A composite steel-concrete beam (CB) features a steel I-beam under tension, a concrete slab under compression, and shear connectors, ensuring the two materials are securely joined at their interface (Brozzetti, 2000; Queiroz et al., 2007; Ahmed, 2018; Papastergiou and Lebet, 2014; Shamass and Cashell, 2019; Ashraf et al., 2000; Liu et al., 2017; Peng et al., 2024). Thus, CB effectively maximizes the structural characteristics of steel and concrete. Compared to traditional steel or concrete beams used in isolation, this composite action significantly enhances the beam’s load-carrying capacity, stiffness, and resistance to bending and deflection. By effectively distributing strains and stresses between the concrete and steel components, CB optimizes overall structural efficiency and durability, making it an advantageous solution for applications requiring high strength and minimal deformation in modern construction (Ahmed, 2018; Uy, 2016; Szewczyk and Szumigała, 2021; Ban and Bradford, 2013; Jun et al., 2018; Thomann and Lebet, 2008; Lin et al., 2014; Thevendran et al., 2000; Li et al., 2024). Web openings in composite beams (CBs) allow for integrating services such as electrical conduits and HVAC systems, facilitating greater versatile and efficient building designs. This arrangement can reduce the building’s height and overall cost. However, these web openings can significantly alter stress distribution and load-carrying capacity, potentially affecting the beam’s overall strength and stability (Rex Donahey et al., 1988; Fahmy, 1996; Manuel Benitez et al., 1998; Guo et al., 2023).

Over the past 30 years, various practical and theoretical studies have explored the behavior of CB with web openings (CBO) (Jun et al., 2018; Thomann and Lebet, 2008; Lin et al., 2014; Thevendran et al., 2000; Li et al., 2024; Rex Donahey et al., 1988; Fahmy, 1996; Manuel Benitez et al., 1998; Guo et al., 2023; Ellobody and Young, 2014; Al-Dafafea et al., 2019; DURIF et al., 2021; Mastan et al., 2024; Du et al., 2021; Ellobody and Young, 2016). Balaguru et al. (Tsavdaridis and Galiatsatos, 2015) examined the repercussions of web openings on the structural performance of CBs, using experimental methods to analyze stress distribution and load capacity, and reported a 15%–20% reduction in load-carrying capacity for CBO compared to CB. Garg et al. (Darwin, 1990) conducted a numerical investigation on CBO and found a 12% decrease in overall stiffness and a 30% increase in deflection for beams. Stress concentration factors were found to be 2.5 times higher around the corners of the openings. Ali et al. (Verre, 2022) concluded that CB with larger openings show increased deflection and deformation under applied loads, and common failure modes include shear failure near the openings and localized bending. Martin Classen et al. (Liu et al., 2016) conducted numerical analysis and stated that finite element models closely match experimental results, demonstrating the validity of the modelling approach. David M. Todd et al. (Dassault Systèmes, 2015) stated that composite sections with web openings have significantly higher bending strength than corresponding non-composite sections. Opening length, height, and eccentricity variations significantly affect the ultimate strength.

To enhance CBO load-bearing capacity, Ellobody and Young (2014) conducted a finite element study on CB featuring both reinforced and unreinforced web openings and revealed that CBO reinforced with horizontal stiffeners had a significantly enhanced load-carrying capacity compared to those with unreinforced web openings. Al-Dafafea et al. (2019) Horizontal stiffeners were the most effective, while short stiffeners showed limited benefits and could cause cracks. Monoliteral stiffeners enhanced global stiffness by 12% for high openings and 5% for smaller openings, while double-sided stiffeners improved it by 50%. Both stiffeners significantly enhanced ultimate strength, but strain distribution showed complexities beyond Vierendeel’s theory, indicating the need for further finite element modelling. Evaluation of openings under local loads demonstrated differences in strength and behavior across opening shapes and stiffening methods, suggesting that current design codes may be overly conservative (Durif et al., 2021).

Few studies have investigated the flexural characteristics of CBOs strengthened with stiffeners, as shown in Table 1, as openings can potentially compromise composite beams’ flexural performance and load-bearing capacity (CBs).

Mastan et al. (2024) conducted a study involving twenty-eight numerical models of reinforced web openings analysed under three-point bending. They concluded that CBOs with longitudinal stiffeners (LS) and transverse stiffeners (TS) can achieve a load-bearing capacity comparable to that of CBs. The present study performed numerical analyses with varying shapes of opening and stiffener areas, comparing CB and CB with circular, rectangular and triangular openings. Additionally, experimental tests were conducted on scaled CB, CBO, and an optimized model to predict ultimate bearing capacity.

Numerical analysis has been carried out on CB, CBO, and CBOs with varying LS and TS areas. The steps involved in numerical analysis are presented in Figure 1. After obtaining the results from the numerical analysis, an experimental investigation was carried out on the optimised model, which was scaled down along with CB and CBC at a ratio of 1:0.32. Scaled-down models’ fabrication and casting were carried out, respectively. Lastly, a numerical, experimental, and theoretical comparison was conducted. This process is illustrated in Figure 1.

The CB is composed of a 4,000 mm-long ISMB 400 steel beam, a one mm-thick deck sheet, shear connectors 90 mm in length and 19 mm in width, and a concrete slab 120 mm thick and 1,200 mm wide. As specified in the AISC Steel Design Guide Series (Darwin, 1990), the stiffener must extend a distance l 1 = max a o 4 ; A r 3 2 t w Past the opening. Here A r = the cross-sectional profile of the stiffener. The data above are depicted in Figure 2 with a circular-shaped opening with a diameter of 0.5D (where D = depth of the steel beam), positioned one-third of the span from the closest edge of the support.

The geometry of the stiffeners is described in two distinct types: LS and TS. The configurations and placement of these stiffeners adopted from BS EN 1994-1-1 are shown in the accompanying Figure 3, and detailed models are provided in Table 2. LS runs parallel to the beam’s central axis, ensuring alignment along its length. TS are aligned orthogonally to the central axis, providing support across the width of the beam.

Finite Element Analysis (FEA) is widely used in civil engineering to model nonlinear behaviours of structures, such as masonry columns strengthened with composites. By incorporating material properties from tensile coupon tests, FEA helps accurately predict structural performance and failure mechanisms under different loading conditions (Verre, 2022).

The maximum capacity of the specimens was evaluated using nonlinear analysis using the dynamic implicit method (Liu et al., 2016). The CB was discretised using solid elements (C3D8R) for the concrete slab, head stud, and steel beam, truss elements (T3D2) for reinforcement, and shell elements (S4) for the deck sheet, respectively (Dassault Systèmes, 2015; Hibbitt et al., 2011; Dong et al., 2021; Numerical et al., 2023; Wang et al., 2019; Han-bing et al., 2010). Based on the material evaluations, Table 3 presents the mechanical attributes of the components in the CB. Tensile coupon evaluations were performed to determine the ultimate tensile and yield strength of all elements except concrete, following ISO 6892–1 guidelines, as shown in Table 3 (Anon ISO 6892-1, 2016). As detailed in Table 4, three cube-shaped compression tests were performed on the concrete for the composite beam sample.

Further interaction among components of CB is assigned following Mastan et al. (2023), as depicted in Supplementary Figure S1. Surface-to-surface interaction is adopted between the steel beam and deck sheet, deck sheet and deck slab, shear connectors and steel beam, respectively. Embedded interaction is adopted for shear connectors and reinforcement in the deck slab.

Following a detailed mesh sensitivity analysis, the ideal element sizes for the model were established to optimise accuracy and computing performance. The mesh refinement procedure was performed in multiple critical zones, including the solid area, rib area of the concrete slab, reinforcement, head studs, deck sheet, and steel beam. Reducing the element size enhanced the accuracy of the results, especially in areas with significant stress gradients, such as around web openings and stiffeners. The investigation indicated that further refinement of the mesh above a specific threshold resulted in little enhancements in accuracy while substantially elevating computing expenses. Consequently, the ideal element sizes were chosen for their capacity to ensure consistent convergence while reducing computational expenditure. The final element dimensions were 80 for the solid region, 30 for the rib region in the concrete slab, 100 for reinforcement, 30 for head studs, 20 for the deck sheet, and 40 for the steel beam, as mentioned in Supplementary Figures S2, S3, ensuring the necessary precision and convergence in the model.

For applying the load and boundary condition, the model is supported by end A, which is pinned, and end B, which rests on a roller. A displacement of 10 mm per step is applied at the mid-span of the slab, which is represented by Supplementary Figures S4–S7. The loading and boundary conditions are also depicted in Supplementary Figure S4.

The CB models were scaled down by a factor of 1:0.32 to maintain geometric similarity and proportion between the model and the prototype, as mentioned in Table 5, ensuring that the mechanical and structural properties of the system were accurately represented. Several key factors were considered when determining this scaling ratio.

• True-to-Scale Proportions: The 1:0.32 ratio was used to emulate the full-scale beam’s section size, concrete slab dimensions, and span within feasible limitations. This scaling ratio guarantees that the scaled model preserves the relative geometry of the components, such as the steel beam, concrete slab, and shear connectors.

The stud connectors and deck sheet, as detailed in Table 5, are illustrated in Figure 4. The shear connectors are welded onto the steel beam using a stud arc welding machine, also shown in Figure 4. After welding, the reinforcement is attached, and the concrete is cast, as illustrated.

As illustrated in Supplementary Figure S9, the CB samples were subjected to three-point flexural testing using a high-capacity hydraulic actuator of 1,000 kN. To evaluate the effectiveness of the test arrangement, the CB samples were initially exposed to 10% of the estimated peak load applied at 0.5 kN/s and subsequently unloaded. The patterns of cracks and distortions around the opening in the web were monitored and documented after each loading stage. The load was subsequently raised by 5 kN increments until failure was observed. The ultimate load for the standard sample CB was calculated utilising the basic plastic analysis approach described in Eurocode 4 (BSI 2004 BS EN 1994, 2004).

For the CB specimen with web openings, the ultimate load was calculated using the approach outlined by Park et al. (2003) Supplementary Figure S10 and Figures 5, 6 illustrate the positioning of sensors and experimental arrangement for the CB specimens, respectively. LDVTs were deployed at mid-span to monitor the deflections of the CB during testing. The strain around the opening in the web was measured using a 45° strain gauge rosette. Additionally, six strain sensors were installed along the length of the LS to assess the strain pattern along these stiffeners. In contrast, four strain measurement devices were mounted on the TS as shown in Supplementary Figure S10.

The three-point flexural test was chosen for this study to directly assess the bending response of composite beams under concentrated loading conditions. This test configuration allows for precise evaluation of the beam’s load-carrying capacity, deflection, and failure modes at mid-span, which is critical for understanding the structural behavior of composite beams with web openings and stiffeners. It was selected over alternative testing methods as it provides a clear and focused analysis of bending behavior with well-established data for comparison.

From the ASCE (Concrete, 1992) a mathematical model is commonly used to evaluate the load-carrying capacity of CBs with openings on the web. The shear force and bending capacity at the centroid of the web opening were determined by analyzing the cumulative impact of shear force and bending moment. The bending capacity and shear force at the centroid of the web opening were determined by examining the combined effects of the bending moment (Du et al., 2021) and shear force (Du et al., 2021). M u = φ M m M m V m γ 3 + 1 − 1 3 (1) V u =   φ V m   V m γ M m 3 + 1 − 1 3 (2) M u = maximum bending moment ; V u = maximum shear capacity M m = nominal bending moment at the centroid of web opening ; V m = nominal shear force at the centroid of web opening ; φ = factored shear force at the centroid of web opening = 0.85 for CB

Structural analysis allows for calculating the moment shear ratio (γ = M/V) relative to the centroid of the opening once its location is determined.

Finite element results, experimental results and theoretical results are compared, and findings are presented in this section.

This investigation assessed the effects of the area of stiffeners on the reinforcement of web openings in composite beams. A total of 18 models with various types of stiffeners and web openings were analyzed. The optimized models were scaled down using a 1:0.32 ratio. Three specimens were then designed and subjected to three-point flexural tests. According to the findings, the following conclusions were derived.

• Among the different shapes of openings considered, circular openings attain high load-bearing capacity, high stiffness and low-stress area percentage compared to rectangular and triangular openings.