PSI - Issue 80

R. Salem et al. / Procedia Structural Integrity 80 (2026) 256–268 Rania Salem / Structural Integrity Procedia 00 (2025) 000 – 000

261

6

strength (Sahoo et al., 2024; Sun et al., 2008; Coogan and Kazmer, 2017).

Each phase is modelled with distinct elastic properties, as summarized in Table 1. Porosity is represented as a near-zero stiffness phase, consistent with homogenization strategies for porous thermoplastics, to approximate its negligible contribution to the overall stiffness of the printed structure. Poisson’s ratio is uniformly set to 0.35, corresponding to bulk ABS, for all solid regions. For voids, however, it is explicitly assigned a value of zero to represent their inability to sustain transverse strain under axial loading. This assumption preserves numerical stability in the simulations while remaining physically consistent with the mechanical insignificance of voids. It is noteworthy that the void phase is explicitly meshed and assigned material properties to facilitate extension of analyses to thermal problems. Unlike in purely mechanical analyses, thermal problems require explicit consideration of air properties. Young's moduli assigned to the solid phases within the RVE are centered at 2000 MPa, approximating the Young's modulus of FFF-printed ABS. A variation of ±500 MPa about this mean value is applied to bonding regions to investigate the effect of their altered mechanical properties, which result from region-specific polymer chain reticulation and shearing mechanisms.

Table 1: Parameters assigned to the different phases

PARAMETER

VALUE

UNIT MPa MPa MPa MPa

2000 2000 ±25% 2000 ±25% 0.001

E REF E HBR E VBR

E P

0.35

-

These parameters are assigned to the different phases according to these assumptions: ▪ Case 1: Degraded bonding: E hbr = E vbr = 1500 MPa , simulating poor interdiffusion across all interfaces (Kumar et al., 2023). ▪ Case 2: Overbonded interfaces: E hbr = E vbr = 2500 MPa, introduced as a numerical test case to evaluate the upper-bound mechanical behavior assuming overbonded interfaces. While this exceeds the reference modulus ( E ref = 2000 MPa ), it serves to benchmark the model's response under idealized bonding conditions. It does not reflect physically realizable material systems unless such stiffened interfacial phases are justified by experimental evidence. ▪ Case 3: Anisotropic (Weak HBR): E hbr = 1500 MPa and E vbr = 2500 MPa , reflecting limited in-layer fusion but improved interlayer bonding under deposition pressure. ▪ Case 4: Anisotropic (Weak VBR): E hbr = 2500 MPa and E vbr = 1500 MPa , indicating strong in-layer but thermally weakened interlayer bonds (Kumar et al., 2023). ▪ Case 5: Fully bonded (Isotropic): E hbr = E vbr = E ref = 2000 MPa , representing uniform, ideal interfacial strength (Kumar et al., 2023). 2.2. Geometry of porosity and bonding regions As outlined above, the geometry of the curvilinear‐porosity RVE is used to analytically quantify both porosity and bonding contributions in the printed cross section. Bonding regions are represented by two different phases formed at the horizontal interfaces between adjacent filaments within a layer and those at the vertical interfaces between successive layers (Fig. 2b) . Their areas are described using the standard formula for a circular segment, leading to: