Experiments on concrete members indicate that the nominal strength of a specimen decreases with increasing specimen size for the same specimen geometry. This phenomenon is known as the size effect in the fracture mechanics of plain and reinforced concrete. Although nominal strength is also highly affected by the width of the distributed load in split-tension cylinder and cube specimens, this effect can be negligible within the practical range of the load-distributed width in diagonal cubes. However, the number of reported theoretical and experimental studies on cube and especially diagonal cube split-tension specimens is much more limited than those on cylindrical specimens. Therefore, in this study, six series of cube and diagonal cube specimens, with three different sizes and with two different maximum aggregate sizes (4 mm and 16 mm) but similar geometries, were tested under different load-distributed widths. The peak loads obtained from the test results were analysed using the multi-fractal scaling law. Subsequently, two prediction formulae are proposed for the design of torsional concrete/reinforced concrete members.

This study introduces a physics-guided self-supervised framework for few-shot ultrasonic defect detection in concrete structures, addressing the dual challenges of scarce labels and domain variability in structural health monitoring (SHM). Our method integrates physics-informed augmentations, contrastive representation learning, and adversarial domain alignment within a mutually reinforcing cycle, enabling robust defect classification with minimal supervision. A Physics-Informed Augmentation Module synthesizes realistic ultrasonic signals, training a Transformer encoder to extract invariant features while suppressing sensor noise. An Adversarial Feature Aligner further improves cross-domain generalization by mitigating distribution shifts across heterogeneous concretes. Experimental validation on three benchmark datasets demonstrates 63–66% accuracy in one-shot cross-domain tasks and up to 89% in five-shot settings. These results represent 12–15 percentage point gains over modern few-shot baselines, with improvements statistically significant at p < 0.001. Compatible with existing ultrasonic hardware, the proposed framework bridges physics-based modeling and machine learning while paving the way for scalable, field-ready SHM solutions for aging infrastructure and resilient smart cities.

This study proposes a field-calibrated, NDT-integrated BIM modeling framework to improve the reliability of post-earthquake assessment for reinforced concrete (RC) buildings. The approach combines destructive and nondestructive testing (NDT) data—including core drilling, Schmidt hammer, ultrasonic pulse velocity (UPV), and Windsor probe—through a site-specific WinSonReb regression model. The calibrated material properties (average compressive strength ≈ 18.6 MPa, CoV > 20%) were embedded into a Building Information Modeling (BIM) environment, producing an as-is, NDT-calibrated BIM model representing a Level-2 static digital twin of the structure. Nonlinear static pushover analyses performed in accordance with TBDY-2018 and ASCE 41-17 showed that the calibrated model exhibits a fundamental period of 0.85 s—approximately 18% longer than the uncalibrated BIM model. This elongation increased displacement demand and caused a shift in performance classification: while the uncalibrated model indicated Life Safety (LS), the calibrated model predicted behavior approaching Collapse Prevention (CP) in the Y direction. Furthermore, calibration reversed the predicted damage hierarchy, from ductile beam hinging to brittle column- and wall-controlled failure near elevator openings, consistent with post-event observations from the 2023 Kahramanmaraş earthquakes. These results demonstrate that integrating field-calibrated NDT data into BIM-based seismic models fundamentally alters both strength estimation and failure-mechanism prediction, reducing epistemic uncertainty and providing a more conservative basis for retrofit prioritization. Although demonstrated on a single case study, the proposed workflow offers a realistic and scalable pathway for NDT-supported seismic performance assessment of existing RC buildings.

This study proposes a holistic workflow to investigate the impact of systematic bias and uncertainty in Non-Destructive Testing (NDT)-based strength estimations on structural performance decisions for earthquake-damaged buildings. Element-level material properties were obtained using a field-calibrated multivariate WINSONREB model (combining Windsor Probe, UPV, and Schmidt hammer tests) and subsequently integrated into a Building Information Modeling (BIM) platform to develop an as-is digital representation of the structure. Quantitative analyses revealed that the proposed model systematically overestimates concrete strength by an average of +5 MPa, particularly in low-strength zones, and that the model's standard error (±5.7 MPa) is sufficient to shift the building's performance classification between Collapse Prevention (CP) and Life Safety (LS) levels under the Turkish Building Earthquake Code (TBEC). These findings highlight the critical importance of adopting a probabilistic framework for interpreting NDT data, as opposed to deterministic assumptions, to adequately capture uncertainty propagation in engineering decision-making.

6 Şubat 2023 Kahramanmaraş depremleri sonrasında Malatya’da orta hasar gören bir betonarme bina üzerinde tahribatsız test (NDT) çalışmaları yürütülmüştür. Bu kapsamda Windsor probu (WP) ve Schmidt çekici (SH) ile elde edilen veriler, karot basınç dayanımı (fc) sonuçlarıyla karşılaştırılarak kalibre edilmiştir. Analizler, WP yöntemi ile fc arasında güçlü bir ilişki (R² ≈ 0.80) bulunduğunu; SH’nin ise yüzey koşullarına duyarlılığı nedeniyle düşük güvenilirlik sergilediğini göstermiştir. SH ve WP verilerinin birlikte değerlendirildiği çoklu regresyon modeli daha yüksek açıklama gücüne ulaşmış (R² = 0.877, p < 0.05) ve literatürde önerilen kombine yaklaşımlarla uyumlu sonuçlar vermiştir. Kalibre edilen malzeme dayanım değerleri, mimari röleve ve yeniden çizilen projeler temel alınarak oluşturulan BIM tabanlı model ile Sta4CAD yazılımında kurulan yapısal analiz modeline entegre edilmiş; TBDY-2018 hükümleri doğrultusunda performans değerlendirmesi yapılmıştır. Bulgular, Eurocode-8’in TBDY-2018’e kıyasla daha yüksek deprem talepleri öngördüğünü ve bu farkların özellikle zayıf zemin koşullarında yapısal güvenlik açısından kritik olabileceğini ortaya koymuştur. Çalışma, deprem sonrası sınırlı karot imkânı (n = 20) bulunan saha koşullarında NDT yöntemlerinin pratik uygulanabilirliğine ilişkin bulgular sunmakta ve gelecekte Bayesian istatistiksel yöntemler ile yapay zekâ tabanlı modellerin (ANN, SVM, ANFIS) dayanım tahmini doğruluğunu artırma potansiyeline işaret etmektedir.