PSI - Issue 84

Emeka John Ude et al. / Procedia Structural Integrity 84 (2026) 1294–1301

1297

3.2. Model setup and commands

The concrete beam was modelled as a 3D eight-node mixed volume/pressure bbarBrick element with the and ASDConcrete3D adopted for its material physical property. The reinforcements were modelled as Displacement Based Beam-Column element with uniaxial trilinear hysteretic material adopted as physical property. The reinforcement geometry was embedded into concrete using a node-to-element interaction to define retained and constrained properties between the concrete and reinforcements nodes respectively. For this, the ASDEmbeddedNodeElement multi-point constraint element in STKO which imposes a penalty-based adherence between the concrete and rebars was adopted. The FRCM was realized as a surface element, where the layered shell formulation was used together with the PlateRebar command to capture the bidirectional effect and possible rotation of the FRCM system, as illustrated in Fig. 3, since this command allows definition of the orientation, ϴ . The effective thickness of the PlateRebar was set to 0.093 mm, consistent with the experimental characterization of the FRCM system. The elastic modulus and tensile strength of the AR‑glass fabric were taken from (Leone et al., 2017), whereas the mortar compressive strength was adopted from the experimental report. A tie constrain was defined for the FRCM and concrete surface using the equalDOF command to simulate perfect bond since the experiment did not report debonding. Appropriate zero length elements and constraints were defined for the loading plates and support to reproduce the experimental setup conditions. A mesh sensitivity analysis was performed, and a mesh size of 2.5 cm appeared to be the best trade-off that represents better the global response of the experimental campaign. The meshed model was thereafter partitioned into 10 to be run in parallel using the OpenSees MP solver. The displacement controlled static analysis was adopted for the numerical simulation and the command set up implored the penalty method as the constraint handler, while the Krylov-Newton was adopted as the solution algorithm.

(a )

(b)

Fig. 3. Model of the reference beam in STKO: (a) concrete, reinforcement and FRCM setup, (b) quadratic mesh and partitioning

3.3. Modelling corrosion For the concept of corrosion, a uniform corrosion the reinforcement in critical region of the beam was assumed (Fig . 4 ), considering cross-section (mass) loss from 10, 20, 30, 40 and 50%. In addition, the adjustment of ductility of the corroded reinforcements follows the formulations in Eqs. (2) – (4) as extensively described in (Rosso et al., 2022).

Fig. 4. Corroded reinforcement sections considered ′ =(1− ) ′ = (1− ) ′ = (1− 1 )

(2) (3)

(4) Finally, a bond slip constitutive model due to corrosion was implemented in the concrete-steel interphase following the formulation by (Blomfors et al., 2018; Lundgren et al., 2012) and well described in (Lin et al., 2019). This