Concrete technology research has marked concrete's compressive strength as a critical indicator for predicting other vital mechanical properties. Existing models for forecasting this strength—from empirical formulas to complex machine learning algorithms—suffer from issues like lack of rationality and dimensional consistency. This study introduces the Modified Dimensional Analysis (MDA) method, which overcomes these challenges by ensuring dimensional homogeneity and mathematical stability, facilitating its use via simple tools like spreadsheets or calculators. We evaluated MDA against conventional models such as Abrams' law and various machine learning approaches, including decision trees, random forests, support vector machines, multiple linear regression, and artificial neural networks. Our findings reveal that MDA exhibits impressive data efficiency, requiring only 60% less training data than all other machine learning algorithms. It demonstrates a testing average R-square value of 0.94 and a maximum RMSE value of 5.55 MPa, which is competitive compared to models trained on 60% more datasets. Furthermore, if adjusted for specific construction sites, MDA will prove to be a robust on-field instrument, responsive to alterations in mix proportions, offering critical, timely insights for site engineers and supervisors. This adaptability allows for immediate, informed interventions, potentially improving construction outcomes.

Climate projections at fine spatial resolutions are required to conduct accurate risk assessment for critical infrastructure and design adaptation planning. Generating these projections using advanced Earth system models (ESM) requires significant computational resources. To address this issue, various statistical downscaling techniques have been introduced to generate fine-resolution data from coarse-resolution simulations. In this study, we evaluate and compare five deep learning-based downscaling techniques, namely, super-resolution convolutional neural networks, fast super-resolution convolutional neural network ESM, efficient sub-pixel convolutional neural network, enhanced deep residual network (EDRN), and super-resolution generative adversarial network (SRGAN). These techniques are applied to a dataset generated by the Energy Exascale Earth System Model (E3SM), focusing on key surface variables such as surface temperature, shortwave heat flux, and longwave heat flux. Models are trained and validated using paired fine-resolution (0.25$$^{\circ }$$ ∘ ) and coarse-resolution (1$$^{\circ }$$ ∘ ) monthly data obtained from a 9-year simulation. Next, blind testing is performed using monthly data obtained from two different years outside of the training and validation set. To evaluate the efficiency of each technique, different statistical metrics are used, including mean squared error (MSE), peak signal-to-noise ratio (PSNR), structural similarity index measure (SSIM), and learned perceptual image patch similarity (LPIPS). The results show that EDRN outperforms other algorithms in terms of PSNR, SSIM, and MSE, but struggles to capture fine-scale features in the data. In contrast, SRGAN, a generative model that uses perceptual loss, excels in capturing fine details at boundaries and internal structures, resulting in lower LPIPS than other methods.

Recent breakthroughs in computing power have made it feasible to use machine learning and deep learning to advance scientific computing in many fields, including fluid mechanics, solid mechanics, materials science, etc. Neural networks, in particular, play a central role in this hybridization. Due to their intrinsic architecture, conventional neural networks cannot be successfully trained and scoped when data is sparse, which is the case in many scientific and engineering domains. Nonetheless, neural networks provide a solid foundation to respect physics-driven or knowledge-based constraints during training. Generally speaking, there are three distinct neural network frameworks to enforce the underlying physics: (i) physics-guided neural networks (PgNNs), (ii) physics-informed neural networks (PiNNs), and (iii) physics-encoded neural networks (PeNNs). These methods provide distinct advantages for accelerating the numerical modeling of complex multiscale multi-physics phenomena. In addition, the recent developments in neural operators (NOs) add another dimension to these new simulation paradigms, especially when the real-time prediction of complex multi-physics systems is required. All these models also come with their own unique drawbacks and limitations that call for further fundamental research. This study aims to present a review of the four neural network frameworks (i.e., PgNNs, PiNNs, PeNNs, and NOs) used in scientific computing research. The state-of-the-art architectures and their applications are reviewed, limitations are discussed, and future research opportunities in terms of improving algorithms, considering causalities, expanding applications, and coupling scientific and deep learning solvers are presented. This critical review provides researchers and engineers with a solid starting point to comprehend how to integrate different layers of physics into neural networks.

Resolving solute transport in heterogeneous porous media is a complex task, because of the sparse experimental data and the high computational cost of numerical simulations. This work proposes a unique two-stage deep learning architecture comprising a dual-branch autoencoder and a geo-guided super-resolution generative adversarial network (Gg-SRGAN) to address this dual challenge. The dual-branch autoencoder addresses the issue of sparsity by constructing a continuous, but coarse representation of concentration and pressure profiles from a sparse, discontinuous profile with up to 85% missing data points. The Gg-SRGAN is then employed to generate a finer representation of field variables from the outputs generated by the dual-branch autoencoder (i.e., downscaling). We train and test our framework using six solute transport cases with varying levels of heterogeneity and compare the results with standalone methods, namely the vanilla autoencoder and vanilla SRGAN, in addition to ground truth profiles generated by the finite element method (FEM). The comparisons are performed based on several statistical metrics, such as absolute point error (APE), mean squared error (MSE), peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), and learned perceptual image patch similarity (LPIPS). The first four cases are used for training, evaluation, and testing. The last two cases are utilized for blind testing to determine the generalizability of the framework. Our results show that the dual-branch autoencoder outperforms the vanilla autoencoder, and the Gg-SRGAN outperforms the SRGAN during both the training and evaluation phases. Moreover, the proposed framework can successfully construct the fine representation of concentration profiles, compared to FEM, using the coarse representation of the pressure, concentration, and domain permeability fields. When tested using the two blind test cases, the proposed dual-branch autoencoder and Gg-SRGAN exhibit superior performance compared to their counterparts in terms of all evaluation metrics.

The duration of a construction project is a key factor to consider before starting a new project. It needs to be accurately estimated from an early stage. Many researchers demonstrated the applicability of regression analysis (RA) in preliminary duration estimation for construction projects; however, RA and similar models fail to simulate the complex behavior of problems in estimating. In contrast, artificial neural networks (ANNs) have several significant benefits that make them powerful and practical for solving complex problems in the field of construction engineering and modeling nonlinearity in the data. Nevertheless, ANNs have constraints because of the absence of structured methodology to decide on various control features and their "black box" nature, which does not explain the underlying input-output process. Moreover, unlike construction cost, construction duration is not determined by the summation of all activities, but only by critical activities. Given these factors, this work presents a feature selection method while applying ANNs for estimating construction duration in the preliminary stage, and proposes a two-stage ANN to take into account the specific nature of construction duration. The results confirm the potential of two-stage ANNs and feature selection by sensitivity analysis to provide a more accurate estimate of construction duration and unlock potential knowledge in the network system to increase user confidence in ANN use.