PSI - Issue 81

Mykola Pidgurskyi et al. / Procedia Structural Integrity 81 (2026) 439–446

441

Fig. 3. Iso-stress contours in the beam web: a), b) end and central openings of a perforated beam; c), d) support and central zones of a solid-web beam.

3. Research Methodology The finite element method was applied to investigate the stresses developing around the perforation openings in steel perforated beams. The numerical models and static analyses within the elastic range of steel behavior were performed in the ANSYS software environment. A global finite element size of 20 mm was adopted, while the mesh around the investigated openings was refined to 2 mm (Fig. 4). This mesh-refinement method demonstrated sufficient accuracy with significantly reduced computational time compared to a full-model mesh with a uniform 2 mm element size (Pidgurskyi et al., 2023; Dutt, 2015).

Fig. 4. Refined finite element mesh around perforation openings of the arched beam with haunches: a) eaves joint; b) apex joint.

A full solid (three-dimensional) model was developed, accounting for the geometry of the perforated web, real support conditions, and applied loads. The beams were modelled using structural steel grade S345 in accordance with regulatory requirements. All perforated beams were designed in compliance with domestic (DBN V.2.6-198:2014, 2014) and international standards (EN 1993-1-1:2006, 2005; EN 1993-1-13:2024, 2024; Fares et al., 2016). The analysis was carried out based on equivalent stresses along the edge of each opening, computed using the von Mises yield criterion. For beams reinforced with circular stiffening rings, the equivalent stresses were evaluated along the inner surface of the ring. 4. Boundary Conditions and Beam Parameters According to previous studies, the beams were analyzed with a minimum inclination of 5°, at which horizontal thrust develops. For comparison purposes, a horizontal beam with the same geometry and perforation pattern was also modelled. The arched beam was designed with a circular outline, and its rise was selected to match the height of the double-pitched beam. All beams were modelled with rigid supports at the column locations, meaning that translations (X, Y, Z) and rotations (UX, UY, UZ) at the beam ends were fully restrained in all directions. For beams without haunches, bearing stiffeners were modelled to reduce stress concentration at the supports (Fig. 5(a)). In addition, no transverse plates were included in the apex joint for this beam type. For beams with haunches, these stiffeners and plates were modelled as structural components; however, since self-weight is included in the loading scheme, their presence also increases the overall load. The haunch lengths were 2880 mm at the eaves joint and 2260 mm at the apex joint, with a height of 395 mm. The plate dimensions corresponded to the web and flange dimensions of the I-section. To reduce stresses around the openings, stiffening rings made from circular steel tubes with a wall thickness of 5 mm (the