Search

Abstract Concrete compressive strength (CS) is a crucial performance parameter in concrete structure design. Reliable strength prediction reduces costs and time in design and prevents material waste from extensive mixture trials. Machine learning techniques solve structural engineering challenges such as CS prediction. This study used Machine Learning (ML) models to enhance the prediction of CS, analyzing 1030 experimental CS data ranging from 2.33 to 82.60 MPa from previous research databases. The ML models included both non-ensemble and ensemble types. The non-ensemble models were regression-based, evolutionary, neural network, and fuzzy-inference-system. Meanwhile, the ensemble models consisted of adaptive boosting, random forest, and gradient boosting. There were eight input parameters: cement, blast-furnace-slag, aggregates (coarse and fine), fly ash, water, superplasticizer, and curing days, with the CS as the output. Comprehensive performance evaluations include visual and quantitative methods and k-fold cross-validation to assess the study’s reliability and accuracy. A sensitivity analysis using Shapley-Additive-exPlanations (SHAP) was conducted to understand better how each input variable affects CS. The findings showed that the Categorical-Gradient-Boosting (CatBoost) model was the most accurate prediction during the testing stage. It had the highest determination-coefficient (R2) of 0.966 and the lowest Root-Mean-Square-Error (RMSE) of 3.06 MPa. The SHAP analysis showed that the age of the concrete was the most critical factor in the predictive accuracy. Finally, a Graphical User Interface (GUI) was offered for designers to predict concrete CS quickly and economically instead of costly computational or experimental tests.

ABSTRACTThis study utilized machine learning models to predict the discharge coefficient for a sharp-crested triangular side orifice (TSO). The chosen models were the Artificial Neural Network (ANN) and Gene Expression Programming (GEP). Development of the models was based on 570 experimental datasets, with 70% allocated for training and the remaining 30% for testing. Five nondimensional parameters were utilized as inputs for the models, including TSO’s crest height to its height (W*=W/H), main channel width to TSO’s base length (L*=B/L), main channel width to TSO’s height (H*=B/H), upstream flow depth to the TSO’s height (Y*=y1/H), and upstream Froude number of the main channel (Fr). While the discharge coefficient (Cd) was defined as the output. Then, the developed models were evaluated by three performance metrics, violin boxplots, and Taylor diagrams to ensure their reliability and accuracy. Furthermore, a sensitivity analysis was conducted to indicate the most effective parameter affecting the Cd value. The findings revealed that both models predicted very well compared to the actual values, with the ANN model emerging as the most reliable predictor. It exhibited the highest determination coefficient (R2), nearing 1, along with the lowest Mean-Square-Error (MSE) and Mean-Absolute-Error (MAE) values, both close to zero. The sensitivity analysis highlighted that the orifice crest height and Froude number significantly impacted the Cd value, contributing to more than 36%. In addition, the predicted discharge coefficient stayed within the range of ± 5.0% of the experimental values. Finally, the developed models demonstrated a high level of equivalence compared to previous studies, especially the ANN model. Therefore, these models are recommended as accurate, robust, and rapid tools to predict the TSO’s discharge coefficient.

Euler deconvolution is widely used for interpreting magnetic anomalies as it estimates the edges and depths of magnetic sources. Since this method was proposed, there has been an intensive effort to mitigate its primary deficiencies, namely, the generation of many spurious solutions and the high noise sensitivity. To select the most significant solutions, we adopt the strategy of constraining the moving window to the source edges, whose locations are estimated using the enhanced horizontal gradient amplitude method. On the other hand, we reduce noise propagation by performing a stable calculation of the vertical derivatives. For this purpose, we use the β-VDR method, a finite-difference method that yields a robust approximation of the vertical derivatives of magnetic data. The accuracy of the proposed technique is demonstrated on synthetic magnetic anomalies, providing the depths more precisely and being insensitive to noise. Application of this technique is also demonstrated on aeromagnetic anomalies from the Olympic Peninsula (USA), where the obtained result is in good agreement with known information of the study region.

The estimation of the Curie point depth (CPD) and heat flow (HF) from magnetic data is useful for geothermal and tectonic studies. However, the methodology of estimating these values is still controversial as numerous studies have used different types of transformed magnetic data. Most studies utilize the total magnetic intensity (TMI) or the reduced to the pole (RTP) data, and since shallow magnetic bodies may affect the final CPD values, low-pass (LP) wavelength filtering has also been used to estimate the CPD. In this study, the two-dimensional radially averaged spectral analysis was applied to the land-based magnetic data from the Sinai Peninsula. TMI and RTP data were both analyzed, using different LP filtered wavelengths, and the final CPD and HF values were compared. The cut-off wavelength parameter was selected in terms of the geologic complexity. The results show that the estimated CPD values are well correlated regardless of whether the magnetic data were either TMI or RTP. The RTP data produced a spatial shift in the CPD smaller than the window size being used. Additionally, the centroid method is primarily dependent on the wavenumber, which varied slightly on applying the RTP. There were slight differences in the estimated CPDs when a short LP filter (e.g., 20 km) was applied, whereas increasing the wavelength (e.g., 40 km) caused the CPD values to increase exceeding the plausible depth limits, and the associated spatial trends varied when compared to the other methods. In areas where the basement rocks were shallow or exposed, a LP filter can be applied with short cut-off wavelength. The estimated CPDs were discussed in the context of seismicity data, gravity-based Moho depths, and HF measurements.

Structural lineaments of Southeast Nigeria were delineated using high-resolution aeromagnetic data. Advanced methods such as enhanced horizontal gradient amplitude (EHGA) involving a 3D model and tilt depth method (TDM) were used in this study. The simulated magnetic model involving the EHGA detector positioned peaks over source borders and created sharp and clear edges for magnetic sources. The TDM is a cutting-edge depth determination method revealing the depth of the contacts and thickness of sedimentary bodies that vary from ∼500 to ∼2,500 m and ∼3,000 to ∼5,000 m, respectively. Lineaments of the study area were extracted from the EHGA map. The structural map showed the dominance of short-ranged (∼0.29 to ∼1.48 km) linear magnetic anomalies. In addition, regional geologic structures (∼2.8 to ∼24.5 km) occur at the western flank of the study area. In general, these structures trend in the NE–SW, NNE–SSW, N–S, E–W, and NW–SE directions. They are indicators of subsurface faults, contacts, and tectonic structures of the thermo-tectonic events associated with Santonian Abakaliki Anticlinorium, Tertiary-Recent Ikom-Mamfe Rift, and structural deformations of the basement rocks associated with the Oban Massif.

During the Late Campanian, sea-level fluctuation occurred in the form of two chert bands and the topmost conglomeratic phosphate beds (CF8a zone). A great transgression event occurred associated with the global warming (CF8b zone) trend indicated by large double-keeled foraminifera as Globotruncana aegyptiaca Nakkady of oligotrophic conditions. Through the event of CF6/CF5 zones, a gradual increase in the relative abundance of Gublerina rajagopalani Govindan and Planoheterohelix planata (Cushman) and other heterohelicids have been shown to tolerate and thrive in a wide range of environmental conditions as in high-stress environments. Sea-level fall at the CF6 zone and the overlying CF5 boundary marked a warming climate for the Middle Maastrichtian Event. In the latest CF4 records, the beginning of the decrease in planktic/benthic ratio, globotruncanids, rugoglobigerinids, and heterohelicids indicating a fall in sea level coincided with the CF4/CF3 and the development of dark grey shales in a regressive regime. The observed low abundance of planktic specimens may be due to the presence of pyrite with black shale interval suggesting low oxygen condition. The latest zones CF3, CF2, and CF1 are generally characterized by heterohelicids blooms specially Planoheterohelix globulosa (Ehrenberg), and a gradual decrease in diversity associated with the warming phase before the K/Pg boundary event, implying high biotic stress to even absent of Guembelitria cretacea species through CF3–CF1 zones. Pseudotextularia elegans (Rzehak) occurs in the zones CF4 and CF2 warming of phases 1 and 2 of Deccan Trap Volcanic. The absence of P. elegans (Rzehak) in the zone (CF3) is postulated due to a tectonic cause (maximum cooling of Deccan Trap Volcanic) whereas its absence in the zone (CF1) was due to regression of phase 2.