The diversity of modern structural glasses is extremely broad, and their mechanical properties vary significantly. In this work, original specimen geometry is proposed for determining the fracture toughness parameters of glass under quasi-static loading conditions. A key feature of the developed configuration is the presence of a stabilization zone, which ensures crack propagation in a given direction. The specimen shape is justified by the results of a numerical simulation of fracture processes in brittle solids. The proposed geometry is characterized by reduced sensitivity to the loading device stiffness and promotes the straight-line propagation of the crack. Specimen fabrication is straightforward and does not require specialized equipment, allowing mechanical testing to be performed using standard tensile testing machines. The developed configuration ensures stable regimes of crack deceleration and complete arrest. The crack tip formed at this stage acquires a natural configuration, and the geometric characteristics of the initial notch do not have a critical influence on the measurement results. The capabilities of the specimen are demonstrated through both numerical and physical experiments in comparison with the classic compact tension (CT) specimen. Physical testing of glass specimens allowed for obtaining the relationship between the crack velocity and the stress intensity factor within a velocity range from 10−6 to 10−2 m/s.
Early architectural decisions on orientation, window-to-wall ratio, shading, and envelope composition fix most of a building’s passive thermal performance long before dynamic energy simulation becomes feasible, leaving these decisions without a structured evidentiary basis. This paper proposes Climate Evidence-Based Design (CEBD), a conceptual framework specializing the evidence-based design tradition from healthcare and workplace research into the climatic, physical, and material evidence needed for integrated, early-stage building decisions. CEBD organizes four components: a diagnosis of why evidence arrives too late to inform early-stage climate-responsive decisions; a six-tier evidence-source structure ordered by increasing directness and analytical control, from design heuristic to measured monitoring, complemented by independent case-specific qualifiers of reliability, contextual relevance, transferability, and uncertainty; Architectural Intent, showing how a project’s declared priorities change which recommendation identical evidence supports; and a mechanism for applying graded evidence before simulation is feasible, without displacing it once it is. A single conceptual model integrates climate evidence, envelope physics, and material sustainability into an operational structure, the Climate Evidence-Based Design Methodology (CEBDM), through which a design team applies graded evidence to early-stage climate-responsive building-performance decisions. CEBD is deliberately bounded to climate-responsive building-performance decisions; it does not attempt to classify or evaluate architectural quality, cultural meaning, tectonic expression, or spatial composition. By making the evidentiary basis of passive design explicit and inspectable, CEBD is a conceptual foundation for future evidence-based design tools. This paper establishes CEBD’s specification and internal logic to support empirical testing; it reports the framework’s construction, together with a retrospective application to a previously published case corpus that demonstrates the applicability of its categories and not the effectiveness of the framework, and identifies the validation studies required as next steps: empirical tests of whether this multidimensional evidence representation is associated with subsequently validated performance outcomes.
Retrofitting existing buildings can contribute to both climate change mitigation and adaptation, yet their interaction has received limited attention across several climatic contexts. This state-of-the-art review investigates retrofit strategies for the existing building stock in European climate zones. The screening process followed PRISMA-ScR guidelines and used Scopus as the database, with three researchers independently assessing the retrieved studies. After screening, 118 publications were analysed to identify research trends, while 48 were selected for detailed assessment of adaptation and mitigation retrofit strategies. The findings show that retrofit performance varies with climate and local conditions. For climate-responsive design, passive measures are still central, although their abiding effectiveness often depends on integration with active systems. Therefore, combinations of envelope improvements, solar control, passive cooling, and selected active technologies appear more effective than isolated measures. These strategies may also involve compromises between energy efficiency, thermal comfort, and environmental impact under present and future climates. The review identifies climate-specific relationships between retrofit measures and adaptation–mitigation outcomes. Current research is affected by climate zone coverage imbalance, insufficient attention to climate-induced hazards, and reliance on simulations. Further investigations should strengthen experimental assessment and adopt multi-hazard approaches considering durability, economic feasibility, and vulnerability.
Accurate prediction of residual flexural strengths is essential for the structural design and performance assessment of steel fibre reinforced concrete (SFRC). While machine learning models have recently demonstrated strong predictive capability, most of the existing approaches rely on heuristic feature selection, unconstrained architectures, and random validation splits that may overestimate generalisation capacity. This study proposes a Hard-Constrained Physics-Informed Neural Network (PCNN) for predicting the residual flexural strengths fR1–fR4 of steel fibre reinforced concrete (SFRC). Non-negativity, monotonic dependence on the reinforcement index, and bounded sequential evolution across crack mouth opening displacement (CMOD) stages are enforced architecturally through custom layers, guaranteeing mechanical admissibility by construction. The model was trained and validated on 235 original beam specimens from 18 experimental groups (expanded to 860 via controlled integer replication) using GroupKFold by Study cross-validation to eliminate data leakage. PCNN achieves fR1 MAE = 1.256 MPa and R2 = 0.606 across held-out experimental campaigns, competitive with Ridge Regression (MAE = 1.144, R2 = 0.688). Monotonicity audits show a 0% violation rate for PCNN versus 7.3% and 10.2% for Random Forest and GBDT, respectively. PCNN is the only model in the comparison that guarantees physical admissibility for all possible input configurations. The results demonstrate that embedding mechanical principles directly into neural network architecture enhances robustness and interpretability without compromising accuracy. The proposed Physics-Informed Neural Networks framework provides a transferable methodology for physics-guided machine learning in structural materials modelling.
Urban soft soils exhibit high compressibility, low shear strength, and significant rheological characteristics, making roadside foundation pits highly susceptible to cyclic traffic disturbances. Repeated vehicle excitation easily induces abnormal deformation and internal force variation in retaining structures, threatening pit stability. To clarify the quantitative response mechanism of soft-soil foundation pits under multi-parameter traffic loads, this study establishes a refined three-dimensional numerical model using PLAXIS 3D Advanced CONNECT Edition V22, based on a typical coastal soft-soil excavation project in Guangdong Province. Different from previous simplified traffic-load simulations, the numerical model incorporated the soil small-strain stiffness constitutive model and realized the coupling of groundwater seepage and dynamic load fields. The influences of four key parameters, namely excavation depth, static vehicle load, dynamic load amplitude, and load frequency, on the deformation and internal force evolution of retaining walls were systematically analyzed. Quantitatively, the traffic-load contribution ratio averaged 27.86% for the additional structural responses. Parametric comparison shows that excavation depth dominates the bending moment response, while static vehicle load is the primary factor controlling wall deformation. Dynamic load amplitude and frequency produce relatively minor and comparable influences. Based on a clear decomposition method of static and dynamic load effects, static vehicle load accounts for 80–90% of the total structural response (wall displacement and bending moment), whereas the dynamic increment induced by pavement unevenness only contributes 10–20%. This study clarifies the differentiated contribution mechanism of multi-factor traffic loads and provides a scientific reference for dynamic simulation, structural optimization, and construction risk control of urban soft-soil foundation pits under vehicle cyclic loads.
