Introduction

The advantages of modern construction using precast concrete is gaining attention due to quality, rapid assembly, reduced labor requirements, and minimized on-site disruption1. Beside these advantages they are considered to be vulnerable during seismic activity2 because of the inadequate connections which are often the weakest links in the structural system3. This necessitates precautions to upgrade the buildings with techniques to improve the performance of structures during earthquake4. Several studies have been conducted to understand the behaviour of precast connections exclusively for seismic failures and strategies have been formulated to ensure structural integrity under lateral loading conditions5.

Researchers have focussed extensively to improve the performance of precast concrete structures through enhanced connection design and member-level strengthening especially when connecting the beam to column, beam to beam, column to column and column to foundation. Studies have also proven to show that the precast system perform comparable to monolithic construction in strength, stiffness, ductility and cyclic stability when reinforcement is confined, and appropriate strengthening mechanisms are followed via transverse reinforcement, FRP wrapping, and steel-plate reinforcement6,7. Precast bolt-connected and steel-plate reinforced concrete substructures have performed effectively during seismic event by redistributing internal forces thereby improving the hysteretic behaviour and controlling the damage occurrence in the existing frames8,9. Also, prestressing and self-centering mechanisms in Precast systems have significantly improved the post-earthquake behaviour thereby showing significant reductions in residual deformations. Numerical studies conducted on Precast substructures were validated experimentally to confirm their ability to achieve stable cyclic response and enhanced seismic resilience10,11. Among various column-to-foundation connection methods adopted in precast construction, three primary approaches are commonly reported: pocket connections, base-plate connections, and sleeve-grouted connections12. Among these, pocket connections—where the column is embedded into a preformed recess within the foundation—have attracted significant attention due to their ability to provide monolithic-like behaviour through a combination of frictional resistance and end bearing13. Three methods are deployed in connecting precast column to foundation, viz. pocket connection, column with base plate connected to foundation and column connected through a sleeve filled with non-shrink grout12. Studies have reported the junction of column and foundation, may experience localized damage during lateral cyclic loading, which can be repaired and restored14. Also, monolithic-like column-foundation precast connection, is easy to assemble and disassemble and therefore is replaceable in case of excessive damage15. Present study focusses on pocket foundation connection, in which the transfer of load from column to foundation is through friction and end bearing within the portion of pocket. The column pocket is roughened to ensure maximum efficiency in the connection16. Hence this connection provides complete rigidity which is in relevance to monolithic connection. The concept of this connection is based on the principle of balance failure. Theoretically balance failure is said to have reached when the compressive strain reaches a value of 0.003 in the extreme fibres of the concrete, simultaneously tension steel reaching the yield strength. Total fixity of the column into foundation is achieved when the insertion into the pocket is 1.5 times the largest cross-sectional dimension of the column17. The end bearing forces may be generated due to bursting pressure and additional links are provided to prevent the bursting pressure18. A non-shrink grout NS3 from Ultratech (strength upto 60 MPa), was used to fill the portion between the column and pocket. In the event of cyclic loading, the load is transferred through the grout from the column to the foundation. Figure 1 depicts the distribution of forces in pocket walls. The performance of the connection depends on the depth of pocket, concrete depth under the pocket and the material between the pocket and column19. Presently, seismic behaviour of precast connections is focused mainly on experimentation, while numerical investigation is relatively less. The resent study focuses on the numerical simulation of the connection in FEA20.

The numerical simulations were performed using ABAQUS 6.14, and key performance indicators—such as ultimate load capacity, strain development, energy dissipation, and ductility—were compared with available experimental data. Studies have proved that accurate modelling can capture realistic hysteretic behaviour in reinforced concrete members, including precast systems7. The present study was conducted to contribute for enhanced seismic resilience of precast column connections by focussing on its design and detailing.

Fig. 1
figure 1

Distribution of forces in column pocket.

Full size image

Significance of the research

Precast concrete structures undergo catastrophic damage by inducing large inelastic deformations during the event of earthquake. All precast systems, especially the connections must be designed to withstand high deformations and enhanced energy dissipation capabilities. Inadequate joint performance in precast structural systems becomes a threat to overall integrity of the structure. Hence connection system in precast structures should adopt robust detailing that is capable of resisting the lateral loads during earthquake. Precast column connected to substructure through pocket foundation is one of the effective methods of connection. In this research, the authors propose two different connection details to investigate the behaviour and performance of the precast column–pocket foundation connection under lateral loading.

Development of precast column foundation connection

To simulate realistic loading and boundary conditions, a four-storeyed reinforced concrete building was modelled using SAP2000, using earthquake-resistant design parameters in accordance with Indian standards. The height of the building was 12.2 m with ground storey height as 3.2 m and other stories as 3 m (Fig. 2). The building was assumed to be located in Chennai, India—classified under Seismic Zone III as per IS 1893 (Part 1): 200221—and founded on moderately stiff soil conditions. The seismic analysis was conducted for 13 load combinations as per IS 1893 (Table 1) and the critical column for the combination of 1.5DL + 1.5EQY was identified and designed as per IS 456: 200022 and detailed as per IS 13,920: 199323.

Fig. 2
figure 2

3-Dimensional view of the model.

Full size image
Table 1 Load Combinations.
Full size table

Based on the analysis results, a critical column–foundation assembly was identified for further evaluation through finite element modeling. The selected column experienced a factored axial load of 1920 kN and a moment of 142.5 kN·m. These forces were scaled down appropriately for 1:2 model simulation, resulting in an axial load of 480 kN and a moment of 17.8 kN·m, which were used for the ABAQUS-based numerical study.

Design of monolithic connection

The monolithic column–foundation connection (Fig. 3) was designed for the above-scaled loads in accordance with IS 45622, IS 13,92023, and SP-16 (1978)24, incorporating ductility-based detailing25. The model dimensions for the 1:2 scaled monolithic connection are shown in Table 2.

Table 2 Dimensions of monolithic connection model.
Full size table

Design and detailing of pocket connection

Two distinct reinforcement detailing strategies were proposed for the pocket-type precast column-to-foundation connections. The pocket creates a recess within the foundation that precisely accommodates the column, ensuring proper alignment and load transfer. The precast column is housed into the pocket with proper alignment and non-shrink grout was selected based on its properties to fill the gap between the column and foundation. The selected grout had the properties of minimal volume change upon curing, and excellent bonding characteristics with a compressive strength of 60 MPa. Also, it ensures stability under various loads preventing the formation of voids or gaps over time. It maintained the connection’s integrity by distributing stresses across the interface and reducing the risk of localized failure. Figure 1 depicts the forces developed in the walls of the pocket. The pocket geometry was designed in accordance with CP 65:199918 and PCI guidelines, and its scaled dimensions are provided in Table 3.

Table 3 Dimensions of pocket foundation model.
Full size table
Fig. 3
figure 3

Reinforcement details of monolithic connection.

Full size image

Connection type PC I

The first connection, PC I (Fig. 4), was designed based on the analytical model proposed by Canha et al.26,27, and detailed per IS 13920:1993 [BIS]. In this model, transverse walls are treated as bending elements subjected to axial and shear forces. The design includes:

  • Transverse reinforcement to resist lateral and frictional forces,

  • Vertical main reinforcement at the corners to handle moment-induced tension,

  • Secondary reinforcement at the wall mid-span to prevent stress concentration.

Corner dowels bent along the transverse wall junctions were anchored at every layer of horizontal reinforcement to improve stress transfer and crack control28. Previous investigations have shown that confinement provided by transverse reinforcement substantially enhances column ductility and axial resistance, which is critical for seismic performance6.

Fig. 4
figure 4

Reinforcement details of PC I connection.

Full size image

Connection type PC II

In PC II (Fig. 5), the transverse walls were treated as structurally independent elements. Each wall was reinforced with vertical and horizontal bars tied at the corners. Additional stirrups were provided at the base of the column over a 300 mm length in the anchorage zone to resist bursting forces caused by end bearing13.

In both configurations, the reinforcement of the pocket walls was extended and lapped with the main reinforcement of the foundation slab, ensuring full development length and structural continuity. Figures 3, 4 and 5 presents the reinforcement details of the monolithic, PC I, and PC II connections, respectively. The specimens were cast and tested in the laboratory, and the experimental results were utilized to develop the finite element model13.

Fig. 5
figure 5

Reinforcement details of PC II connection.

Full size image

Development of finite element model

Concrete structures exhibit nonlinear behaviour under lateral loading conditions and the powerful framework for analysing such behaviour under cyclic loading can be achieved through finite element analysis. In this study, the structural behaviour of precast column–foundation connections were simulated using ABAQUS 6.14, incorporating appropriate material models for concrete, reinforcing steel, and grout. The objective was to capture key performance characteristics such as stress distribution, cracking, energy dissipation, and ductility, and to compare them with experimental observations.

Element types

Concrete

The nonlinear response of concrete under cyclic loading can be best captured using the Concrete Damage Plasticity (CDP) model in ABAQUS29. The CDP model accounts for the stiffness degradation and irreversible damage associated with tension and compression failures in quasi-brittle materials like concrete30.

Under cyclic loading, concrete experiences progressive degradation due to crack formation, plastic strain accumulation, and crushing. The CDP model effectively captures these phenomena using parameters such as tensile damage factor (dt), compressive damage factor (dc), tension stiffening, and plastic strain evolution. The tension stiffening model simulates post-cracking behaviour by allowing stress transfer across cracks through reinforcement31.

The uniaxial tension-compression stress–strain behaviour is shown in Fig. 6. The relevant material parameters32 for M30 grade concrete used in this study are presented in Tables 4, 5., 6 and 7. These include:

  • Compression hardening data (stress–strain).

  • Compression damage.

  • Tension stiffening.

  • Tension damage.

The elastic and plasticity parameters for the CDP model are summarized in Table 8.

Table 4 Compression hardening of concrete.
Full size table
Table 5 Concrete compression damage.
Full size table
Table 6 Tension stiffening of concrete.
Full size table
Table 7 Concrete tension damage.
Full size table
Fig. 6
figure 6

Uniaxial load cycle (tension-compression-tension).

Full size image

Elasticity and plasticity parameters required to define the yield function for the CDP model for various CDP input parameters are given in Table 8.

Table 8 Concrete damage parameters.
Full size table

Steel

Reinforcing steel bars were modelled using an elastic–plastic material model with isotropic hardening. The stress–strain relationship accounts for the initial linear-elastic region, followed by yield plateau and strain hardening until rupture. The properties used for Fe 415 grade steel are listed in Table 9, including:

• Young’s modulus (200 GPa).

• Yield strength (415 MPa).

• Poisson’s ratio (0.33).

• Post-yield plastic strain behaviour.

Table 9 Steel model parameters.
Full size table

Modelling of connections

The precast and monolithic connection geometries were modelled using the following ABAQUS elements (ASCE guidelines):

Concrete: C3D8R (8-node linear brick element with reduced integration).

Steel reinforcement: B31 (2-node linear beam element in space).

Concrete and rebar interaction was defined using embedded region constraints33. The concrete–grout interface between the precast column and the pocket was modelled using a surface-to-surface contact formulation with hard contact in the normal direction and a frictional behaviour in the tangential direction. A Coulomb friction coefficient was assigned based on typical values reported for roughened concrete–grout interfaces, allowing shear stress transfer through friction and bearing action. This approach enables realistic simulation of load transfer mechanisms, including frictional resistance, end bearing, and confinement effects within the pocket region. The embedded region constraint was used only for concrete–reinforcement interaction, ensuring perfect bond between concrete and steel while allowing interface nonlinearity at the concrete–grout contact. These modelling assumptions collectively govern the development of stresses and hysteretic response under reverse cyclic loading. Beam sections were used for reinforcement detailing, and appropriate mesh densities were assigned as shown in Table 10.

Boundary conditions and loading protocol

The column base was fully restrained in all degrees of freedom to simulate fixity with the foundation (Fig. 7). The top of the column was allowed to move only in the lateral direction, simulating lateral cyclic displacement while keeping axial load constant. The loading protocol (Fig. 8) applied reverse cyclic displacement using a Smooth Step amplitude function, with incremental displacement amplitudes:

  • 1 mm up to 5 mm.

  • 2 mm up to 15 mm.

  • 5 mm up to 40 mm.

Each amplitude level consisted of three complete cycles, simulating increasing levels of lateral deformation until failure. Figures 7 and 834, illustrate the boundary conditions and loading history.

Table 10 Details of meshing.
Full size table
Fig. 7
figure 7

Loading and boundary conditions.

Full size image

Results and discussions

This section provides the results and discusses the performance indicators such as strength capacity, material straining, hysteretic response, displacement ductility, and energy dissipation under reverse cyclic loading evaluated through finite element simulations and the structural responses of the three types of column–foundation connections: monolithic, PC I, and PC II.

Strength

Strength capacity is the highest load the specimen can withstand under displacement-controlled cyclic loading34. All models were subjected to increasing displacement cycles up to 40 mm. As shown in Fig. 9, the monolithic specimen exhibited the highest ultimate strength in both positive and negative loading directions. Reduction in strength and stiffness in precast connections in comparison to monolithic systems is often attributed to the discontinuities at the interface and reliance on grout or mechanical connectors which interrupt the concrete continuity and load paths, leading to diminished load transfer and increased slip. This led to enhanced detailing in precast connections to interface the discontinuities and improve the shear capacity. In this study, the precast pocket connection PC II performed closer to the monolithic system than PC (I) The average strength reduction with respect to the monolithic connection was 22.1% reduction in strength for PC I and 15.8% reduction in strength for PC (II) The greater strength reduction observed in PC I compared to PC II is primarily attributed to the concentration of stresses along the transverse pocket walls and corners, where force transfer is less uniformly distributed. In PC I, load transfer is governed by combined action of grout, transverse walls, and corner dowels, leading to localized stress concentrations and earlier cracking, which limits the effective mobilization of reinforcement under cyclic loading. Also, the detailed reinforcement in PC II provided enhanced load transfer and confinement during cyclic action.

Fig. 8
figure 8

Loading protocol.

Full size image
Fig. 9
figure 9

Strength of specimen in (a) positive and (b) negative direction.

Full size image

Ultimate stress and strain in concrete and steel

Figures 10, 11 and 12 illustrate the principal strain contours and damage zones in concrete and steel for all three specimens.

  • Monolithic Connection: Strain was concentrated at the column–foundation interface, with visible crack propagation.

  • PC I: Stress was distributed from the grout to the pocket walls, with moderate strain in the column face and some localized damage near pocket corners.

  • PC II: Exhibited better stress diffusion with reduced straining in critical zones, confirming the efficiency of the independent reinforcement strategy.

By allowing each pocket wall to function as an independent load-resisting component, the reinforcing technique in PC II improved its performance and encouraged more consistent stress diffusion from the grout–column contact into the foundation. The axial and shear forces were spread through the vertical and horizontal reinforcement that are separately tied in PC II lessening the stress concentration at pocket corners in contrast to PC I, where force transfer was concentrated along transverse walls and corner dowels. This behaviour was supported by finite element studies, which demonstrated lower peak stresses and delayed crack initiation in PC II as a result of better confinement and decreased deformation incompatibility at the column–foundation interface. The precast specimens showed lower overall material straining than the monolithic model, suggesting better post-seismic repairability potential.

Load displacement hysteresis behaviour

Figures 13 and 14 presents the load–displacement hysteresis responses of the monolithic, PC I, and PC II connections under reverse cyclic loading. All three models exhibited relatively low pinching, indicating effective bond behaviour and stable load transfer between the column, grout, and foundation. However, it should be noted that ABAQUS does not explicitly model reinforcement bar slip or bond degradation; consequently, pinching effects associated with interface slip and bond deterioration may be underestimated compared to experimental observations. As displacement increases, the envelope curves reveal a progressive stiffness degradation for all connections, which is characteristic of reinforced concrete connections subjected to cyclic loading. The monolithic connection maintained the highest initial stiffness and peak strength, while PC I showed more pronounced stiffness degradation due to stress concentration and localized cracking near the pocket corners. In contrast, PC II consistently maintained stiffness closer to the monolithic response throughout the loading cycles, indicating improved force redistribution and enhanced confinement provided by the independent reinforcement detailing.

Overall, the hysteresis behaviour confirms that while precast connections inherently experience some stiffness degradation due to interface discontinuities, optimized reinforcement strategies—as implemented in PC II—can significantly mitigate these effects. The fuller and more stable hysteresis loops observed for PC II suggest improved energy dissipation capacity and reduced damage accumulation, reinforcing its suitability for seismic applications where sustained cyclic performance is critical.

Displacement ductility

Ductility was assessed based on ASCE guidelines, defined as the ratio of ultimate displacement to yield displacement35. Table 11 presents the displacement ductility factors of the monolithic and precast connection specimens under reverse cyclic loading. The monolithic connection exhibited a ductility factor of 2.0, corresponding to yielding at relatively higher displacement, which reflects its higher initial stiffness and strength but comparatively limited post-yield deformation capacity. In contrast, both precast configurations demonstrated substantially higher ductility, with PC I and PC II exhibiting ductility factors of 3.33 and 4.44, respectively. The enhanced ductility observed in the precast connections is primarily attributed to their ability to accommodate larger inelastic deformations through controlled cracking, interface slip, and energy dissipation within the pocket region. In particular, the reduced yield displacement of PC I and PC II indicates earlier yielding, which promotes stable hysteretic behaviour and delays sudden strength degradation under cyclic loading. Among the precast configurations, PC II showed the highest ductility due to its independent reinforcement strategy, which enabled improved stress redistribution, enhanced confinement, and reduced strain concentration at critical regions.

Overall, the results indicate that although monolithic connections possess higher stiffness, well-detailed precast pocket connections—especially PC II—offer superior deformation capacity, which is a critical requirement for seismic resilience. The significantly higher ductility of PC II highlights its effectiveness in sustaining large cyclic displacements with reduced damage accumulation, making it a favorable connection strategy for application in seismic zones.

Table 11 Displacement ductility.
Full size table
Fig. 10
figure 10

Maximum straining of concrete and steel of monolithic connection.

Full size image
Fig. 11
figure 11

Maximum straining of concrete and steel of PC I connection.

Full size image
Fig. 12
figure 12

Maximum straining in the material concrete and steel of PC II connection.

Full size image
Fig. 13
figure 13

Load-displacement hysteresis behaviour.

Full size image
Fig. 14
figure 14

Load-displacement envelope.

Full size image

Energy dissipation capacity

Energy dissipation (Fig. 15), an important index for evaluating seismic performance, was quantified for each cycle using the area enclosed within each hysteresis loop (Park 1988 & 1989). The cumulative energy dissipated across all cycles is shown in Fig. 16.

  • PC I dissipated more energy throughout the test compared to the other two specimens.

  • Up to 25 mm displacement, PC II and monolithic specimens had similar energy absorption.

  • Beyond 25 mm, PC II surpassed the monolithic connection.

The total energy dissipated (as % more than monolithic):

  • PC I: 62.5% higher.

  • PC II: 36.8% higher.

This confirms that precast pocket connections, especially with optimized reinforcement (PC II), can offer enhanced energy dissipation under seismic loading.

Energy dissipation is a key indicator of seismic performance and was evaluated from the cumulative area enclosed by the hysteresis loops, following the approach proposed by Park36,37. From Fig. 16, PC I dissipated the highest cumulative energy among all specimens, which can be attributed to its earlier yielding, increased cracking, and greater damage accumulation within the pocket region under cyclic loading. The higher energy dissipation in PC I is therefore associated with more pronounced inelastic action and localized damage. Up to a displacement of 25 mm, PC II and the monolithic connection exhibited comparable energy dissipation, indicating similar stiffness retention and damage evolution during the initial loading stages. Beyond this displacement level, PC II surpassed the monolithic specimen in cumulative energy dissipation, reflecting its enhanced ductility and stable hysteretic response at larger deformation demands. This behaviour suggests that the independent reinforcement strategy in PC II effectively promotes controlled energy dissipation through distributed cracking and improved confinement, while limiting premature strength degradation.

Overall, the results demonstrate that precast pocket connections are capable of dissipating significant seismic energy when appropriately detailed. While PC I achieved the highest total energy dissipation due to increased damage concentration, PC II provided a more desirable balance between energy dissipation and structural integrity, achieving 36.8% greater energy dissipation than the monolithic connection with reduced localized damage, thereby confirming its suitability for seismic applications.

Discussion on energy dissipation with ductility implications and design relevance

Energy dissipation, quantified from the cumulative area enclosed by the hysteresis loops36,37, showed clear correspondence with the displacement ductility trends observed in the specimens. PC I exhibited the highest cumulative energy dissipation (62.5% higher than the monolithic connection), which is consistent with its earlier yielding and increased inelastic deformation; however, this higher energy dissipation was accompanied by greater localized damage and stiffness degradation. In contrast, PC II demonstrated a balanced response, where its significantly higher ductility (µ = 4.44) enabled sustained energy dissipation at larger displacement levels, allowing it to surpass the monolithic connection beyond 25 mm displacement while maintaining structural integrity.

From a design perspective, these results indicate that energy dissipation alone should not be considered the sole indicator of seismic performance. While PC I dissipated more energy, its behaviour was governed by damage concentration, whereas PC II achieved efficient energy dissipation through controlled deformation and improved confinement, resulting in superior ductility and reduced strain localization. Therefore, for precast concrete connections in seismic zones, reinforcement strategies similar to PC II are recommended, as they provide an optimal balance between strength retention, ductility, and energy dissipation, which is essential for achieving resilient and repairable seismic performance.

Fig. 15
figure 15

Energy dissipated in each cycle.

Full size image
Fig. 16
figure 16

Cumulative energy dissipated.

Full size image

Future research should extend the present numerical framework by incorporating soil–structure interaction and foundation flexibility to more realistically capture seismic boundary conditions in precast column–foundation systems. Additional experimental and numerical investigations are required to examine the influence of construction tolerances, grout properties, and long-term cyclic degradation on the seismic performance of pocket connections. Furthermore, expanding the analysis to full precast frame systems and developing performance-based design recommendations for different precast connection typologies would enhance the practical application of precast concrete technology in seismic regions.

Conclusion

This study presented a numerical investigation of precast column-to-foundation pocket connections subjected to reverse cyclic loading, comparing two novel reinforcement detailing strategies—PC I and PC II—with a conventional monolithic connection. The study lies in the detailed finite element–based comparison of different pocket reinforcement philosophies and their influence on stress distribution, ductility, and energy dissipation under seismic loading. The finite element simulations, conducted using ABAQUS 6.14, provided valuable insights into the structural behaviour, strength, ductility, and energy dissipation characteristics of these connection types.

The following conclusions can be drawn from the study:

  • The finite element model successfully captured the load–displacement trends observed in experimental studies, with an average deviation of less than 15%, confirming the reliability of numerical simulation as a tool for assessing seismic behaviour in precast concrete connections.

  • The monolithic connection exhibited the highest strength, while PC II showed superior performance among the precast configurations. The average strength reduction with respect to the monolithic specimen was 22.1% for PC I and 15.8% for PC II, highlighting the effectiveness of the independent reinforcement strategy.

  • Precast specimens exhibited lower strain concentrations in critical regions compared to the monolithic specimen, indicating improved damage distribution and enhanced post-seismic repair potential.

  • Ductility values for PC I and PC II were 66.5% and 122% higher, respectively, than the monolithic specimen, demonstrating the ability of well-detailed precast connections to accommodate large inelastic deformations under seismic loading.

  • The total energy dissipated by PC I and PC II was 62.5% and 36.8% greater than that of the monolithic connection, confirming the effectiveness of optimized pocket detailing in resisting cyclic loading.

Overall, the results demonstrate that precast pocket connections with optimized and independently reinforced detailing can achieve seismic performance comparable to, or exceeding, that of traditional monolithic connections. Among the studied configurations, PC II exhibited the most balanced behaviour, combining high ductility, enhanced energy dissipation, and acceptable strength capacity, making it a promising solution for seismic applications. Future research focus can incorporate foundation flexibility, conduct parametric optimization of pocket detailing, and extending the numerical framework to full precast frame systems to further support performance-based seismic design of precast concrete structures.