This study aims to improve the estimation of concrete compressive strength by combining destructive and non-destructive testing (ultrasound and sclerometer) to obtain a more precise and reliable assessment for quality control and the diagnosis of concrete structures. The experimental data were analyzed using a program linked to the Pundit Lab measurement tool to establish a regression model linking the ultrasonic propagation velocity (UPV) to the compressive strength of self-compacting concrete (SCC). SCC was formulated using a standard mix design with 450 kg/m3 of cement, 10% limestone filler (45 kg/m3), a water/cement ratio of 0.44, a gravel/sand ratio of 1.0, and a polycarboxylate-based superplasticizer. The fresh properties were evaluated using cone slump tests, L-box tests, V-funnel tests, and sieve stability tests, confirming the self-compacting nature of the mix. After 28 days of curing, five cylindrical specimens (160 × 320 mm) were subjected to ultrasonic (UPV), sclerometric, and compression tests. The results showed an average compressive strength of 39.06 MPa, a modulus of elasticity of 41.5 GPa, and a shear modulus of 16.7 GPa. The developed regression model improved the accuracy of compressive strength estimation by combining ultrasonic and sclerometric measurements, thus providing a reliable method for quality control and evaluation of existing concrete structures.
Asphalt pavement performance is highly temperature dependent due to the viscoelastic behavior of the binder. Although rheological characteristics of bitumen (e.g., softening point, elastic recovery) are widely analyzed, thermal properties as diffusivity (or its components conductivity and specific heat capacity) are rarely tested. The most up-to-date design processes use asphalt master-curves to calculate temperature-dependent stiffness for versatile weather conditions, yet the temperature susceptibility of the mixture is neglected. In this paper a full approach is presented in order to determine the sensitivity of the pavement to changing in thermal properties. Surface temperatures are based on real-life data, in-depth temperatures are calculated via 1-D Fourier's heat transfer equation with the primary governing variable of diffusivity. The stiffness of the mixture is predicted by the Witczak–Bari model for every centimeter of depth to get a full stiffness profile. Multi-layer elastic theory is used to calculate pavement response. To evaluate conditional numerical sensitivities within the selected parameter space, 200 different combinations of thermal conductivity and specific heat capacity were tested as theoretical inputs. The connection between fatigue life and these numerical parameters was found non-linear. It was discovered that an increase of 1 W/mK in thermal conductivity can cause an average fatigue life reduction of 7.2%, while a 100 J/kgK change in specific heat capacity increases fatigue life with 1.0%. The ultimate general sensitivity analysis to thermal diffusivity showed that in the typical range of this property relative fatigue life can vary between 88.5–104.0% of the base scenario.
Vision-language models (VLMs) have shown strong performance on multimodal reasoning tasks, yet their ability to perform quantitative analysis of technical drawings remains largely unexplored. This study evaluates the instruction-guided zero-shot performance of three open-source VLMs, on the task of estimating room areas from color-coded residential floor plan images. Of the three models tested, two produced quantitatively evaluable structure outputs. A dataset of 100 floor plans, comprising 472 plan-level room-type evaluation comparisons derived from 894 ground-truth room instances, was used to compare predicted and ground truth areas. Results show moderate predictive correlation (R2 approximate to 0.71), but substantial estimation errors (MAPE approximate to 43%), with noticeably larger errors for smaller rooms. A classical pixel counting baseline, using color segmentation, achieved MAPE = 5.68% at the room-instance level, and MAPE = 1.58% when evaluated at the floor plan level, highlighting the limitations of the tested VLMs for precise geometric estimation under this controlled setup. These findings suggest that while the evaluated open-source VLMs can approximate spatial proportions in simplified color-coded layouts, their quantitative accuracy remains significantly below that of deterministic image analysis methods.
Strategic investment in shared mobility infrastructure necessitates balancing multiple conflicting criteria. This study develops and applies an integrated Best-Worst Method (BWM) and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) framework to prioritize urban zones for shared mobility hub investment in Budapest, Hungary. Vector normalization was adopted in the TOPSIS stage for its scale-invariance properties, ensuring robust cross-criteria comparability. The methodology incorporates 17 sub-criteria across five main dimensions: Transport Connectivity, Demand Density, Infrastructure Readiness, Environmental Impact, and Social Equity. Expert judgments were synthesized using BWM and applied to evaluate four distinct urban typologies: the Historic Central Business District (Zone A), Mixed-Use Residential Corridor (Zone B), Major Transport Hub & University Area (Zone C), and Residential Suburb (Zone D).Results identify Zone B (Újlipótváros - Angyalföld) as the top priority, offering the best balance between high demand and implementation feasibility. Zone A (Belváros - Lipótváros), the historic CBD, ranks second, as its superior connectivity is offset by severe spatial constraints. Zone C (Kelenföld - Albertfalva), a major transport hub, places third, while Zone D (Péterhegy - Kőbánya), a residential suburb, ranks fourth due to lower current connectivity despite strong environmental and social equity scores.The analysis reveals a core planning trade-off: zones with the highest immediate demand face the greatest physical implementation challenges. The proposed framework provides municipal planners with a transparent, quantitative tool for optimizing phased infrastructure investment, with future work recommended to incorporate real-time and time-dependent mobility data for dynamic scenario analysis.
Polycarboxylate ether (PCE) superplasticizers are widely used in cementitious systems. They are also used to disperse solid particles in aqueous suspensions, including cement particles and nano-additives such as graphene oxide (GO). Their performance depends on both polymer formulation and dosage. This study compared four commercial PCEs in a NaOH-adjusted alkaline model medium. The electrokinetic and hydrodynamic behavior of GO suspensions was evaluated over GO:PCE mass ratios from 1:0 to 1:6. Zeta potential (ZP) and dynamic light scattering (DLS) were used as comparative descriptors. The control GO suspension showed a mean ZP of −58 mV, a Z-average hydrodynamic size of 531 nm, and a polydispersity index (PDI) of 0.37. After PCE addition, the PDI decreased to 0.26–0.30, indicating lower DLS-derived apparent heterogeneity relative to the control suspension. The response depended on PCE formulation. ACE 300 and MC-PowerFlow 2695 gave the most negative ZP values, reaching −76 mV and −65 mV, respectively, at a GO:PCE ratio of 1:5. At 1:6, the magnitude decreased relative to the peak values. By contrast, the ViscoCrete series showed a more moderate ZP response, from −53 to −58 mV, while the smallest apparent sizes were obtained with the ViscoCrete series, reaching 477 nm. These results show that GO dispersion in this alkaline model medium cannot be inferred from a single descriptor or from applying a single dosage rule across different PCE formulations. Joint interpretation of electrokinetic and hydrodynamic measurements is therefore needed for comparative screening.
