The proposed design combines the strength and plastic energy dissipation of steel with the flexibility and shear damping of rubber. Steel bars were welded between rigid upper and lower plates to form the principal load-bearing framework, while compounded recycled rubber filled the spaces around them. Before assembly, steel surfaces were treated with Chemlok 205 primer and Chemlok 220 rubber-to-metal adhesive.

Rubber sheets and the steel framework were then placed in a mould and joined by hot-compression vulcanization at 128–130 °C for 2–2.5 hours under pressure. The resulting continuous rubber core was intended to coordinate deformation among the bars and preserve a load path after local steel fracture, unlike a conventional steel bar damper whose response can be abruptly interrupted by fracture. The experimental matrix contained nine specimens: three steel bar dampers, designated BD; three rubber-only dampers, designated RD; and three recycled rubber–steel bar composites, designated RBD.

Each steel and composite specimen contained eight bars, each 10 mm in diameter, arranged in a 2 × 4 layout. The rubber and composite cores had a projected area of 26,250 mm², and the connection plates measured 200 × 250 mm and were 10 mm thick. Core heights of 75, 100 and 125 mm allowed configuration type and height to be evaluated separately.

The selected rubber contained 20% devulcanized rubber from truck-tread powder and 80% Standard Malaysian Rubber Constant Viscosity 60. Material tests gave it a tensile strength of 21.5 MPa, elongation at break of 530%, compressive modulus of 6.0 MPa at 20% strain and damping ratio of 6.31%. Each specimen underwent 13 displacement-controlled loading steps with three complete cycles at each step, allowing as many as 39 cycles.

Displacement amplitude increased from 2 to 50 mm, and loading rates ranged from 0.25 to 0.80 mm/s. Testing stopped when severe visible damage occurred or the maximum resisting force in the current cycle fell below 85% of the peak resisting force. The vertical arrangement generated cyclic shear deformation in each energy-dissipation core.

Because the loading path kept the distance between the connection plates essentially constant, geometric elongation of the steel bars could produce secondary axial tension at large displacements, although axial force was not independently applied or measured. All three configurations were tested with the same installation and boundary conditions. Damage observations showed a pronounced change in failure evolution.

Steel-only specimens fractured after pronounced bending of their bars: BD8–75 failed on cycle 19 at about 14 mm displacement, while the 100 and 125 mm versions failed on cycles 28 and 31 at amplitudes of 27.5 and 35 mm. The longer core increased the effective bending length and delayed fracture. Rubber-only specimens completed all 39 cycles and reached 50 mm, although they developed varying degrees of interface debonding and edge warping without through-thickness tearing of the rubber core.

The composites combined bar yielding and fracture with rubber shear deformation and interfacial debonding. Their continuous rubber core continued to connect the upper and lower plates after local bar fracture, limiting the sudden load degradation seen in the steel-only devices. The largest benefit occurred in the shortest composite core.

RBD8–75 reached its major failure on cycle 25 at approximately 22.5 mm, compared with cycle 19 and 14 mm for BD8–75. Its peak load was 33.12 kN, 35.2% above the steel specimen’s 24.49 kN. Peak loads for the 100 and 125 mm composites were respectively 30.5% and 43.0% higher than those of their steel-only counterparts.

After local steel damage, the taller composites retained residual load capacity and overall connectivity, and no through-thickness tearing occurred in their rubber cores. Numerical results showed that the height of the composite core reduced bar curvature and end plastic-strain concentration; at peak load, maximum equivalent plastic strains in the 100 and 125 mm models were 12.5% and 15.8% lower than in the 75 mm model. Stiffness declined as cyclic deformation accumulated, but the steel framework continued to support the composite after the bars began yielding.

Across displacement amplitudes of approximately 2–50 mm, effective stiffness decreased by about 62–64% for the rubber-only dampers and 73.3–82.6% for the steel dampers. Increasing the height from 75 to 125 mm reduced the rubber-only device’s maximum load by 40.2% while leaving its maximum displacement near 50 mm. For the steel damper, the same change reduced maximum load by 41.8% but increased maximum displacement by about 150%, from 14.01 to 35.02 mm.

The results expose a design trade-off: a taller steel core accommodates more distortion but generally develops less peak force because the bars bend over a longer length. Energy measurements showed that the composite produced more than a modest increase in resistance. Final cumulative dissipation by RBD8–75 was 217.5% higher than that of BD8–75, while RBD8–100 exceeded its steel counterpart by 29.0%.

At 125 mm, the final cumulative difference was only 0.4%, although the composite dissipated more over most of the displacement range shared by the two specimens. Maximum single-cycle energy dissipation was respectively 104.8%, 21.9% and 2.3% higher for the 75, 100 and 125 mm composites than for their steel counterparts. The rubber therefore made its largest contribution when short bars were especially vulnerable to concentrated deformation.

RBD8–100 provided the best balance among the tested composites, combining a peak load of 25.14 kN with the highest single-cycle and cumulative dissipation values. The researchers also developed a three-dimensional ABAQUS model with geometric nonlinearity, bilinear elastoplastic steel, ductile steel damage and a Neo-Hookean hyperelastic rubber. The model was validated against the three RBD specimens, not all nine experimental specimens.

Numerical-to-experimental peak-force ratios ranged from 0.941 to 1.095. Predicted stiffness degradation differed from the measurements by 0.30–2.10%, and errors for single-cycle energy dissipation ranged from 5.09% to 8.71%. The model was less accurate for the 75 mm composite’s cumulative energy dissipation and residual deformation, with errors of 37.07% and 34.52%, respectively.

It also omitted rubber viscoelasticity, the Mullins effect, rate dependence and progressive rubber damage, limiting its direct representation of dynamic or long-term behaviour. Separate numerical parameter studies indicated that increasing the steel-to-rubber area ratio from 1.53% to 4.69% raised the maximum absolute force at 10 mm displacement by about 249% and average energy dissipation per cycle by about 439%. The equivalent damping ratio increased more modestly, from 11.97% to 18.63%.

Raising steel yield strength from 250 to 500 MPa increased maximum absolute load at 22.5 mm by 42.9%, average single-cycle energy dissipation by 67.0% and equivalent damping ratio by 17.8%. Within the explored ranges, the researchers identified a 3.45% steel-to-rubber area ratio and yield strength near 400 MPa as balanced choices, not universal optimum values. A simplified bilinear backbone model based on the experiments and parametric models linked core height, area ratio and yield strength to initial stiffness, characteristic strength and post-yield stiffness; characteristic-parameter errors for the three experimental composites were below 5.1%, and backbone-curve errors ranged from 3.68% to 5.40%.

The work points toward a way to combine metal yielding with rubber shear deformation while reusing waste tyre rubber, but the evidence remains limited to laboratory-scale quasi-static component tests and numerical models. Repeated experiments over broader parameter ranges, full-scale and structural-subassembly tests, dynamic cyclic loading, realistic earthquake records, and evaluations of temperature, aging, environmental exposure and interface durability are still needed before the performance or practical deployment of the damper can be assessed beyond the studied conditions.

Subject of Research: Component-level cyclic performance of recycled-rubber–steel bar composite dampers

Article Title: Mechanical performance of novel recycled rubber–steel bar composite dampers: Experimental and numerical investigation

Article References: Alhokabi, A. A., Tan, C. G., Zou, C., Ibrahim, Z., Bunnori, N. M., El‑Shafie, A., Nazri, F. M., Ghaedi, K., & Gordan, M. (2026). Mechanical performance of novel recycled rubber–steel bar composite dampers: Experimental and numerical investigation. Case Studies in Construction Materials, 25, Article e06629. https://doi.org/10.1016/j.cscm.2026.e06629

Image Credits: AI Generated

DOI: 10.1016/j.cscm.2026.e06629

Keywords: recycled rubber, steel–rubber composite, cyclic loading, energy dissipation, hysteretic response, damage tolerance, finite element analysis, seismic dampers, waste tyre rubber, structural resilience, Mechanical, performance

A new study published in the Bulletin of Earthquake Engineering proposes a way to resolve that trade-off, and its central idea is disarmingly simple: give the earthquake something cheap and expendable to break. Researchers Min Jae Park of Pukyong National University in Busan, Ali Ghamari, Imran Karimi of Tabriz University, and Radhika Sridhar of Walailak University in Thailand have developed a hybrid precast concrete beam-column connection built around a replaceable I-shaped steel ductile link that functions as a structural seismic fuse. Like the fuse in an electrical circuit, the device is designed to absorb damage on behalf of the more valuable components around it, concentrating plastic deformation in a sacrificial element that can be unbolted and swapped out after a major event.

The mechanics of the system reflect a deliberate strategy of controlled damage. The steel link is bolted inside a steel-jacketed joint region at the intersection of the precast beam and column. During an earthquake, as the frame sways and the joint demands rotational capacity from the connected members, the I-shaped link yields in a controlled fashion, dissipating seismic energy through plastic deformation. Crucially, the design ensures that yielding is confined exclusively to the fuse itself. The precast beams, the columns, and the joint region are all shielded from the inelastic demands that would otherwise crack concrete, buckle reinforcement, and leave a structure permanently weakened or economically unrepairable. Once the shaking stops, the damaged link can be removed with ordinary tools and a new one installed, restoring the connection to its original capacity.

To understand how such a fuse should be proportioned, the research team carried out an extensive parametric and numerical study, anchored against experimental data to ensure the models reflected real structural behavior rather than numerical artifacts. Four key design variables came under scrutiny. The first was the ratio between the strength of the steel link and the strength of the concrete beam, denoted by the Greek letter psi, which governs whether the fuse or the surrounding concrete yields first. The second was the rho factor, capturing the effect of the relative resistance distribution within the connection. The third and fourth were the geometric properties of the I-shaped link itself: the thickness of its flanges and the depth of its cross-section, the two dimensions that most directly control how much bending resistance and deformation capacity the fuse can deliver before it is exhausted.

The results of the numerical campaign were strikingly consistent. Across the range of parameters examined, every model exhibited stable hysteretic behavior, the looping force-deformation signatures that engineers read as evidence of a structure absorbing and releasing seismic energy cycle after cycle without rapid deterioration. All models achieved moment capacities exceeding the plastic moment of the beam, meaning the fuse could reliably mobilize the full strength of the connection, and the system displayed minimal stiffness degradation, indicating that the connection did not progressively soften and loosen as the shaking continued. In seismic engineering, where degrading hysteresis loops often signal a connection heading toward collapse, this combination of stable loops and preserved stiffness is precisely the behavior a resilient system should show.

Perhaps the most consequential finding concerned which of the four design variables actually mattered most. Intuition might suggest that the strength ratio between the fuse and the concrete beam, psi, would dominate the response, since it determines where yielding occurs. The study found otherwise. The depth of the I-shaped link, h, exerted the strongest influence on the system’s behavior, outweighing the strength ratio, the rho factor, and the flange thickness. That result carries practical weight for designers: it means that tuning the depth of the sacrificial element is the most powerful lever available for shaping how the connection performs, a conclusion that was not obvious before the parametric study was run and one that could redirect how future fuses are detailed.

Beyond identifying the dominant parameter, the researchers distilled their findings into proposed design equations for the system, giving practitioners a quantitative route from target performance to fuse dimensions. Design equations of this kind are the bridge between a promising laboratory concept and a buildable technology, because they allow engineers to size the link so that it yields before the surrounding precast elements, sustains the large deformations expected in a design-level earthquake, and retains enough residual capacity to protect the joint through the full duration of strong ground motion. The study’s grounding in established seismic frameworks, including references to modern minimum design load provisions and seismic provisions for structural steel buildings, situates the proposal within the codes that already govern practice.

The motivation for this line of research is written in the damage surveys of recent disasters. Investigations following the February 2023 Kahramanmaras earthquake sequence in Turkey documented widespread failures of reinforced concrete and precast construction, with brittle connections and inadequate detailing cited among the recurring deficiencies. Earlier events, notably the 2011 Christchurch earthquake in New Zealand, exposed how conventionally detailed reinforced concrete frames can suffer damage so extensive that repair becomes uneconomical, forcing demolition of buildings that technically did not collapse. Those experiences have pushed the international research community toward a resilience paradigm that goes beyond life safety: structures should not merely stand up during an earthquake but should emerge from it repairable, functional, and quickly returnable to service.

Replaceable fuse technology has become one of the most active fronts in that paradigm shift, with researchers developing damage-control joints, bio-inspired hinges, artificial controllable plastic hinges, and self-centering connections for both cast-in-place and precast frames. What distinguishes the new study is its integration of the fuse concept into a hybrid precast connection that preserves the construction speed advantages of precast systems while solving their historical connection problem. The bolted arrangement inside a steel-jacketed joint means the fuse is not cast into the concrete or welded into place, so replacement is a matter of unbolting rather than cutting, chipping, and recasting. For building owners and city planners, that distinction translates directly into downtime: a connection repaired in days rather than months keeps schools open, hospitals running, and businesses operating after a quake.

The work, supported by grants from the National Research Foundation of Korea and the Korea Agency for Infrastructure Technology Advancement, along with facilities provided by Walailak University, arrives at a moment when the global building stock in earthquake-prone regions is expanding fastest precisely where seismic risk is highest. Precast construction will remain attractive to developers for its speed and economy, and the question is not whether it will be used but whether it will be used safely. By demonstrating through validated numerical modeling that a bolted, replaceable, ductile steel fuse can deliver stable hysteretic performance, protect the primary structure, and be tuned through a single dominant geometric parameter, the researchers have offered the industry a concrete, and constructible, answer. The next steps toward widespread adoption will involve experimental validation at full scale and codification of the proposed design equations, but the study makes a compelling case that the weakest link in precast construction can be redesigned into its strongest feature: a sacrificial component that takes the hit so the building does not have to.

Subject of Research: Seismic-resilient precast concrete beam-column connections with replaceable ductile steel fuses

Article Title: Towards resilient precast concrete frame featuring a hybrid system with replaceable ductile steel fuses

Article References: Park, M. J., Ghamari, A., Karimi, I., & Sridhar, R. (2026). Towards resilient precast concrete frame featuring a hybrid system with replaceable ductile steel fuses. Bulletin of Earthquake Engineering. https://doi.org/10.1007/s10518-026-02668-6

Image Credits: AI Generated

DOI: 10.1007/s10518-026-02668-6

Keywords: precast concrete, seismic fuse, ductile steel link, beam-column connection, earthquake engineering, hysteretic behavior, replaceable plastic hinge, structural resilience, parametric study, stiffness degradation, repairable structures, Bulletin of Earthquake Engineering