, extending approximately 550 mm from the column faces [³⁶]. Additionally, minor shear cracks were observed below the beam-column joint in PBC2. The failure modes of the monolithic and precast beam-column specimens were analyzed based on the maximum equivalent stress (Von Mises stress) and plastic equivalent strain (PEMAG), as depicted in Figures 27–29. The monolithic specimen (MBC) exhibited the highest tensile stress of approximately 500 MPa, whereas the precast specimens PBC1 and PBC2 recorded peak values of 441 MPa and 474 MPa, respectively. These stress values were obtained from the von Mises stress contours and primarily reflect the behaviour of the embedded steel components (e.g., reinforcement bars and steel plates), which possess yield strengths in the 290–545 MPa range. The concrete regions, in contrast, exhibited much lower stress levels. To evaluate damage in concrete, the equivalent plastic strain (PEMAG) values were considered more representative than stress contours. The corresponding peak plastic strain values recorded were 0.0375 for MBC, 0.0416 for PBC1, and 0.0445 for PBC2, highlighting variations in deformation behaviour under cyclic loading. Highlighting variations in deformation behaviour under cyclic loading [³⁷]. In MBC, strain development was progressive and uniformly distributed across the joint with increasing drift ratio. At higher drift levels, strain concentration became more pronounced in the plastic hinge region, forming flexural cracks. The stress distribution remained relatively uniform, with peak stress occurring near the joint due to bending moments. Under ultimate load conditions, localized concrete crushing was observed near the beam-column joint, yet the monolithic nature of the specimen facilitated effective load transfer, preventing premature failure [³⁸].

Figure 27
Distribution contour of maximum equivalent stress (left column) and equivalent plastic strain (right column) for monolithic specimen (MBC).

Figure 28
Distribution contour of maximum equivalent stress (left column) and equivalent plastic strain (right column) for precast specimen (PBC1).

Figure 29
Distribution contour of maximum equivalent stress (left column) and equivalent plastic strain (right column) for precast specimen (PBC2).

In contrast, PBC1 exhibited a nonlinear strain response, with significant variations at the beam-column interface due to the presence of the hybrid steel connector. Localized strain concentrations were observed at the joint, indicating differential movement between the precast elements. At higher drift ratios, joint opening and minor slippage contributed to increased deformation. The stress distribution in PBC1 was highly localized, with stress peaks concentrated near the steel connector. While the hybrid steel connector effectively transferred forces, stress accumulation around the steel plates and bolts was observed in the numerical model, indicating regions of potential vulnerability. However, no signs of fatigue, yielding, or damage were observed experimentally in these regions. Under ultimate loading conditions, stress redistribution was evident, leading to stiffness degradation and increased joint deformation [³⁹].

PBC2 demonstrated a more balanced strain distribution than PBC1, indicating improved stress transfer across the joint. Strain variations were less severe, suggesting enhanced resistance to cyclic loading. At higher drift levels, strain localization remained controlled, reducing excessive deformation at the joint. The stress distribution in PBC2 showed a smoother transition between the beam and column, with lower stress concentrations than PBC1. The hybrid steel connector in PBC2 facilitated more effective load distribution, mitigating stress accumulation at the joint. However, stress concentration at the steel connection was still present, emphasizing the importance of precise detailing in precast joint design [⁴⁰]. Although stress concentrations were observed numerically around the hybrid steel connector, no signs of physical damage, bolt slippage, or connector failure were visible during or after experimental testing. Due to the embedded nature of the connection, internal deformation or yielding of the steel components could not be directly inspected. However, the absence of distress in the concrete surrounding the connection zone and the sustained load transfer capacity throughout testing indicate that the connector remained structurally intact and functioned effectively under cyclic loading.

The numerical results were consistent with the experimentally observed failure mechanisms, as concrete deformation was predominantly localized in the beam face and the beam-column joint for all specimens. No significant deformation was observed in the columns of either the monolithic or precast specimens, confirming the reliability of the numerical model in capturing the structural response under cyclic loading [⁴¹].

5. CONCLUSIONS

This study investigated the structural performance of monolithic and precast beam-column connections under cyclic lateral loading. Two newly developed precast connections (PBC1 and PBC2), incorporating hybrid steel connectors, were experimentally tested and numerically analyzed to evaluate their failure behaviour, stiffness degradation, load-carrying capacity, hysteresis behaviour, energy dissipation, and displacement ductility. The results were compared with a monolithic specimen (MBC) to assess the efficiency of the precast alternatives. The key conclusions drawn from this study are summarized below:

1. Crack Formation

   1. MBC exhibited severe cracking near the beam-column joint, with diagonal cracks forming at later drift levels due to high-stress concentration and the development of a plastic hinge.

   2. PBC1 displayed localized cracking at the beam-column interface, influenced by the hybrid steel connector. Cracks propagated more gradually, resulting in reduced damage in the joint core.

   3. PBC2 showed a more balanced crack distribution, with cracks extending farther from the column face and minimal cracking in the joint region, indicating improved stress transfer.

   4. Numerical simulations closely matched experimental observations, accurately predicting stress concentration zones and crack propagation patterns.

2. Load-Carrying Capacity and Stiffness:

   1. The maximum load-carrying capacities recorded in experiments were 6.54 kN (MBC), 9.4 kN (PBC1), and 11.2 kN (PBC2), representing 42.8% and 71.3% strength increases for PBC1 and PBC2, respectively, compared to MBC. These performance increases are based on single-specimen tests for each configuration. No statistical variability or confidence interval could be reported due to the limited experimental samples.

   2. Numerical predictions were slightly higher, with peak loads of 6.81 kN (MBC), 10.0 kN (PBC1), and 11.22 kN (PBC2), likely due to idealised material behaviour.

   3. PBC2 exhibited the highest initial stiffness, attributed to its taller joint support and extended lap length for reinforcement anchorage.

   4. Stiffness degradation was most pronounced in PBC1, while PBC2 retained better stiffness, highlighting the effectiveness of the hybrid steel connector in minimizing stiffness loss during repeated cycles.

3. Hysteresis and Ductility:

   1. PBC2 exhibited wider hysteresis loops, indicating enhanced energy dissipation, with 2.96% greater energy absorption than PBC1 and 30% higher than MBC.

   2. PBC1 demonstrated the highest displacement ductility factor (3.35), followed by MBC (2.60) and PBC2 (2.22).

   3. Despite its lower ductility factor, PBC2 maintained better stiffness retention, suggesting that the hybrid steel connector stabilized cyclic performance.

4. Energy Dissipation:

   1. The total energy dissipation values were 1273.14 kN·mm (MBC), 1821.6 kN·mm (PBC1), and 2184.6 kN·mm (PBC2).

   2. PBC2 exhibited the highest energy dissipation, 41.72% greater than MBC and 30.1% greater than PBC1, confirming the hybrid steel connector’s superior seismic energy absorption capability.

5. Numerical Accuracy:

   1. Numerical models accurately captured crack patterns, failure modes, and stress-strain distributions observed in experiments.

   2. Numerical analysis slightly overestimated load-carrying capacity due to assumptions of perfect bonding and idealized material properties, while experimental specimens exhibited micro-cracking and material degradation.

   3. Stiffness degradation trends were consistent between numerical and experimental results, validating the reliability of the finite element model.

6. Implications for Precast Design:

   1. The hybrid steel connector proposed in PBC2 significantly improved structural performance, making it a viable alternative to monolithic connections in precast frames.

   2. Optimizing joint detailing and reinforcement configurations could enhance ductility and energy dissipation without compromising stiffness retention.

   3. Future research should focus on long-term durability, fatigue performance, and large-scale implementation of hybrid precast beam-column joints in seismic-prone regions.

   4. The findings suggest that PBC2 provides a more efficient balance of strength, stiffness, and energy dissipation, making it a promising solution for precast frame applications under seismic loading.

   5. Compared to hybrid systems, the proposed connector integrates cross-shaped anchorage with bolted assembly and asymmetrical reinforcement, which is explicitly designed for interior joints. The system reduces on-site labour by eliminating grouting and the need for temporary props, requiring only minimal formwork during final concrete pouring. This approach offers a practical and novel solution for seismic-resistant precast construction.

7. Practical Implications and Future Research

   1. The proposed hybrid steel connector in PBC2 improved structural performance, making it a viable alternative to monolithic connections in precast frames.

   2. Optimizing joint detailing and reinforcement configurations could enhance ductility and energy dissipation without compromising stiffness retention.

   3. Future research should focus on long-term durability, fatigue performance, and large-scale implementation of hybrid precast beam-column joints in seismic-prone regions.

While the current experimental findings are based on half-scale specimens, it is essential to recognise potential scale effects in full-scale implementations. In buildings with column heights exceeding 3 m, changes in slenderness, joint rotation, and connector engagement could affect performance. The slip and rotation capacity of the hybrid connector, in particular, may be influenced by increased moment arm and deformation demands. Hence, further full-scale testing is recommended to validate the connector’s seismic efficiency and mechanical reliability under real-world conditions.

6. ACKNOWLEDGMENTS

Vasanthakumar T: Participated in planning, implementing the parametric study, researching methodology, drafting the original manuscript, and designing the analytical and experimental parametric study. Kalpana V.G: Reviewed and supervised the experimental works. Sasikumar P: Copy editing of the manuscript.

7. BIBLIOGRAPHY

• ^[1] Abbas, M., Elbaz, K., Shen, S.L., et al, “Earthquake effects on civil engineering structures and perspective mitigation solutions: a review”, Arabian Journal of Geosciences, v. 14, pp. 1350, 2021. doi: http://doi.org/10.1007/s12517-021-07664-5.
  » https://doi.org/10.1007/s12517-021-07664-5
• ^[2] Rocha, P., Ferreira, N.; Pimenta, F., et al, 2023. “Impacts of prefabrication in the building construction industry”, Encyclopedia, v. 3, n. 1, pp. 28–45. doi: https://doi.org/10.3390/encyclopedia3010003.
  » https://doi.org/10.3390/encyclopedia3010003
• ^[3] BUREAU OF INDIAN STANDARDS, Ductile Design and Detailing of Reinforced Concrete Structures Subjected to Seismic Forces: Code of Practice, IS 13920–2016, New Delhi, Bureau of Indian Standards.
• ^[4] AMERICAN CONCRETE INSTITUTE, 318-19 Building Code Requirements for Structural Concrete and Commentary, Baltimore, American Concrete Institute, 2019.
• ^[5] EUROPEAN COMMITTEE FOR STANDARDIZATON, Eurocode 8: Design of Structures for Earthquake Resistance – Part 1: General Rules, Seismic Actions and Rules for Buildings, EN 1998-1, 0 Brussels, CEN, 2004.
• ^[6] Navaratnam, S., Ngo, T., Gunawardena, T., et al, “Performance review of prefabricated building systems and future research in Australia”, Buildings, v. 9, n. 2, pp. 38, 2019. doi: https://doi.org/10.3390/buildings9020038.
  » https://doi.org/10.3390/buildings9020038
• ^[7] Ketiyot, R., Hansapinyo, C., “Seismic performance of interior precast concrete beam-column connections with T-section steel inserts under cyclic loading”, Earthquake Engineering and Engineering Vibration, v. 17, n. 2, pp. 355–369, 2018. doi: http://doi.org/10.1007/s11803-018-0446-9.
  » https://doi.org/10.1007/s11803-018-0446-9
• ^[8] Araújo, D., Prado, L.P., Da Silva, E.B., et al, “Temporary beam-to-column connection for precast concrete frame assembly”, Engineering Structures, v. 171, pp. 529–544, 2018. doi: http://doi.org/10.1016/j.engstruct.2018.05.089.
  » https://doi.org/10.1016/j.engstruct.2018.05.089
• ^[9] Guo, X., Gao, S., Wang, L., et al, “Bearing capacity of embedded channel-shaped steel connections at precast concrete beam end”, Engineering Structures, v. 175, pp. 177–190, 2018. doi: http://doi.org/10.1016/j.engstruct.2018.08.043.
  » https://doi.org/10.1016/j.engstruct.2018.08.043
• ^[10] Li, B., Kulkarni, S.A., Leong, C.L., “Seismic performance of precast hybrid-steel concrete connections”, Journal of Earthquake Engineering, v. 13, n. 5, pp. 667–689, 2009. doi: http://doi.org/10.1080/13632460902837793.
  » https://doi.org/10.1080/13632460902837793
• ^[11] Choi, H.K., Choi, Y.C., Choi, C.S., “Development and testing of precast concrete beam-to-column connections”, Engineering Structures, v. 56, pp. 1820–1835, 2013. doi: http://doi.org/10.1016/j.engstruct.2013.07.021.
  » https://doi.org/10.1016/j.engstruct.2013.07.021
• ^[12] Nzabonimpa, J.D., Hong, W.K., Park, S.C., “Experimental investigation of dry mechanical beam–column joints for precast concrete-based frames”, Structural Design of Tall and Special Buildings, v. 26, n. 1, pp. e1302, 2017. doi: http://doi.org/10.1002/tal.1302.
  » https://doi.org/10.1002/tal.1302
• ^[13] Girgin, S.C., Misir, I.S., Kahraman, S., “Experimental cyclic behaviour of precast hybrid beam-column connections with welded components”, International Journal of Concrete Structures and Materials, v. 11, n. 2, pp. 229–245, 2017. doi: http://doi.org/10.1007/s40069-017-0190-y.
  » https://doi.org/10.1007/s40069-017-0190-y
• ^[14] Chen, W., Xie, Y., Guo, X., et al, “Experimental investigation of seismic performance of a hybrid beam–column connection in a precast concrete frame”, Buildings, v. 12, n. 6, pp. 801, 2022. doi: https://doi.org/10.3390/buildings12060801.
  » https://doi.org/10.3390/buildings12060801
• ^[15] Yu, J., Zhang, E., Xu, Z., et al, “Seismic performance of precast concrete frame beam-column connections with high-strength bars”, Materials (Basel), v. 15, n. 20, pp. 7127, 2022. doi: http://doi.org/10.3390/ma15207127. PubMed PMID: 36295202.
  » https://doi.org/10.3390/ma15207127
• ^[16] Septiarsilia, Y., Iranata, D., Suswanto, B., “Hybrid beam-column connection of precast concrete structures: a review”, E3S Web of Conferences, v. 434, pp. 02019, 2023. doi: https://doi.org/10.1051/e3sconf/202343402019.
  » https://doi.org/10.1051/e3sconf/202343402019
• ^[17] Rave-Arango, J.F., Blandón, C.A., Restrepo, J.I., et al, “Seismic performance of precast concrete column-to-column lap-splice connections”, Engineering Structures, v. 172, pp. 687–699, 2018. doi: http://doi.org/10.1016/j.engstruct.2018.06.049.
  » https://doi.org/10.1016/j.engstruct.2018.06.049
• ^[18] Xue, W., Bai, H., Dai, L., et al, “Seismic behaviour of precast concrete beam‐column connections with bolt connectors in columns”, Structural Concrete, v. 22, n. 3, pp. 1297–1314, 2021. doi: http://doi.org/10.1002/suco.202000007.
  » https://doi.org/10.1002/suco.202000007
• ^[19] BUREAU OF INDIAN STANDARDS, Indian Standard Code of Practice for Plain and Reinforced Concrete, IS 456–2000, New Delhi, Bureau of Indian Standards.
• ^[20] BUREAU OF INDIAN STANDARDS, Criteria for Earthquake Resistant Design of Structures Part 1 General Provisions and Buildings, IS 1893–2016 (Sixth Revision), New Delhi, Bureau of Indian Standards.
• ^[21] Zdraveski, F., Mickoski, H., Petreski, Z., “Application of similitude laws for experimental investigations of dynamic properties of tall prototype steel structure”, Annals of the Faculty of Engineering Hunedoara, v. 14, n. 2, pp. 25, 2016.
• ^[22] BUREAU OF INDIAN STANDARDS, Code of Practice for General Construction in Steel, IS: 800-2007, New Delhi, Bureau of Indian Standards.
• ^[23] ACI Standard, Acceptance criteria for moment frames based on structural testing and commentary, Farmington Hill, ACI Standard, 2005.
• ^[24] Dassault, S., Abaqus analysis user’s guide, Providence, RI, USA: Dassault Systèmes Simulia Corp., 2014.
• ^[25] Woon, K.S., Hejazi, F., “Vertical cyclic performance of precast frame with hook-end u-shaped high-damping rubber joint”, KSCE Journal of Civil Engineering, v. 23, n. 5, pp. 2215–2226, 2019. doi: http://doi.org/10.1007/s12205-019-1115-x.
  » https://doi.org/10.1007/s12205-019-1115-x
• ^[26] Park, R., “Evaluation of ductility of structures and structural assemblages from laboratory testing”, Bulletin of the New Zealand society for earthquake engineering, v. 22, n. 3, pp. 155–166, 1986.
• ^[27] Saqan, E.I., “Evaluation of ductile beam-column connections for use in seismic-resistant precast frames”, D.Sc. Thesis, University of Texas at Austin, Texas, 1995.
• ^[28] Saravanan, M.M., Sasikumar, P., “Assessment of axial load carrying capacity of fully encased composite columns: comparative study with different codes”, Matéria (Rio de Janeiro), v. 29, n. 3, pp. e20240303, 2024. doi: http://doi.org/10.1590/1517-7076-rmat-2024-0303.
  » https://doi.org/10.1590/1517-7076-rmat-2024-0303
• ^[29] Sasikumar, P., “Behaviour of high strength concrete encased steel composite short columns subjected to axial load”, Revista de la Construcción. Journal of Construction, v. 24, n. 1, pp. 407–416, 2025.
• ^[30] Sasikumar, P., “Experimental study on the fully encased composite short columns made with high strength fibre reinforced concrete”, Asian Journal of Civil Engineering, v. 25, pp. 3239–3250, 2024. doi: http://doi.org/10.1007/s42107-023-00975-w.
  » https://doi.org/10.1007/s42107-023-00975-w
• ^[31] Sasikumar, P., Manju, R., “Performance of high strength concrete encased steel composite columns”, Revista Romana de Materiale - Romanian Journal of Materials, v. 52, n. 4, pp. 374–384, 2022.
• ^[32] Lu, C., Dong, B., Pan, J., et al, “An investigation on the behaviour of a new connection for precast structures under reverse cyclic loading”, Engineering Structures, v. 169, pp. 131–140, 2018. doi: http://doi.org/10.1016/j.engstruct.2018.05.041.
  » https://doi.org/10.1016/j.engstruct.2018.05.041
• ^[33] Bahrami, S., Madhkhan, M., Shirmohammadi, F., et al, “Behaviour of two new moment resisting precast beam to column connections subjected to lateral loading”, Engineering Structures, v. 132, pp. 808–821, 2017. doi: http://doi.org/10.1016/j.engstruct.2016.11.060.
  » https://doi.org/10.1016/j.engstruct.2016.11.060
• ^[34] Gou, S., Ding, R., Fan, J., et al, “Seismic performance of a novel precast concrete beam-column connection using low-shrinkage engineered cementitious composites”, Construction & Building Materials, v. 192, pp. 643–656, 2018. doi: http://doi.org/10.1016/j.conbuildmat.2018.10.103.
  » https://doi.org/10.1016/j.conbuildmat.2018.10.103
• ^[35] Khaloo, A., Doost, R.B., “Seismic performance of precast RC column to steel beam connections with variable joint configurations”, Engineering Structures, v. 160, pp. 408–418, 2018. doi: http://doi.org/10.1016/j.engstruct.2018.01.039.
  » https://doi.org/10.1016/j.engstruct.2018.01.039
• ^[36] Nzabonimpa, J.D., Hong, W.K., Kim, J., “Nonlinear finite element model for the novel mechanical beam-column joints of precast concrete-based frames”, Computers & Structures, v. 189, pp. 31–48, 2017. doi: http://doi.org/10.1016/j.compstruc.2017.04.016.
  » https://doi.org/10.1016/j.compstruc.2017.04.016
• ^[37] Feng, D.C., Wu, G., Lu, Y., “Finite element modelling approach for precast reinforced concrete beam-to-column connections under cyclic loading”, Engineering Structures, v. 174, pp. 49–66, 2018. doi: http://doi.org/10.1016/j.engstruct.2018.07.055.
  » https://doi.org/10.1016/j.engstruct.2018.07.055
• ^[38] Meydanli Atalay, H., Ozden, S., “Cyclic behaviour of beam-column connections in precast structures”, Frontiers in Built Environment, v. 7, pp. 639778, 2021. doi: http://doi.org/10.3389/fbuil.2021.639778.
  » https://doi.org/10.3389/fbuil.2021.639778
• ^[39] Alameddine, F., Ehsani, M.R., “Seismic behaviour of gravity load designed reinforced concrete buildings”, ACI Structural Journal, v. 88, n. 1, pp. 71–78, 1991.
• ^[40] Girgin, S.C., Misir, I.S., Kahraman, S., “Experimental cyclic behaviour of precast hybrid beam-column connections with welded components”, International Journal of Concrete Structures and Materials, v. 11, n. 2, pp. 229–245, 2017. doi: http://doi.org/10.1007/s40069-017-0190-y.
  » https://doi.org/10.1007/s40069-017-0190-y
• ^[41] Vidjeapriya, R., Jaya, K.P., “Experimental study on two simple mechanical precast beam-column connections under reverse cyclic loading”, Journal of Performance of Constructed Facilities, v. 27, n. 4, pp. 402–414, 2023. doi: http://doi.org/10.1061/(ASCE)CF.1943-5509.0000324.
  » https://doi.org/10.1061/(ASCE)CF.1943-5509.0000324

Publication Dates

• Publication in this collection
  27 Oct 2025
• Date of issue
  2025

History

• Received
  22 June 2025
• Accepted
  11 Sept 2025