Cyclic Behavior Enhancement of Existing RC Bridge Columns with UHPC Jackets: Experimental and Parametric Study on Jacket Thickness

Abstract

Ultra-high-performance concrete (UHPC) jackets present a promising solution for enhancing the seismic resilience of seismically deficient reinforced concrete (RC) bridge columns. This study investigates jacket thickness optimization through integrated experimental and numerical analyses. Quasi-static cyclic tests on a control column and a specimen retrofitted with a 30-mm UHPC jacket over the plastic hinge region demonstrated significant performance improvements: delayed damage initiation, controlled cracking, a 24.6% increase in lateral load capacity, 139.5% higher energy dissipation at 3% drift, and mitigated post-peak strength degradation. A validated OpenSees numerical model accurately replicated this behavior and enabled parametric studies of 15-mm, 30-mm, and 45-mm jackets. Results identified the 30-mm thickness as optimal, balancing substantial gains in lateral strength (~12% higher than other thicknesses), ductility, and energy dissipation while avoiding premature failure modes—insufficient confinement in the 15-mm jacket and strain incompatibility-induced brittle failure in the 45-mm jacket. These findings provide quantitative design guidance, establishing 30 mm as the recommended thickness for efficient seismic retrofitting of existing RC bridge columns.

1. Introduction

Recent seismic events revealed critical vulnerabilities in existing reinforced concrete (RC) bridge infrastructure, particularly in columns designed to outdated standards [1,2]. Historical data indicate that approximately 82% of bridge failures stem from deficiencies in the plastic hinge zone, manifested primarily as inadequate flexural strength and ductility [3,4]. These observations have spurred significant advances in retrofitting technologies, with ultra-high-performance concrete (UHPC) emerging as a promising solution due to its exceptional combination of mechanical properties and durability characteristics [5,6].

UHPC achieves compressive strengths exceeding 120 MPa through optimized particle packing distributions and water-to-binder ratios below 0.20 [7,8]. The incorporation of 2–3% volumetric steel fibers enables tensile strain-hardening behavior, effectively restricting crack widths to below 50 μm even at tensile strain up to 0.5% [9,10]. These exceptional properties derive from three fundamental characteristics: (1) a silica fume-enhanced dense microstructure, (2) an ultra-low permeability matrix, and (3) efficient fiber bridging mechanisms through high interfacial bond strength [11,12,13]. Consequently, UHPC demonstrates superior performance in seismic retrofitting applications where conventional materials exhibit insufficient deformation capacity and premature brittle failure [14,15].

Contemporary research substantiates ultra-high-performance concrete’s (UHPC) multifaceted retrofitting capabilities through systematically integrated technological advancements. Lateral load-carrying capacity of UHPC jackets achieves approximately 40% enhancement through optimized fiber alignment [16], while torsion-resistant systems demonstrate 40% improved energy dissipation using steel grid-reinforced mortar composites under bidirectional loading [17]. Recent studies further document 110–125% displacement capacity increases with UHPC jackets [18,19,20], and 80–90% reductions in concrete spalling damage through stay-in-place formwork applications [21,22]. UHPC’s damage resilience is demonstrated by encasement techniques restoring 92% stiffness in degraded T-beams while maintaining >1.5 million fatigue cycles with <8% residual strain [23]. These structural innovations stem from computational material science breakthroughs, particularly machine learning-optimized hybrid fibers (0.9% micro-steel + 1.2% PVA) achieving triple-impact resistance via controlled phase interactions [24]. Despite these advancements, critical knowledge gaps persist regarding optimal retrofit layer thickness under cyclic loading—a key parameter governing interfacial stress distribution and seismic damage evolution. Practical implementation challenges, including stringent surface preparation and uncertain long-term bond performance [25,26], underscore the need for simplified application methodologies that harmonize structural efficacy with constructability.

Compared to conventional retrofitting methods like FRP wrapping or steel jacketing, UHPC jackets offer superior structural and durability advantages. FRP, though lightweight and easily installed, suffers from adhesive dependency, environmental degradation, and low ductility under cyclic loads [4,19]. Steel jacketing enhances confinement but is hindered by corrosion risks, added weight, and complex installation. UHPC combines passive confinement and active structural contribution via ultra-high compressive/tensile strength. Steel fibers enable strain-hardening, post-cracking ductility, and improved seismic energy dissipation. Surface texturing enhances mechanical interlocking, reducing reliance on chemical bonds [25]. Its ultra-low permeability and corrosion resistance ensure long-term durability in aggressive environments like coastal or freeze-thaw zones.

The critical research gap lies in determining optimal UHPC jacket configurations for field implementation. While previous investigations [27,28,29] have established fundamental performance benefits, no systematic methodology exists for determining the most effective jacket thickness-to-performance ratio. This study addresses this gap through an integrated experimental and computational approach, examining three jacket thickness variations (15-mm, 30-mm, 45-mm) through quasi-static cyclic loading tests and nonlinear finite element analysis. The research provides quantitative guidelines for engineers seeking to optimize retrofit effectiveness while considering practical construction constraints.

2. Experimental Program

2.1. Specimen Description

Two 1/5-scaled cantilever column specimens, including one typical bridge column designed according to 1990s seismic-deficient designs [30] as a benchmark and another strengthened with a 30-mm-thick UHPC jacket applied to cover the potential plastic hinge region at the column base (abbreviated as U30), as shown in Figure 1a, were designed and fabricated. The scaled column specimen measured 1400 mm in height with a 300 × 300 mm² cross-section, supported by a footing block and capped with a 1200 × 400 × 400 mm³ (L × W × H) pier stub, maintaining 1:5 geometric scaling to represent typical highway bridge substructures as shown in Figure 1b–d, equivalent longitudinal reinforcement ratio (ρ_l = 1.51%), matching transverse reinforcement ratio (ρ_s = 1.02%), as tabulated in

Table 1. Specimen list.

Specimen	UHPC Jacket		Column Section (mm)	ρl (%)	ρs (%)
	t (mm)	h (mm)
RC	–	–	300 × 300	1.51	1.02
U30	30	400

. The 400-mm-high UHPC jacket in the retrofitted specimen was specifically designed to cover the plastic hinge zone while satisfying the 150 mm minimum plastic hinge length requirement specified in current seismic provisions [31], with the 30 mm thickness selected as a practical solution balancing structural enhancement with construction feasibility. To ensure valid comparisons, all structural parameters except for the UHPC jacket thickness were held constant across specimens, including column geometry, reinforcement configuration, material properties, and axial load levels.

2.2. Materials

The UHPC mixture comprised cement, water, a polycarboxylate-based high-range water reducer, and steel fibers (

Table 2. Properties of steel fiber.

Length (mm)	Diameter (mm)	Tensile Strength (MPa)	Shape	Surface
13	0.2	1560	Straight	Smooth

). The composite mixture incorporated 2% steel fibers by volume, with a specific mix proportion provided in

Table 3. Mix proportion of UHPC (kg/m³).

W/B 1	Water	Cement	Fly Ash	Silica Fume	Quartz Sand	Steel Fiber	HPWR 2
0.17	170	730	110	160	1032	156	30

¹ W/B is the water-to-binder ratio. ² HPWR is the high-performance water reducer.

. Mechanical blending was achieved using a single horizontal-shaft forced mixer. Except for the structural components, standardized test specimens were fabricated for material properties, undergoing identical curing conditions. Tensile properties of UHPC were determined using dog-bone shaped specimens conforming to Chinese standards [32], tested under uniaxial loading conditions as illustrated in Figure 2a,b. The resultant stress-strain relationship, as shown in Figure 2c, demonstrated significant post-cracking ductility, with steel fiber bridging mechanisms producing distinct strain-hardening characteristics before the strain corresponding to the ultimate tensile strength. As a result, the mean first cracking strength (f_cr-uhpc) and ultimate tensile strengths (f_t-uhpc) are 5.26 MPa and 7.81 MPa, respectively. Parallel compressive testing of 100 mm cubic specimens yielded an average compressive strength (f_c-uhpc) of 127.9 MPa and elastic modulus (E_uhpc) of 5.1 × 10⁴ MPa, confirming the material’s exceptional mechanical properties. Additionally, the UHPC mixture demonstrated superior durability characteristics, exhibiting restrained autogenous shrinkage (300 με) and low creep coefficients (0.4) due to its optimized particle packing and silica fume enhancement. Furthermore, its dense microstructure provided exceptional corrosion resistance, limiting chloride diffusion to less than 0.02% of conventional concrete values through crack-width control below 50 μm under tensile strain.

The bridge column specimens in this experimental program employed C40 concrete, with its complete material composition specified in

Table 4. Mix proportion of concrete (kg/m³).

W/B 1	Water	Cement	Fly Ash	Silica Fume	Sand	Gravel	HPWR 2
0.34	139	310	80	20	789	1089	4.1

¹ W/B is the water-to-binder ratio. ² HPWR is the high-performance water reducer.

. Standard 150 mm cubic specimens were tested to evaluate the concrete’s mechanical properties, demonstrating an average cubic compression strength of 57.0 MPa with a coefficient of variation of 2.0%. For reinforcement, the columns incorporated HRB400-grade steel bars, utilizing 12 mm diameter deformed bars as longitudinal reinforcement and 6 mm diameter smooth bars for lateral confinement. The mechanical properties of these reinforcing steels, including yield strength and ultimate tensile strength, are comprehensively documented in

Table 5. Test properties of UHPC.

Material	First Cracking Strength		Ultimate Tensile Strength		Compressive Strength
	Mean 1 (MPa)	C.V 2 (%)	Mean (MPa)	C.V. (%)	Mean (MPa)	C.V. (%)
UHPC	5.26	1.2	7.81	7.4	127.9	4.0

¹ Mean is the mean strength value. ² C.V. is the coefficient of variation.

.

2.3. Fabrication of Specimens

The experimental program involved the fabrication of four reinforced concrete bridge column specimens, with fabrication processes documented in Figure 3. Reinforcement assemblies with standardized configurations (Figure 3a) were prefabricated prior to concrete casting. To enhance the bond between the UHPC jacket and concrete body, a specialized form lining with surface texturing, as shown in Figure 3b, was implemented, creating controlled geometric patterns on the cast surfaces. This texturing technique generated shear connectors at the interface, enhancing interfacial load transferring capacity. Following initial concrete curing and demolding of the temporary steel formwork (Figure 3c), the concrete surface underwent treatment (Figure 3d). Subsequently, customized timber formwork was installed (Figure 3e) for UHPC pouring. The finalized retrofitted column base configuration, demonstrating seamless integration of the UHPC jacket, is shown in Figure 3f. Shear key interface treatment was employed using textured form lining, forming geometric interlocks to enhance mechanical bond. This method ensures effective shear transfer and reduces slip at the UHPC–concrete interface under cyclic loading.

2.4. Test Setup and Loading Protocol

The experimental investigation was conducted at the National Engineering Research Center of Geological Disaster Prevention Technology in Land Transportation, affiliated with Southwest Jiaotong University. Specimen foundations were fixed to the strong floor through post-tensioned ground anchors combined with anti-slip hydraulic jacks, as depicted in Figure 4a. Axial loading simulation employed an electro-hydraulic servo actuator to impose a constant axial load of 385 kN at the column top, representing dead load from the superstructure. This vertical actuator incorporated closed-loop force control, utilizing real-time feedback from an integrated load cell to maintain the precisely specified 385 kN axial load rigidly throughout the entire subsequent lateral testing sequence. Following the stable application and verification of the axial load, the lateral cyclic loading protocol commenced. Lateral force application utilized another servo-controlled actuator linked to the reaction wall, implementing a quasi-static testing protocol through displacement-controlled cyclic excitation.

The loading sequence followed a displacement-controlled protocol with 0.5% incremental drift levels, following the loading regime detailed in Figure 5, with a maximum applied drift is 10%. Three loading cycles were performed at each drift, with positive displacement defined as actuator extension (push direction) and negative displacement corresponding to retraction (pull phase). Comprehensive instrumentation incorporating an array of linear variable differential transformers (LVDTs) and surface-mounted strain gauges enabled continuous monitoring through a synchronized data acquisition system. The test setup and site photo are shown in Figure 4b, illustrating the implementation of experimental methodology. Component hysteretic properties characterized herein provide constitutive inputs for dynamic analysis but do not directly represent system-level inertia or multi-directional effects during earthquakes.

3. Experimental Results and Discussion

3.1. Cracking Pattern and Failure Mechanism

Grids with a size of 50 mm were delineated on the surfaces of the specimens to facilitate the monitoring of crack propagation during the loading tests. Figure 6 presents the comparative crack pattern between the control and retrofitted specimens, revealing fundamental differences in failure mechanisms. For the unmodified RC specimen, initial fracture detection occurred at 0.5% lateral drift, progressing to severe damage characterized by widespread diagonal fracturing, progressive concrete surface crushing and spalling, as shown in Figure 6a. The failure mode was dominated by shear-compression interaction, with cracks forming at the base and spreading upward in a diagonal pattern, eventually leading to core concrete degradation and bar exposure.

In contrast, the retrofitted specimen exhibited significantly delayed damage initiation, with first observable microcracks emerging at 1.0% drift. Crack development was more stable and primarily flexural, initiating near the plastic hinge zone and remaining limited in width and length throughout the loading process. The retrofitted configuration demonstrated superior damage tolerance, as demonstrated in Figure 6b by constrained crack propagation within localized zones, intact concrete integrity, and maintained reinforcement functionality. The 30-mm UHPC jacket effectively redistributed tensile stresses through composite interaction while preventing brittle failure modes. No signs of jacket rupture or debonding were observed, indicating enhanced confinement and effective stress redistribution. The internal concrete, as shown in Figure 6c, reflects minimal damage under the protection of the retrofitted jacket. This is primarily attributed to the high tensile strength of UHPC, which provides lateral confinement, thereby delaying crack initiation and effectively controlling crack width. This comparative analysis suggests that the application of a UHPC jacket demonstrates a substantial improvement in the seismic performance of existing RC bridge piers, primarily by shifting the failure mechanism from brittle shear to controlled flexural behavior. Furthermore, it is found that shear transfer operates through mechanical interlock from 2-mm textured surfaces, chemical adhesion, and friction. The surface texturing formed shear keys that prevented slippage until 3% drift, while debonding in U45 at 4% drift resulted from steel-concrete stiffness mismatch, underscoring material compatibility’s critical role in interface performance.

3.2. Hysteresis Behavior

Figure 7 compares the lateral load-displacement hysteretic responses of the RC column and the 30 mm UHPC-jacketed column. The hysteresis loops of the RC specimen display progressive pinching near neutral displacement, with peak load decay beyond ±30 mm. In contrast, the U30 specimen maintains fuller hysteresis loop profiles throughout loading cycles, achieving a 33% higher load capacity and reduced pinching at extreme displacements (±40–50 mm). This phenomenon suggests that the incorporation of exterior UHPC jackets in the plastic hinge region effectively enhances the hysteresis behavior of RC bridge columns. This enhancement is attributed to the dual function of the exterior UHPC jacket: confining the column body and serving as an additional compression zone. The superior hysteresis performance of UHPC-retrofitted specimens (Figure 7b) stems from three interlinked mechanisms: Firstly, energy dissipation shifts from random concrete crushing to controlled RC/UHPC interface friction. Secondly, attenuated stiffness degradation sustains load amplitudes at large drifts (e.g., >3% rad), enlarging loop area. Thirdly, confinement delays rebar buckling (>4% rad vs. 2.5% rad in RC controls), prolonging stable energy absorption.

The skeleton curves, depicted as the lines connecting the peak loads in the initial cyclic loading at each drift ratio, are illustrated in Figure 8. Throughout the elastic stage, all skeleton curves exhibit similar behavior. However, as the drift ratio increases, the flexural strength of retrofitted columns with UHPC jackets notably surpasses that of RC, with the strength difference increasing as the drift ratio increases. Subsequent to entering the nonlinear stage, the flexural strength of all specimens initiates a decline, with the RC specimen experiencing the most rapid decrease. The key parameters of all skeleton curves, derived from the equivalent energy principle (Figure 9), are tabulated in

Table 6. Test results of reinforcing bar.

D (mm)	fy		fu (MPa)
	Mean 1 (MPa)	C.V 2 (%)	Mean (MPa)	C.V. (%)
12	530.3	0.2	636.1	0.0
6	513.4	0.5	629.8	0.5

¹ Mean is mean strength value. ² C.V. is the coefficient of variation.

. Post-peak (drift >5%), the RC specimen shows rapid strength decay, collapsing to 55% of its peak capacity at 6% drift. In contrast, the U30 curve declines gradually, retaining 82% of its peak strength at equivalent displacement, indicating improvements in flexural strength and deformation tolerance. Additionally, the use of UHPC jackets results in the enhancement in Py, Pmax, Δy, and Δmax by no less than 17.1%, 17.1%, 2.1% and 17.1%, respectively. Pmax of U30 is increased by 24.6%, respectively, compared to that of RC.

3.3. Stiffness Degradation

The determination of stiffness involves calculating the slope between the positive peak and negative peak points of the hysteretic loop, as expressed by Equation (1) and illustrated in Figure 10a.

$$K_{\mathrm{i}}={\displaystyle \frac{\left|+P_{\mathrm{i}}\right|+\left|-P_{\mathrm{i}}\right|}{\left|+P_{\mathrm{i}}\right|-\left|-P_{\mathrm{i}}\right|}}$$

(1)

where +P_i and −P_i are the peak loads corresponding to the positive and negative displacements of +Δi and −Δi, respectively. The degradation of stiffness across all specimens is graphically presented in Figure 10b. The initial stiffness of the retrofitted columns exceeds that of the RC specimen, with an observed increase correlating with the thickness increase of the UHPC jacket. This phenomenon is attributed to the expanded cross-sectional area resulting from the jacket, leading to heightened stiffness. As loading progresses, a decrease in stiffness is witnessed across both specimens with increasing drift ratios. Nevertheless, retrofitted columns consistently exhibit significantly greater stiffness compared to RC specimens at each drift ratio.

3.4. Energy Dissipation

The energy dissipation capacity, quantified by the cumulative hysteresis loop area (Figure 11a), highlights the superior performance of UHPC-retrofitted columns over conventional RC columns. The energy dissipation E_i for each loading cycle is calculated as the area enclosed by the force–displacement hysteresis loop, which can be numerically approximated using:

$$E_i=\int F\cdot \Delta$$

(2)

where F is the lateral load and Δ is the corresponding displacement. The total cumulative energy dissipation up to a given drift level is obtained by summing the energy of all previous cycles:

$$E_{total}=\sum _{i=1}^n{E}_i$$

(3)

where n is the number of completed cycles.

As illustrated in Figure 11b, both systems exhibit comparable energy dissipation at 0.5% drift, primarily due to initial elastic behavior. Beyond this threshold, the U30 specimen demonstrates significantly higher energy dissipation, achieving a 139.5% increase at 3.0% drift compared to RC columns. This enhancement is primarily attributed to the jacket confining the plastic hinge region more effectively, thereby mitigating steel buckling and facilitating uniform energy dissipation via distributed damage. In contrast, RC columns rely on localized energy dissipation via concrete cracking and steel yielding, leading to rapid degradation at higher drifts. The UHPC retrofit thus optimizes post-yield behavior, translating to improved seismic resilience.

4. Numerical Studies

4.1. Analytical Models

Numerical simulations were performed through the fiber element method implemented in the Open System for Earthquake Engineering Simulation (OpenSEES) platform. The analytical model, shown in Figure 12, incorporated nodal spacing at 100 mm intervals from the footing top to a height of 1200 mm. In the retrofitted zone, beam-column elements with A-A section properties connected the nodes, whereas the remaining regions utilized identical elements with B-B section properties. Lateral confinement effects from transverse ties and UHPC jackets [33,34,35] were explicitly modeled. The longitudinal reinforcement employed the “reinforcing steel material” constitutive model (Figure 13a), which incorporates stress degradation mechanisms caused by bar buckling [36] and fatigue accumulation [37]. Key parameters of this steel material model are summarized in

Table 7. Summary of skeleton curves.

Specimen	Δy (mm)	Py (kN)	Δmax (mm)	Pmax (kN)
RC	7.30	100.40	17.99	118.12
U30	7.45	125.09	20.99	147.16

. For concrete and UHPC modeling, the “Concrete 02” material Figure 13b was adopted, with material properties for confined concrete, unconfined concrete, and UHPC systematically documented in

Table 8. Parameters of ‘Reinforcing Steel Material’.

Specimen	Buckling Module				Fatigue Module
	lsr	β	r	γ	Cf	α	Cd
RC	6.66	1	0.5	0.8	0.26	0.500	0.389
Retrofitted columns	6.66	1	0.5	0.8	0.26	0.425	0.400

,

Table 9. Parameters of ‘Concrete02’ in RC.

	fc (MPa)	εc	f’u (MPa)	ε′u	λ	ft (MPa)	Ets (MPa)
Unconfined concrete	57	0.00238	11.40	0.00547	0.1	0	0
Confined concrete	62.7	0.00358	12.55	0.10740

,

Table 10. Parameters of ‘Concrete02’ in retrofitted columns.

	t (mm)	fc (MPa)	εc	f’u (MPa)	ε′u	λ	ft (MPa)	Ets (MPa)
UHPC confined concrete	15	59.29	0.0029	11.85	0.086	0.1	0	0
	30	60.43	0.0031	12.09	0.093
	45	61.58	0.0033	12.32	0.100
UHPC and tie-confined concrete	15	65.02	0.0040	13.00	0.120
	30	66.17	0.0043	13.23	0.129
	45	67.32	0.0045	13.46	0.135

and

Table 11. Parameters of the Concrete02 model for UHPC jacket.

	fc (MPa)	εc	f’u	ε′u (MPa)	λ	ft (MPa)	Ets (MPa)
UHPC	127.9	0.0035	25.58	0.105	0.1	7.81	2057

.

4.2. Numerical Validation

The numerical simulations (SIM) and experimental results (EXP) for both hysteresis loops and skeleton curves are presented in Figure 14 and Figure 15, demonstrating close alignment between the two datasets. This agreement validates the accuracy of the adopted material models and parameters, confirming their effectiveness in capturing the cyclic behavior of the tested columns. Notably, the numerical predictions for both the RC and U30 specimens exhibit strong correlation with experimental measurements, reinforcing the reliability of the analytical approach in simulating the structural response under cyclic loading conditions.

4.3. Parametric Study of Jacket Thickness

To optimize seismic retrofitting efficiency, OpenSees simulations analyzed bridge columns with UHPC jackets of 15 mm (U15), 30 mm (U30), and 45 mm (U45) thickness under a constant jacket height of 400 mm. While U30 experimental results anchor the model, parametric thickness variations (U15/U45) derive from extrapolated simulations. This assessed key seismic indicators—lateral strength, ductility, and energy dissipation—to guide cost-effective strategies. Performance was evaluated through hysteresis curves, skeleton curves, and energy dissipation plots as shown in Figure 16 and Figure 17.

As depicted in Figure 16 and Figure 17, U15 exhibited the poorest performance, with a peak lateral load of 125 kN, rapid strength decay after 3% drift, and severely limited energy dissipation due to insufficient confinement. In contrast, U30 achieved a 20% higher peak load with gradual post-peak degradation. Critically, its energy dissipation matched U45 up to 3% drift, demonstrating efficient hysteretic behavior without excessive material use. U45 matched U15’s peak load and showed abrupt strength drops beyond 40 mm due to interfacial debonding, despite achieving marginally higher energy dissipation only at extreme drifts of more than 3% drift.

U30 emerged as the optimal solution, delivering 12% higher lateral strength than U45 alongside comparable energy dissipation in the critical service-drift range. While U45 achieved slightly greater cumulative energy dissipation beyond 3% drift, this marginal gain is offset by its higher material costs, interfacial failure risks, and significantly lower load resistance. The 30 mm thickness thus balances structural resilience (strength/ductility) with energy dissipation efficiency, confirming its technical and economic superiority for seismic retrofits.

The performance of UHPC jackets observed in this study aligns with findings from previous experimental investigations. Al-Amoodi et al. [38] examined the compressive performance of RC columns retrofitted with 20, 30, and 40 mm UHPFRC jackets under monotonic uniaxial loading and reported that a 30-mm thickness provides a substantial balance between load-carrying enhancement and material efficiency, with up to 132% strength increase for square columns. While their study focused on axial behavior, our simulation-based cyclic analyses further demonstrate that the 30-mm jacket not only enhances strength but also promotes ductile behavior up to 10% drift, delaying failure mechanisms such as debonding and fiber pull-out.

Moreover, Zhang et al. [39] investigated the cyclic performance of RC piers retrofitted with ECC and BFRP-ECC jackets and found significant improvements in energy dissipation and reduced damage concentration near the column base. Similarly, our U30 specimens exhibited improved energy absorption, retaining over 70% of peak lateral capacity at 3% drift. Unlike Zhang et al., who studied hybrid materials, our study isolates UHPC behavior and offers a parametric insight across thicknesses, revealing that both insufficient (15 mm) and excessive (45 mm) jacketing may trigger brittle failure modes due to either confinement inadequacy or strain incompatibility.

Thus, this study advances the current understanding by identifying the 30-mm jacket as a practical and mechanical optimum under seismic loading through a detailed parametric analysis, which remains largely unaddressed in existing literature. Further experimental validation under dynamic multiaxial loading is recommended to verify these findings.

5. Discussion

The 30-mm UHPC jacket (U30) presents a favorable balance between seismic performance and constructability. It achieves a 24.6% strength gain while benefiting from the self-consolidating properties of UHPC, which reduce formwork complexity and labor requirements. Additionally, its material cost is approximately 40% lower than that of the 45-mm alternative (U45). In moderate seismic zones, thinner jackets such as the 15-mm option (U15) may offer a viable solution, though further validation is needed.

However, several limitations should be acknowledged. Firstly, the analysis is based on a fixed column geometry, without considering size effects, varying axial loads, or construction tolerances that may affect confinement behavior in practice. While parametric conclusions are supported by validated numerical models, the use of simplified material laws and idealized interfaces excludes critical factors such as surface roughness, curing conditions, and long-term degradation mechanisms.

Additionally, it should be noted that while the retrofitted columns maintained stable hysteretic behavior up to 10% drift, the failure behavior beyond this range remains uncertain. Numerical results suggest that extreme displacements (>12%) may induce debonding at the jacket–concrete interface, fiber pull-out, or reinforcement instability. These mechanisms could trigger brittle degradation and rapid strength loss. Further testing under larger displacement demands is needed to fully characterize the ultimate failure modes under extreme seismic events.

Lastly, future studies incorporating larger sample sizes and stochastic material modeling will quantitatively address uncertainty propagation to refine design recommendations.

6. Conclusions

In this study, the efficiency of retrofitting RC bridge columns through the application of UHPC jackets was investigated both experimentally and numerically. The main conclusions drawn from the combined analysis of experimental and numerical results are outlined below:

• Improved Damage Control: The 30-mm UHPC jacket effectively delayed crack initiation and controlled damage propagation, shifting failure mode from brittle shear in the RC specimen to ductile flexural behavior in the retrofitted column.
• Enhanced Seismic Performance: Compared to the unretrofitted RC specimen, the U30 column demonstrated a 24.6% increase in peak lateral load capacity and a 139.5% increase in energy dissipation at 3% drift, validating the seismic efficiency of UHPC jacketing.
• Numerical Model Validation: A fiber-based OpenSees model accurately replicated the experimental behavior, supporting the adopted constitutive models for concrete, UHPC, and steel under cyclic loading.
• Thickness Optimization Insight: Parametric simulations of 15 mm, 30 mm, and 45 mm jackets revealed that the 30-mm thickness optimally balances lateral strength, ductility, and energy dissipation. Thinner jackets failed to provide sufficient confinement, while thicker ones induced premature brittle failure due to strain incompatibility.
• Design Recommendation: Based on both experimental validation and simulation results, a 30-mm UHPC jacket is recommended as a cost-effective and constructible solution for seismic retrofitting of RC bridge columns in plastic hinge regions.