PSI - Issue 82

Chiara Bedon et al. / Procedia Structural Integrity 82 (2026) 65–71 Chiara Bedon et al. / Structural Integrity Procedia 00 (2026) 000–000

70 6

The analysis of results from Table 1 suggests that – among other influencing parameters – the thickness of glass has a key role in ensuring any possible mechanical interlock of fragments in the post-breakage stage. Considering both SII and SIII configurations, as well as EVA or SG interlayers, the V1 specimens can still offer – even after the imposed cyclic deflections – a minimum stiffness contribution, as it can be seen from the values presented for t g = 10 mm. This effect is generally less pronounced for the specimens with t g = 4 mm. A similar trend can also be seen for both 10 mm and 4 mm thick glass layers when subjected to V2, both in SII and SIII. In terms of possible comparative analysis of present and past experimental results, a direct correlation is always hard to generalize. However, considering for example the short-term/small-amplitude loads applied in (Bedon et al., 2025) to large-scale EVA-bonded, 2-ply AN samples, some preliminary considerations can be drawn to support further studies (Table 2). In this regard, it is important to remind that the present study voluntarily imposed preliminary cracks randomly distributed (close to mid-span), rather than regular ones only, as it is known that their position / size / depth further affects the bending stiffness of the broken LG element (Bedon & Fasan, 2024; Bedon & Santos, 2023). The specimen size is another key influencing parameter that should be taken into account, as also observed in (Bedon et al., 2025; Zhou et al., 2025). In this study, the experiments proved that the bending stiffness of cracked LG specimens suffers for a progressive modification during the imposed cyclic protocol. Most importantly, the size effect can be roughly seen from Table 2, where a size ratio of 10 characterizes the listed specimens. Considering that the present and past specimens were both composed of AN glass and bonded by 0.76 mm EVA film, the average experimental results shows that there are no doubts that the partially fractured 2-ply LG section suffers for a certain loss of bending stiffness and load-bearing capacity. Besides, the reported E g,cr values are clearly associated to non-null mechanical contribution of the cracked glass layer, which is particularly relevant when the cracked layer is on the compressive side (SII) but especially in the large-scale setup. At the small-scale level, a 40-50% residual contribution can be still ensured in SII, which suggests a possible optimization of safety rules and structural design assumptions. Besides, future experimental investigations and extended studies should further explore the interconnected influencing parameters. Table 2. Qualitative comparison of present and literature results, in terms of calculated (average) equivalent modulus of elasticity for the cracked glass layer ( E g,cr ) in SII and SIII configurations, based on different experimental investigations (all samples with EVA interlayer).

Small-scale specimen (present study) [Cyclic loading protocol]

Large-scale specimen (Bedon et al., 2025) [Single loading step]

Size

50 mm × 100 mm

500 mm × 1000 mm

≈1/23000 in SII ≈1/10600 in SIII

Imposed deflection/span ratio

≈ 1/1750 in SII and SIII

t int [mm]

t g [mm]

E g,cr [GPa]

D E [%]

t g [mm]

E g,cr [GPa]

D E [%]

10

23.65

-66.2

SII

0.76

8

69.9

-0.14

4

7.62

-89.1

10

31.69

-54.7

SIII

0.76

8

37.9

-45.8

4

2.73

-96.1

4. Conclusions The post-breakage security and structural safety analysis and quantification for laminated glass (LG) elements in constructions is a critical issue, and a rather challenging task, due to many influencing parameters and interconnected aspects that are hard to control and quantify by simplified approaches. From a practical point of view, it is assumed that the glass type (and thus the associated size and shape of shards) is one of the most important influencing parameters for the residual capacity assessment in case of damage. As such, annealed (AN) glass – which is the weakest in tension and is associated to few large shards – should not suffer for severe stiffness reduction in the post breakage stage, compared to heat-strengthened and even fully tempered glass types. In this paper, the attention was specifically focused on the cyclic experimental analysis of small-scale, 2-ply LG specimens composed of AN glass.