Analysis of Structural Performance and Design Optimization of Prefabricated Cantilever Systems Under Traffic Loads

Abstract

Prefabricated cantilever systems (PCSs) are essential for mountainous road infrastructure, yet their structural behavior under traffic loads remains insufficiently studied. This study innovatively integrates scaled experiments, finite element simulations, and field test data to develop and validate a full-scale PCS model under extreme traffic conditions. The results reveal that the beam–column junction is highly vulnerable to stress concentrations, risking concrete cracking. To address this, a novel prestressed reinforcement design is proposed, optimizing rebar placement to reduce local stresses and enhance structural integrity. Ultimate load analysis confirms that prestressing improves stiffness, load resistance, and ductility. This study provides a systematic framework for PCS optimization, promoting its application in complex engineering environments.

1. Introduction

The construction of roads in mountainous regions is often complicated by steep and uneven terrain. To minimize excavation and earthworks, cantilever structures have gained widespread adoption as a key solution for roadbed support [1,2]. Conventionally, such structures are built using cast-in-place techniques, necessitating extensive on-site concrete pouring and curing. These processes not only demand substantial labor but also prolong construction schedules [3,4]. In response to these challenges, prefabricated cantilever structures (PCSs) have been developed as an efficient and practical alternative.

PCSs are highly adaptable to rugged mountainous landscapes and have become the preferred structural solution for mountain road construction [5,6,7]. As illustrated in Figure 1, these systems are composed of multiple prefabricated elements, including inner and outer longitudinal beams, cantilever beams, anchor rods, columns, retaining plates, and their corresponding foundations. Unlike conventional cast-in-place methods, these prefabricated components are produced in a controlled factory environment, ensuring consistent quality and precision. Once transported to the construction site, they are rapidly assembled, significantly improving construction speed while minimizing on-site labor requirements [8]. The assembly process also enhances overall safety, particularly in hazardous construction settings [9]. In contrast to conventional roadbeds, such as those formed by cutting slopes or filling embankments, prefabricated cantilever roadbeds are less influenced by topographical variations, leading to smoother roadway alignments and minimized elevation discrepancies, thereby demonstrating superior adaptability [10,11,12,13].

Several studies have explored the structural performance of PCSs [14,15,16,17,18,19,20,21,22,23,24]. Given that the beam–column connection plays a crucial role in the stability of these systems, research on their seismic behavior has shown that while various connection techniques have negligible effects on the longitudinal seismic response and risk distribution of primary structural elements, they substantially influence the transverse seismic behavior and associated risk patterns [14]. Regarding connection strategies for PCSs, Chen et al. [15] conducted both scaled physical tests and finite element (FE) simulations to examine the mechanical properties of bolted joints, demonstrating their effectiveness in preserving structural stability. Moreover, factors such as cantilever beam width and internal reinforcement grid configurations have a notable impact on the beams’ ultimate load-bearing capacity [16,17,18]. Different cantilever beam types also exhibit varying responses under extreme loading conditions [19,20]. Collectively, these findings offer essential insights for optimizing the design of prefabricated cantilever structures.

Studies have shown that vehicle loads play a critical role in the fatigue deterioration and potential failure of cantilever structures. Consequently, accurately evaluating and predicting the influence of traffic load distribution is essential for ensuring the long-term structural performance of these systems [22,23,24,25,26]. At present, the impact of vehicular loads on cantilever structures is primarily examined through numerical simulations [27,28,29,30,31]. Large-scale vehicle load testing is vital for assessing structural integrity and has been employed to detect potential deficiencies arising from material imperfections, construction inconsistencies, or long-term operational wear [32,33,34]. Pinkney et al. [32] analyzed the behavior of modular truss structures under varying vehicular loads, highlighting that different loading approaches significantly affect bridge performance. Additionally, the static and dynamic responses of structures subjected to various vehicle types, including heavy trucks and emergency vehicles, have been extensively investigated, demonstrating that both vehicle classification and loading configurations lead to distinct stress patterns within the structure [35,36,37,38]. Sun et al. [39] utilized FE modeling to simulate vehicle and lane loading conditions, concluding that FE-based approaches offer superior accuracy and cost efficiency compared to conventional static load testing. These studies provide crucial insights for refining vehicle load simulation methodologies. Since PCS represents an innovative type of cantilever structure, its distinctive connection mechanisms will inevitably influence internal stress distributions under traffic-induced forces. However, comprehensive research on the overall stress behavior of PCS under traffic loads remains lacking. In conclusion, while considerable advancements have been achieved in the analysis of cantilever structures, there is still a notable gap in understanding the performance of PCSs under extreme traffic loading conditions.

Thus, this research systematically fills the existing gap by integrating experimental studies with advanced simulations, focusing on the innovative design of PCSs. This study develops and validates a novel scaled-down finite element (FE) model for PCSs, established through rigorous experimental testing. Furthermore, a pioneering full-scale PCS FE model is created and validated using real paving test data, enabling a comprehensive analysis of structural responses under diverse traffic-induced effects. Three critical loading scenarios—external eccentric load, internal eccentric load, and full load—are meticulously evaluated. Additionally, the research introduces a groundbreaking approach to structural optimization by proposing the incorporation of prestressed rebars to mitigate stress concentrations in cantilever beams. These findings not only enhance our understanding of PCS behavior but also provide valuable insights for optimizing and practically applying PCS roadbeds in the challenging context of mountainous road construction, setting a new benchmark for future research in this domain.

2. Development and Validation of the Scaled-Down PCS Model

This section begins by confirming the model’s reliability through a combination of simulations and experimental tests on a scaled-down PCS. Following this, a full-scale model is developed to analyze and optimize the structural response under varying traffic load distributions.

2.1. Design and Testing of the Scaled-Down PCS

Drawing from the actual engineering structure depicted in Figure 1c, a 1:7.5 scaled-down model of the PCS was designed for both experimental testing and finite element (FE) simulation. As shown in Figure 2a, the scaled structure adopts a three-beam, two-span configuration. The geometric dimensions of the scaled model are presented in Figure 2b. The columns stand at a height of 1250 mm, with a rectangular cross-section measuring 200 mm × 200 mm. The cantilever beams feature a variable cross-section and have a total length of 1700 mm, with the smallest cross-section measuring 130 mm × 100 mm and the largest measuring 130 mm × 200 mm. The cantilever beams are spaced 600 mm apart. The connection between the cantilever beams and the columns is made using 12 bolts, each with a diameter of 16 mm, which are secured with double nuts to prevent loosening. To replicate the anchoring effect of the mountainous terrain on the structure, ground anchor beams were incorporated. These anchor beams are connected to ground anchor rods, which are secured with bolts and steel sections, and are inclined at an angle of 60° relative to the ground. In the constructed test structure, the cantilever beams and columns were made from C50 concrete, while the other components were constructed using C30 concrete. The experimental loading configuration is illustrated in Figure 3. A dual-channel electro-hydraulic servo system with a 100 T capacity was employed for loading. The load was applied to the three cantilever beams via a distribution beam, with the loading points located 45 cm from the column center. During the test, the loading rate was initially set at 0.5 mm/min, which was reduced to 0.2 mm/min once cracks were detected. The loading process was terminated when the structural load capacity fell to 85% of the peak load, after surpassing the peak value.

2.2. Scaled-Down PCS Model Construction

The finite element (FE) model developed for this study is depicted in Figure 4. Concrete, bolts, and nuts are modeled using eight-node hexahedral elements with reduced integration (C3D8R). Reinforcement bars are represented by two-node linear truss elements (T3D2), while anchor rods are modeled using two-node spatial linear beam elements (B31). Steel plates are discretized with four-node quadrilateral surface elements with reduced integration (S4R). The interaction between concrete (host element) and embedded elements (reinforcement and anchor rods) is simulated using the embedded element method. The base of the structure and the inclined anchor rods are fixed to prevent displacement and rotation.

Surface-to-surface contact is implemented to model interactions between key components, such as cantilever beams and columns, columns and retaining plates, outer longitudinal beams and columns, retaining plates and foundation, and steel plates and cantilever beams. A penalty friction coefficient of 0.4 is assigned to tangential behavior, while normal behavior is represented by hard contact. Binding constraints are applied to prevent relative displacement and rotation between columns and foundation, inner longitudinal beams and cantilever beams, bolts and prefabricated holes, and cantilever beams and external longitudinal beams. The anchor rods are embedded within the internal longitudinal beams.

In constructing the scaled model, particular attention was given to mesh size selection to ensure both accuracy and computational efficiency. Since ABAQUS 2020 uses dimensionless units, we defined the basic unit as meters. The mesh sizes were set as follows: 0.0034 m for bolts, 0.05 m for perforated steel plates, 0.1 m for anchor rods and reinforcement, and 0.02 m for concrete components. These mesh sizes were chosen to satisfy the dual requirements of precision and computational efficiency, enabling reliable finite element analysis while minimizing computational time.

Concrete specimens subjected to loading undergo both tensile cracking and compressive failure [15]. To model the stress–strain behavior of concrete, the Concrete Damage Plasticity (CDP) model available in ABAQUS is employed [40,41]. By incorporating damage variables, the CDP model effectively captures the mechanical response of concrete under monotonic, low-cycle, and dynamic loads, demonstrating favorable convergence. Figure 5 presents the uniaxial compressive and tensile stress–strain curves, as specified by the GB50011-2010 [42] Concrete Structure Design Code. The compressive stress–strain relationship is defined as follows:

$${\sigma }_c=\left\{\begin{array}{cc}{\displaystyle \frac{nx}{n-1+x}}& x\le 1\\ {\displaystyle \frac{x}{{\alpha }_c{(x-1)}^2+x}}& x>1\end{array}\right.$$

(1)

$$x={\displaystyle \frac{\epsilon }{{\epsilon }_{c,r}}},{\rho }_c={\displaystyle \frac{f_{c,r}}{E_c{\epsilon }_{c,r}}},n={\displaystyle \frac{1}{1-{\rho }_c}}$$

(2)

where ${\sigma }_c$ is the tensile stress; $f_{c,r}$ is the uniaxial tensile strength of concrete; ${\epsilon }_{c,r}$ is the peak tensile strain corresponding to $f_{c,r}$; and ${\alpha }_c$ is the parameter for the descending branch of the uniaxial tensile stress–strain curve. The tensile stress–strain relationship is given by

$${\sigma }_t=\left\{\begin{array}{cc}1.2x-0.6x^6& x\le 1\\ {\displaystyle \frac{x}{{\alpha }_t{(x-1)}^{1.7}+x}}& x>1\end{array}\right.$$

(3)

$$x={\displaystyle \frac{\epsilon }{{\epsilon }_{t,r}}},{\rho }_t={\displaystyle \frac{f_{t,r}}{E_c{\epsilon }_{t,r}}}$$

(4)

where ${\sigma }_t$ is the tensile stress; $f_{t,r}$ is the uniaxial tensile strength of concrete; ${\epsilon }_{t,r}$ is the peak tensile strain corresponding to $f_{t,r}$; and ${\alpha }_t$ is the parameter for the descending branch of the uniaxial tensile stress–strain curve. In the CDP model, concrete failure is represented by tensile cracking following tensile yield and compressive crushing following compressive yield. The yield function incorporates equivalent plastic strain ($\widetilde{{\epsilon }_t^{pl}}$) and compressive equivalent plastic strain ($\widetilde{{\epsilon }_c^{pl}}$). The model uses non-elastic strains, including tensile cracking strain ($\widetilde{{\epsilon }_t^{ck}}$) and compressive inelastic strain ($\widetilde{{\epsilon }_c^{in}}$). The compressive inelastic strain, as illustrated in Figure 5a, is computed as follows:

$$\left\{\begin{array}{c}\widetilde{{\epsilon }_c^{in}}={\epsilon }_c-{\epsilon }_{0c}^{el}\\ {\epsilon }_{0c}^{el}={\displaystyle \frac{{\sigma }_c}{E_0}}\\ \widetilde{{\epsilon }_c^{pl}}=\widetilde{{\epsilon }_c^{in}}-{\displaystyle \frac{d_c}{(1-d_c)}}\times {\displaystyle \frac{{\sigma }_c}{E_0}}\end{array}\right.$$

(5)

where ${\sigma }_c$ and ${\epsilon }_c$ are the compressive stress and strain, respectively; ${\epsilon }_{0c}^{el}$ is the elastic deformation under compression; $\widetilde{{\epsilon }_c^{pl}}$ is the equivalent plastic strain under compression; and $d_c$ is the damage factor. The tensile inelastic strain, shown in Figure 5b, is calculated as follows:

$$\left\{\begin{array}{c}\widetilde{{\epsilon }_t^{in}}={\epsilon }_t-{\epsilon }_{0t}^{el}\\ {\epsilon }_{0t}^{el}={\displaystyle \frac{{\sigma }_t}{E_0}}\\ \widetilde{{\epsilon }_t^{pl}}=\widetilde{{\epsilon }_t^{in}}-{\displaystyle \frac{d_t}{(1-d_t)}}\times {\displaystyle \frac{{\sigma }_t}{E_0}}\end{array}\right.$$

(6)

where ${\sigma }_t$ and ${\epsilon }_t$ are the tensile stress and strain, respectively, ${\epsilon }_{0t}^{el}$ is the elastic deformation under tension, $\widetilde{{\epsilon }_t^{pl}}$ is the equivalent plastic strain under tension, $d_t$ is the damage factor. The method for calculating the concrete damage factor follows Lee and Fenves [43], expressed as follows:

$$d_{torc}=1-\sqrt{{\displaystyle \frac{{\sigma }_{torc}}{E_0\cdot {\epsilon }_{torc}}}}$$

(7)

where $d_{torc}$ represents the tensile or compressive damage factor, ${\sigma }_{torc}$ is the tensile or compressive stress of the concrete, and ${\epsilon }_{torc}$ is the tensile or compressive strain of the concrete.

In this study, C30 concrete is modeled as an elastic material, whereas C50 concrete is treated as elasto-plastic. For C50 concrete, the compressive and tensile strains are inserted into Equations (1)–(6) to derive the corresponding stresses, thereby accurately representing its stress–strain relationship. Additionally, Equation (7) is used to calculate the damage factors associated with non-elastic strains, which account for the material’s degradation under loading conditions. The detailed damage data for C50 concrete are summarized in

Table 1. Parameters of the CDP model for C50 concrete.

C50 Concrete
Compressive Behavior			Tensile Behavior
Stress (MPa)	Cracking Strain	Damaged Factor	Stress (MPa)	Inelastic Strain	Damaged Factor
38.19948	0	0	2.34521	0	0
43.10004	0.00009	0.03588	2.69435	0.00001	0.0381
46.86577	0.00018	0.05958	2.99498	0.00001	0.05135
49.21496	0.0003	0.09141	3.21325	0.00002	0.07359
50	0.00047	0.13119	3.3	0.00003	0.10934
46.18227	0.00097	0.23777	3.08616	0.00005	0.17877
42.67332	0.00126	0.29604	2.73549	0.00007	0.25975
38.95815	0.00156	0.35185	2.39215	0.00009	0.33492
24.30413	0.00294	0.56056	2.09564	0.00011	0.40014
22.32143	0.00319	0.58953	1.84995	0.00014	0.45551
20.58501	0.00344	0.61531	0.98002	0.00025	0.67277
19.06016	0.00367	0.63835	0.91809	0.00027	0.69023
12.92085	0.005	0.73606	0.86384	0.00028	0.70585
12.2334	0.00521	0.74764	0.81598	0.00029	0.71989
11.60991	0.00542	0.75829	0.77345	0.00031	0.73259
11.0424	0.00563	0.76811	0.73544	0.00032	0.74411
10.5241	0.00584	0.77718	0.70128	0.00034	0.75463
10.04921	0.00604	0.78559	0.48623	0.00047	0.82462
9.61277	0.00625	0.7934	0.47243	0.00048	0.82936
9.21053	0.00645	0.80068	0.4595	0.0005	0.83382
8.17443	0.00706	0.81979	0.30496	0.00076	0.88947
7.87662	0.00726	0.82539	0.167	0.00153	0.94216
3.63523	0.01333	0.91146	0.15656	0.00166	0.9462
2.12355	0.02106	0.94601	0.14757	0.00179	0.94966
1.92258	0.02298	0.95082	0.13974	0.00192	0.95268

and

Table 2. Parameters of the CDP model for concrete.

Expansion Angle	Eccentricity	fb0/fc0	K	υ
35°	0.1	1.16	0.6667	0.0005

, where the compressive strength ($f_{c,r}$) is 50 MPa with a corresponding strain (${\epsilon }_{c,r}$) of 0.00192, and the tensile strength ($f_{t,r}$) is 3.3 MPa with a strain (${\epsilon }_{t,r}$) of 0.000128.

The steel plates in the structure are made of Q345 steel with a yield strength of 345 MPa, and the reinforcing bars are HRB400. Both the steel plates and reinforcing bars are modeled using a bilinear hardening model, as follows:

$${\sigma }_s=\left\{\begin{array}{cc}{E}_s\cdot {\epsilon }_s& 0\le {\epsilon }_s\le {\epsilon }_y\\ k\cdot E_s\cdot {\epsilon }_s& {\epsilon }_y<{\epsilon }_s\le {\epsilon }_u\end{array}\right.$$

(8)

where ${\sigma }_s$ is the stress in the steel; ${\epsilon }_s$ is the strain in the steel; $E_s$ is the initial elastic modulus of the steel; and $k$ is the ratio of the hardening modulus to the initial elastic modulus, taken as 0.01 in this study. The material parameters used are summarized in

Table 3. Used material parameters.

Material	Density (kg/m3)	Elastic Modulus (GPa)	Poisson’s Ratio
C50	2400	35	0.2
C30	2400	30	0.2
Q345	7850	210	0.3
HRB400	7850	200	0.3

. Figure 6 depicts the stress–strain curve of the bilinear hardening model, which includes an elastic segment (OA) and a hardening segment (AB). The Von Mises yield criterion is used to define the yield stress as follows:

$${\sigma }_S=\sqrt{{\displaystyle \frac{1}{2}}\cdot [{({\sigma }_1-{\sigma }_2)}^2+{({\sigma }_2-{\sigma }_3)}^2+{({\sigma }_3-{\sigma }_1)}^2]}$$

(9)

where ${\sigma }_s$ is the yield stress; and ${\sigma }_1$, ${\sigma }_2,$ and ${\sigma }_3$ are the first, second, and third principal stresses, respectively.

2.3. Scaled-Down PCS Model Validation

Figure 7 compares the simulated and experimental results for the scaled-down PCS model. The load–displacement curve in Figure 7a shows that the simulation results closely match the experimental data, especially before reaching the maximum load capacity. The ultimate load capacities obtained from the simulation and experiment are 144.1 kN and 142 kN, respectively, with a simulation error of only 1.47%. Figure 7b,c illustrates the damage distributions of the cantilever beam, as observed experimentally and in the simulation, respectively. Under vertical displacement loading, cracks propagate from the loading point toward the base of the cantilever beam, resulting in failure. The simulation results accurately reflect the experimental damage pattern. The high degree of similarity between the simulated and experimental results validates the accuracy of the developed model, which effectively captures the actual stress and damage behavior of the structure. Therefore, the material models and boundary conditions applied in the scaled-down model can be extended to the full-scale PCS model.

3. Full-Scale PCS Model Development and Validation

This section details the stress measurement process conducted during the paving of the full-scale PCS roadbed structure. Following this, the construction and validation of the full-scale model, which is derived from the scaled-down model and experimental data, are discussed.

3.1. Full-Scale PCS Testing

Figure 8a depicts the onsite paving procedure for the full-scale PCS roadbed. Figure 8b illustrates the stress measurement locations on the cantilever beams. Six critical measurement points are identified in Figure 8c as follows: the two beam ends, the junction between varying and constant cross-sections, and the midpoint of the varying cross-section. Stress concentrations are anticipated at the beam ends and at the junction between varying and constant cross-sections. Moreover, notable bending and shear stresses may develop at the midpoint of the varying cross-section due to the paving loads.

3.2. Full-Scale PCS Development

Figure 9 shows the developed FE model for the full-scale PCS roadbed. To minimize the impact of mountain size on the structural response, the mountain body is modeled to be three times the size of the PCS. Figure 9b,c illustrate that the interactions between the cantilever structures, concrete, and reinforcement in the model are consistent with those in the scaled-down model. The bottom and corners of the mountain body are fixed to restrict rotation and displacement, as depicted in Figure 9d.

3.3. Full-Scale PCS Model Validation

Figure 10a presents the stress distribution derived from the FE simulation, while Figure 10b compares the stress measurements obtained from both the FE simulation and onsite experiments. The stress values at the measured points show strong consistency between the simulation and experimental results. Although the FE simulation generally predicts slightly lower stress values compared to the onsite measurements, the similar trends and closely aligned stress values validate the model’s accuracy. It is important to note that variations in material properties and uncertainties in boundary conditions can impact both field test outcomes and numerical simulations [44,45]. Future studies should incorporate statistical considerations to account for these variabilities, enhancing the reliability of the findings. Discrepancies between the test and simulation results may also be attributed to environmental factors, such as temperature, humidity, and wind, which can influence onsite tests. Incorporating probabilistic analyses in future work could help evaluate the effects of load variations and diverse environmental conditions on the structure by integrating long-term monitoring data, providing deeper insights into the fatigue and durability of prefabricated cantilever systems under real-world traffic conditions.

4. Full-Scale PCS Response Analysis and Optimization Under Traffic Load

4.1. Structural Response Analysis

The following three distinct traffic load conditions are considered in this study: (1) external eccentric loading, (2) internal eccentric loading, and (3) full loading. Eccentric loading is considered the most critical stress condition [46,47,48,49]. External eccentric loading applies the load to the outer side of the cantilever, away from the mountain, resulting in outward bending moments. In contrast, internal eccentric loading places the load on the cantilever’s inner side, near the mountain, generating inward bending moments. Full loading distributes the load evenly across the entire structure, representing the maximum load the structure experiences during regular operation. Figure 11a illustrates the schematic of load distributions for these three scenarios. The lane load and vehicle load are determined according to highway design requirements, as shown in Figure 11b,c. The final applied traffic loads for each scenario are presented in Figure 11d through Figure 11f.

In the PCS, the cantilever beams are responsible for bearing a variety of loads, including both vertical and horizontal forces, making them the primary load-bearing components. On the other hand, the columns predominantly resist axial forces and are less influenced by these loads. As a result, the analysis is primarily focused on the cantilever beams.

Figure 12, Figure 13 and Figure 14 present the stress responses of the cantilever beam under the following three distinct loading conditions: external eccentric loading, internal eccentric loading, and full loading. Figure 12 illustrates that, under external eccentric loading, the concrete stress in the cantilever beam is predominantly concentrated at the joint between the cantilever beam and the column, with lower stress observed at the two ends of the cantilever beam. The maximum concrete stress value is 1.9 MPa (Figure 12a). The maximum stress in the reinforcement mesh is concentrated at the bottom of the beam–column joint, with a peak stress of 11.2 MPa (Figure 12b). Bolts exhibit significant stress concentrations at their top sections, with a maximum stress value of 18.7 MPa (Figure 12c).

Figure 13 shows that, compared to external eccentric loading, internal eccentric loading results in higher stress values and a larger stress area in the concrete at the cantilever-column joint, with a maximum principal stress of 2.6 MPa (Figure 13a). The stress in the reinforcement mesh and bolts significantly increases, reaching the maximum principal stresses of 90.5 MPa (Figure 13b) and 77.8 MPa (Figure 13c), respectively. These findings indicate that internal eccentric loading exerts a greater impact on the structure’s stability, highlighting its critical role in the overall performance of the cantilever beam. It is noteworthy that under external eccentric loading, the stress in the concrete is lower than that under internal eccentric loading. This is because the internal section of the beam is longer than the external section. Therefore, even though the internal section is adjacent to the mountainside, the longer internal side leads to higher stress in the internal part, whereas the external section experiences comparatively lower stress under external eccentric loading.

Under full loading, the concrete stress distribution in the cantilever beam is uniform across its middle and upper sections, with the highest principal stress observed at the bottom of the joint, reaching a peak value of 2.9 MPa (Figure 14a). The maximum stress in the reinforcement mesh is similarly concentrated at the bottom of the joint, with a value of 24.4 MPa (Figure 14b). The bolts experience significant stress at their top sections, with the maximum stress reaching 59.2 MPa (Figure 14c).

The stress analysis across different loading conditions reveals that the beam–column joint experiences the highest stress concentration under internal eccentric loading. In this case, both the reinforcement mesh and bolts exhibit significantly higher stress compared to other scenarios, while the concrete also endures considerable stress. However, given that the reinforcement’s yield strength is 345 MPa, the structural reinforcement remains highly reliable. On the other hand, the maximum tensile strength of C50 concrete, at only 4.1 MPa [50,51,52], indicates that the safety margin of the concrete is considerably lower than that of the reinforcement. Consequently, to enhance the structure’s performance under long-term traffic loads, further optimization efforts are incorporated in this study.

4.2. Structural Optimization Based on the Numerical Design

The analysis reveals that the concrete in the cantilever beam experiences relatively high stress, with significant stress concentration and limited safety margins. To improve these safety margins, it is essential to reduce stress in the concrete. One effective solution is the addition of prestressed rebars, which help improve stress distribution and mitigate stress concentration. Prestressing not only counteracts tensile stress in the concrete but also significantly enhances the structure’s crack resistance and overall stiffness [53,54,55]. Based on the findings from Section 3.1, where concrete stress is mainly concentrated at the bottom of the inner end of the beam, three configurations for prestressed rebar placement are proposed, which are illustrated in Figure 15, as follows: Case 1, where prestressed rebars are placed at the upper inner side of the cantilever beam; Case 2, where prestressed rebars are positioned on both the upper inner and outer sides of the cantilever beam; and Case 3, which builds on Case 2 with additional prestressed rebars at the bottom of the inner beam. Each configuration includes six rebars with a 28 mm diameter, spaced 100 mm apart (D28@100), and prestressed to 300 MPa. The main reason for adopting a very low prestress level (300 MPa) in prestressed elements is to prevent the occurrence of unacceptable cracks due to excessive stresses on the components during transportation [22,56]. A lower prestress level reduces the risk of damage during transportation, especially when the components are subjected to transportation vibrations or other external loads. The reduced stress helps effectively decrease stress concentration in the material, thus preventing cracks from forming.

Figure 16, Figure 17 and Figure 18 display the stress distribution of both the original and optimized cantilever beams with prestressed rebars under the following three loading conditions: external eccentric load, internal eccentric load, and full load. Figure 16 illustrates the impact of prestressed rebars on the stress distribution under external eccentric loading. The addition of prestressed rebars significantly reduces the tensile stress at the top of the inner side of the beam and the beam–column joint. In Case 1, where prestressed rebars are placed only on the inner side of the beam, the maximum stress is reduced to 1.5 MPa (Figure 16b), although stress concentration remains, particularly at the top of the beam–column joint. In Cases 2 and 3, where rebars are added to both the inner and outer top surfaces, the stress concentration is notably decreased, with the maximum stress values dropping to 1.4 MPa (Figure 16c) and 1.3 MPa (Figure 16d), respectively. This demonstrates that positioning prestressed rebars on both the inner and outer top surfaces significantly reduces stress concentration, thereby improving the bending and shear resistance of the cantilever beam under external eccentric loading.

Figure 17 demonstrates that the addition of prestressed rebars greatly reduces stress levels in the cantilever beam, particularly at the midsection and the bottom of the beam–column joint, under internal eccentric loading. In Case 1, where prestressed rebars are positioned only on the inner side, the high-stress area at the beam–column joint is reduced, but a concentrated stress zone still exists (Figure 17b). In contrast, Cases 2 and 3, where additional prestressed rebars are placed on the outer beam, further decrease the high-stress area, improving the overall uniformity of stress distribution and reducing the maximum stress. These cases show only a small high-stress region at the bottom of the inner beam–column joint (Figure 17c,d). This suggests that positioning prestressed rebars on both the upper inner and outer sides of the beam significantly improves the uniformity of stress distribution under internal eccentric loading.

Figure 18 illustrates that the inclusion of prestressed rebars significantly lowers stress levels at both the bottom of the outer beam–column joint and the tensile zone at the inner beam–column joint under full loading. In Case 1, with prestressed rebars only on the inner side, the maximum stress is reduced to 2.8 MPa (Figure 18b), but the stress in the high-stress area remains notable. In Case 2, the further optimization of prestressed rebar placement lowers the maximum stress to 2.7 MPa (Figure 18c), alleviating stress concentration and achieving a more uniform distribution of tensile stress, although some significant stress remains at the outer beam–column joint. In Case 3, the addition of prestressed rebars at the bottom of the inner beam further reduces maximum stress to 2.6 MPa (Figure 18d), completely eliminating the high-stress area at the outer beam–column joint. These findings demonstrate that positioning prestressed rebars on the upper inner and outer surfaces, as well as the bottom inner beam, optimizes stress distribution under full loading. The results above demonstrate that placing prestressed rebars on the top inner and outer surfaces of the cantilever beam (Case 2) significantly improves stress distribution under both internal and external eccentric loading.

4.3. Load-Bearing Capacity of Optimized Structures

The incorporation of prestressed rebars and the application of prestressing have a substantial effect on the load-bearing capacity of cantilever structures [57,58,59]. To fully assess the performance of the optimized cantilever structure, further simulation analysis of its load-bearing capacity was conducted. As illustrated in Figure 19, a displacement control method was employed to apply a downward vertical displacement at the most unfavorable position of the entire structure. This method facilitates the investigation of the load-bearing capacity and overall stability of the structure under various optimization scenarios.

Figure 20 presents the displacement-load curves and final damage distribution for both the original and optimized cantilever structures. As shown in Figure 20a, during the initial loading phase, the original model exhibits the smallest slope, reflecting its lowest initial stiffness. With the addition of prestressed rebars, the initial stiffness of Cases 1, 2, and 3 gradually increases. Notably, Cases 2 and 3 demonstrate significantly higher initial stiffness compared to both the original model and Case 1, emphasizing the positive effect of prestressed rebars on the overall stiffness of the structure. Regarding peak load, the original model has the lowest capacity, while the optimized models show progressive improvements, with Case 3 achieving the highest peak load, indicating a substantial enhancement in load-bearing capacity. Finally, the analysis of the descending segment reveals that the original model’s load-bearing capacity declines the fastest, while the optimized models exhibit a more gradual reduction. Cases 2 and 3, which have the lowest descending segments, suggest that the optimized structures offer superior ductility. The damage distribution shown in Figure 20b reveals that, in the original model, damage is primarily concentrated at the upper part of the inner beam and the beam–column joint. In Case 1, the prestressed rebars placed at the upper inner side of the beam significantly alleviate damage at the beam–column joint, demonstrating the effectiveness of prestressed rebars in mitigating stress concentration and enhancing the overall structural integrity. In Cases 2 and 3, a further reduction in the damage of the beam–column joint is observed, indicating that placing prestressed rebars on both the inner and outer sides of the beam effectively reduces damage at this critical junction, improves stress distribution, and prevents the propagation of damage. In summary, the analysis demonstrates that placing prestressed rebars at the top of the cantilever beam significantly enhances the overall structural stiffness, peak load capacity, and ductility, while also delaying the onset of structural damage. This optimization leads to a more balanced stress distribution, thereby playing a vital role in improving the stability and safety of the structure.

5. Conclusions

This research investigates the performance of a novel prefabricated cantilever structure (PCS) roadbed subjected to various traffic load distributions, incorporating laboratory and field tests along with numerical simulations, all based on an actual mountainous road construction case. The mechanical behavior of the structure was examined under the following three extreme vehicle loading conditions: (1) external eccentric load, (2) internal eccentric load, and (3) full load. Based on the findings, the structure underwent optimization by incorporating prestressed rebars into the cantilever beams. The key conclusions are as follows:

(1)

The prefabricated cantilever structure model presented in this study effectively replicates the real-world stress and failure characteristics of the structure. In the scaled model, the load-bearing capacity curves derived from both simulations and laboratory experiments demonstrate a high degree of consistency. For the full-scale model, comparisons between simulated and actual stresses on the cantilever beams during paving further corroborate the model’s accuracy, verifying its robustness and reliability.

(2)

Under varying vehicular loads, the connection between the cantilever beam and the column is identified as a critical weak point, characterized by significant stress concentration. While the reinforcements in this region endure high stress, the tensile strength of the concrete is relatively low. However, once cracking occurs in the concrete, the tensile stress will be borne by the steel bars, and failure could result from either the yielding of the steel bars or the failure of the concrete compression surface, potentially leading to the collapse of the entire structure.

(3)

The incorporation of prestressed rebars in the tension zone of the cantilever beams optimizes the internal stress distribution, alleviating stress concentration, and leading to a more uniform overall stress distribution, thus improving the structural performance. Positioning prestressed rebars at both the inner and outer top surfaces of the cantilever beams significantly enhances stress distribution under both external and internal eccentric loads. Additionally, adding prestressed rebars at the bottom of the inner cantilever beam further refines stress distribution under full loading conditions.

(4)

Incorporating prestressed rebars into the cantilever beams improves the structure’s overall stiffness, load-bearing capacity, and ductility, while delaying the initiation of structural damage. Simulations of the optimized structure under ultimate load conditions show that the inclusion of prestressed rebars, especially at the top inner and outer cantilever beam surfaces, notably enhances the initial slope and peak load of the displacement-load curve, along with a reduced descending slope. The damage distribution analysis indicates that adding prestressed rebars, particularly at the top inner and outer sides, significantly mitigates damage at critical beam–column connections, improving stress distribution and preventing rapid damage propagation.

This study presents a reliable approach for assessing and optimizing the performance of innovative prefabricated cantilever structures, significantly advancing their practical engineering application. The proposed structure offers a cost-effective and reliable solution for highway construction in mountainous regions with challenging terrain. By simplifying the construction process and reducing costs, this structure provides a practical alternative for engineers facing complex topographical constraints. However, certain limitations exist in the current research. For instance, variations in material properties and uncertainties in boundary conditions were not considered. Additionally, the long-term effects of environmental factors on structural performance were not addressed. Future research will build on the models developed here to explore the structure’s performance under various factors, such as environmental conditions, materials, and geometry. Furthermore, our finite element model will incorporate seismic effects and conduct partial structural failure analyses, such as stability assessments after anchor rod failure, to provide further support for the safety and reliability of the structure. Long-term monitoring of actual structures will also be conducted to gain deeper insights into their fatigue and durability performance.