Search

Despite its rapid development, Physics-Informed Neural Network (PINN)-based computational solid mechanics is still in its infancy. In PINN, the loss function plays a critical role that significantly influences the performance of the predictions. In this paper, by using the Least Squares Weighted Residual (LSWR) method, we proposed a modified loss function, namely the LSWR loss function, which is tailored to a dimensionless form with only one manually determined parameter. Based on the LSWR loss function, an advanced PINN technique is developed for computational 2D and 3D solid mechanics. The performance of the proposed PINN technique with the LSWR loss function is tested through 2D and 3D (geometrically nonlinear) problems. Thoroughly studies and comparisons are conducted between the two existing loss functions, the energy-based loss function and the collocation loss function, and the proposed LSWR loss function. Through numerical experiments, we show that the PINN based on the LSWR loss function is effective, robust, and accurate for predicting both the displacement and stress fields. The source codes for the numerical examples in this work are available at https://github.com/JinshuaiBai/LSWR_loss_function_PINN/. [ABSTRACT FROM AUTHOR]

Our recent study has found that physics-informed neural networks (PINN) tend to be local approximators after training. This observation led to the development of a novel physics-informed radial basis network (PIRBN), which is capable of maintaining the local approximating property throughout the entire training process. Unlike deep neural networks, a PIRBN comprises only one hidden layer and a radial basis "activation" function. Under appropriate conditions, we demonstrated that the training of PIRBNs using gradient descendent methods can converge to Gaussian processes. Besides, we studied the training dynamics of PIRBN via the neural tangent kernel (NTK) theory. In addition, comprehensive investigations regarding the initialisation strategies of PIRBN were conducted. Numerical examples demonstrated that PIRBN is more effective than PINN in solving nonlinear partial differential equations with high-frequency features and ill-posed computational domains. Moreover, the existing PINN numerical techniques, such as adaptive learning, decomposition and different types of loss functions, are applicable to PIRBN. The programs that can regenerate all numerical results are available at https://github.com/JinshuaiBai/PIRBN. • We found that PINNs tend to be local approximators during training process. • We proposed the physics-informed radial basis network (PIRBN). • We proved that training a PIRBN can converge to Gaussian Processes. • We proved that PIRBNs can maintain the local approximating property during training. • PIRBN is effective for problems with high-frequency features and ill-posed domains. [ABSTRACT FROM AUTHOR]

• A new PINN-based topology optimization framework is proposed. • The proposed framework employs PINN to conduct structural analysis to replace FEA. • The proposed framework does not require any labelled data to train PINN. • The proposed framework can effectively optimize the problem domains. • Comparisons and discussions are provided between the topology optimization methods. Physics-Informed Neural Networks (PINNs) have recently attracted exponentially increasing attention in the field of computational mechanics. This paper proposes a novel topology optimization framework: Physics-Informed Neural Network-based Topology Optimization (PINNTO). Unlike existing machine-learning based topology optimization frameworks, PINNTO employs an energy-based PINN to replace Finite Element Analysis (FEA) in the conventional structural topology optimization, to numerically determine the deformation states, which is a key novelty in the proposed methodology. A supervised neural network that respects governing physical laws defined via partial differential equations is trained to develop the corresponding network without any labelled data, with the intention of solving solid mechanics problems. To assess feasibility and potential of the proposed PINNTO framework, a number of topology-optimization-related case studies have been implemented. The subsequent findings illustrate that PINNTO has the ability to attain optimized topologies with neither labelled data nor FEA. In addition, it has the capability to generate comparable designs to those produced by the current successful approaches such as Solid Isotropic Material with Penalization (SIMP). Based on the results of this study, it can also be deduced that PINNTO can acquire optimal topologies for various types of complex domains given that the boundary conditions and loading configurations are correctly imposed for the associated energy-based PINN. Consequently, the proposed PINNTO framework has demonstrated promising capabilities to solve problems under conditions when the usage of FEA is challenged (if not impossible). In summary, the proposed PINNTO framework opens up a new avenue for structural design in this ' data-rich ' age. [ABSTRACT FROM AUTHOR]

Neural Particle Method (NPM) is a newly proposed Physics-Informed Neural Network (PINN) based, truly meshfree method for hydrodynamics modeling. In the NPM, PINN is applied to globally approximate field variables, and the high-order Implicit Runge–Kutta (IRK) method is used to treat the time integration. The NPM can easily achieve incompressibility and deal with the free-surface problem. However, the NPM is currently limited to inviscid hydrodynamics problems and is computationally expensive. In this work, we developed a general NPM (gNPM) for viscous hydrodynamics modeling. In the gNPM, a single pressure is output as the predicted pressure field rather than the multiple pressures in the original NPM. Thus, the size of the neural network is greatly reduced, making the gNPM computationally more efficient. Besides, the spatial derivatives in the governing equations are calculated with respect to the current spatial coordinates rather than the predicted space. In this manner, the gNPM is more straightforward to be implemented. Furthermore, by considering the viscous term in the conservation of momentum, the gNPM can be applied for viscous hydrodynamics modeling. The effectiveness and robustness of the gNPM have been demonstrated through several hydrodynamics benchmark cases with different boundary conditions. We highlight that the proposed gNPM is able to cope with highly uneven particle distributions, while the traditional meshfree method can produce severe failure. • A general Neural Particle Method (gNPM) is proposed for hydrodynamics modeling. • The gNPM is computationally more efficient and easier to be implemented than the NPM. • The effectiveness of the gNPM has been demonstrated through benchmark cases. • The gNPM is more robust than the SPH for uneven particle distributions. [ABSTRACT FROM AUTHOR]

Musculoskeletal conditions affect an estimated 1.7 billion people worldwide, causing intense pain and disability. These conditions lead to 30 million emergency room visits yearly, and the numbers are only increasing. However, diagnosing musculoskeletal issues can be challenging, especially in emergencies where quick decisions are necessary. Deep learning (DL) has shown promise in various medical applications. However, previous methods had poor performance and a lack of transparency in detecting shoulder abnormalities on X-ray images due to a lack of training data and better representation of features. This often resulted in overfitting, poor generalisation, and potential bias in decision-making. To address these issues, a new trustworthy DL framework has been proposed to detect shoulder abnormalities (such as fractures, deformities, and arthritis) using X-ray images. The framework consists of two parts: same-domain transfer learning (TL) to mitigate imageNet mismatch and feature fusion to reduce error rates and improve trust in the final result. Same-domain TL involves training pre-trained models on a large number of labelled X-ray images from various body parts and fine-tuning them on the target dataset of shoulder X-ray images. Feature fusion combines the extracted features with seven DL models to train several ML classifiers. The proposed framework achieved an excellent accuracy rate of 99.2%, F1Score of 99.2%, and Cohen's kappa of 98.5%. Furthermore, the accuracy of the results was validated using three visualisation tools, including gradient-based class activation heat map (Grad CAM), activation visualisation, and locally interpretable model-independent explanations (LIME). The proposed framework outperformed previous DL methods and three orthopaedic surgeons invited to classify the test set, who obtained an average accuracy of 79.1%. The proposed framework has proven effective and robust, improving generalisation and increasing trust in the final results. [ABSTRACT FROM AUTHOR]

Physics-Informed Neural Networks (PINNs) have recently gained increasing attention in the field of topology optimization. The fusion of deep learning and topology optimization has emerged as a prominent area of insightful research, where minimization of the loss function in neural networks can be comparable to minimization of the objective function in topology optimization. Inspired by concepts of PINNs, this paper proposes a novel framework, ' Complete Physics-Informed Neural Network-based Topology Optimization (CPINNTO) ', to address various challenges in topology optimization, particularly related to structural optimization. The key innovation of the proposed framework lies in introducing the first complete machine-learning-based topology optimization framework through integration of two distinct PINNs. Herein, the Deep Energy Method (DEM) PINN is implemented to determine the deformation state of corresponding structures numerically. In addition, derivation of the objective function with respect to design variables is replaced with automatic differentiation in sensitivity-analysis PINN (S-PINN). The feasibility and potential of the CPINNTO framework have been assessed through several case studies while highlighting strengths and limitations of utilizing PINNs in topology optimization. Subsequent findings indicate that CPINNTO can achieve optimal topologies without labeled data nor FEA. The numerical examples demonstrate that CPINNTO is capable of stably obtaining optimal structures for various topology optimization applications, including compliance minimization problems, multi-constrained problems, and three-dimensional problems. Resulting designs exhibit favorable compliance values comparable to the designs obtained via density-based topology optimization. In summary, the proposed CPINNTO framework opens up novel and interesting possibilities for structural topology optimization. • A novel Complete PINN-based topology optimization framework (CPINNTO). • Two PINNs to replace structural and sensitivity analyses in topology optimization. • CPINNTO does not require any labeled data or FEA to train PINNs. • CPINNTO can effectively optimize problem domains of both 2-D and 3-D structures. [ABSTRACT FROM AUTHOR]

• Identify gaps in state-of-the-art research to support healthcare AI trustworthiness. • Explore the explainable AI and fusion in healthcare. • Explore legitimacy, morality, and robustness standards for AI policymakers. • Assessment quality, bias risk, and data fusion for medical AI trustworthiness. • Examine eight research gaps to guide future study and assist researchers. In the last few years, the trend in health care of embracing artificial intelligence (AI) has dramatically changed the medical landscape. Medical centres have adopted AI applications to increase the accuracy of disease diagnosis and mitigate health risks. AI applications have changed rules and policies related to healthcare practice and work ethics. However, building trustworthy and explainable AI (XAI) in healthcare systems is still in its early stages. Specifically, the European Union has stated that AI must be human-centred and trustworthy, whereas in the healthcare sector, low methodological quality and high bias risk have become major concerns. This study endeavours to offer a systematic review of the trustworthiness and explainability of AI applications in healthcare, incorporating the assessment of quality, bias risk, and data fusion to supplement previous studies and provide more accurate and definitive findings. Likewise, 64 recent contributions on the trustworthiness of AI in healthcare from multiple databases (i.e., ScienceDirect, Scopus, Web of Science, and IEEE Xplore) were identified using a rigorous literature search method and selection criteria. The considered papers were categorised into a coherent and systematic classification including seven categories: explainable robotics, prediction, decision support, blockchain, transparency, digital health, and review. In this paper, we have presented a systematic and comprehensive analysis of earlier studies and opened the door to potential future studies by discussing in depth the challenges, motivations, and recommendations. In this study a systematic science mapping analysis in order to reorganise and summarise the results of earlier studies to address the issues of trustworthiness and objectivity was also performed. Moreover, this work has provided decisive evidence for the trustworthiness of AI in health care by presenting eight current state-of-the-art critical analyses regarding those more relevant research gaps. In addition, to the best of our knowledge, this study is the first to investigate the feasibility of utilising trustworthy and XAI applications in healthcare, by incorporating data fusion techniques and connecting various important pieces of information from available healthcare datasets and AI algorithms. The analysis of the revised contributions revealed crucial implications for academics and practitioners, and then potential methodological aspects to enhance the trustworthiness of AI applications in the medical sector were reviewed. Successively, the theoretical concept and current use of 17 XAI methods in health care were addressed. Finally, several objectives and guidelines were provided to policymakers to establish electronic health-care systems focused on achieving relevant features such as legitimacy, morality, and robustness. Several types of information fusion in healthcare were focused on in this study, including data, feature, image, decision, multimodal, hybrid, and temporal. [ABSTRACT FROM AUTHOR]

The degree of compaction in the levee building materials is a crucial factor that affects the piping phenomena. The density and compaction of the soil strata determine the structural soundness of the levee. Segments with reduced density or compaction can become weak spots during floods. To assess part of the Helena levee (2,500 m) in Arkansas (AR), the United States, an extensive ground-penetrating radar (GPR) fieldwork was conducted. This reliable method will undoubtedly improve the assessment procedure of the levee structure by identifying the weak spots within the structure that result from poor compaction of the levee core layers in a short time with high accuracy. However, interpreting the GPR data can be challenging and requires specialised knowledge. Obtaining meaningful insights typically involves a time-consuming process of extensive manual processing and visual inspection. To address this issue, this article proposes a novel, reliable deep-feature fusion framework for GPR data to identify horizontal variation in the soil compaction of a levee. To address data scarcity, a new type of transfer learning in the same domain is adopted, and four deep learning models (Xception, Inception, EfficientNet and MobileNet) are used to extract features. The combined features are then used to train and test five machine learning classifiers (Neural Network, Support Vector Machine, K-Nearest Neighbour, Logistic Regression, and Naive Bayes). The best combination of deep Learning and machine learning is four models with the neural network classifier which achieved the highest results by obtaining an accuracy of 98.2%, an F1 score of 97.6%, and an area under the curve of 99.9%. The proposed framework faced an additional challenge when subjected to an unseen dataset of 1,511 images reserved primarily for testing. Remarkably, it achieved an accuracy rate of 95.7% with the neural network classifier. This article presents a new research direction that has substantial potential in various domains, including civil engineering, the petroleum sector, road safety, agriculture, and more. • A Reliable deep feature fusion technique for the Ground Penetrating Radar (GPR) Data. • A new type of transfer learning is adopted to address training data scarcity issues. • Feature fusion is performed on four DL models. • Five machine learning classifiers have been employed. • The proposed framework achieved an accuracy of 98.2%, an F1 score of 97.6% and an AUC of 99.9 %. [ABSTRACT FROM AUTHOR]

Interfacial shear strength determines the load transfer efficiency between nanofillers and polymer matrix, and thus the overall mechanical performance of polymer nanocomposites. Through atomistic simulations, this work systematically assessed the sliding behaviours of one-dimensional carbon nanothread (NTH) from a bundle configuration embedded in the poly (methyl methacrylate) (PMMA) matrix. It is found that the functionalized NTH exhibits remarkably enhanced interfacial shear strengths due to the strong mechanical interlocking effect, about an order of magnitude higher than the pristine sample. However, excess volume is generated within a functionalized bundle structure, leading to a weaker non-bonded interaction for a functionalized NTH within the bundle than that only embedded in the polymer matrix. With covalent cross-linking, the surrounding NTHs can be simultaneously dragged out from the PMMA matrix while pulling out the central NTH. Such phenomenon attributes to the filler-filler interactions and filler-matrix interactions. Further studies reveal that no extra free volume is generated within the bundle structure at high temperatures, which retains the effective load transfer efficiency. In comparison, free volume tends to be generated at the interface between NTHs and PMMA matrix at a higher temperature. These findings could be beneficial for fibre design and high-performance polymer nanocomposites with 1D nanomaterials. Cross-linked carbon nanothreads can be simultaneously pulled out from the polymer matrix with the central filler, suggesting a remarkably enhanced interfacial load transfer efficiency. [Display omitted] [ABSTRACT FROM AUTHOR]

Polymer nanocomposites with excellent mechanical performances have been increasingly sought after in engineering applications such as biotechnology, aerospace, and automotive areas. Through molecular dynamics (MD) simulation, this work systematically assessed the tensile performance of poly(methyl methacrylate) (PMMA) nanocomposite reinforced by randomly dispersed two-dimensional diamond-diamane. It is found that randomly dispersed diamane effectively enhances the tensile properties of PMMA with surface functionalization, and the enhancement effect can be remarkably augmented by cross-linking. Simulations reveal that the enhancement effect can be effectively tailored by the alignment of the diamane fillers. The PMMA nanocomposites exhibit much better tensile performance when the diamane fillers are uniformly aligned along the in-plane direction of the filler. Additional investigations show that larger diamane filler is preferred without cross-linking, while smaller diamane filler should be considered when cross-linking is present. Overall, the impacts of different factors on the tensile properties of PMMA nanocomposites are analysed in-depth in this work, which provides atomistic insights for the preparation of polymer nanocomposites with desired mechanical properties. Effective enhancement on the tensile properties of polymer nanocomposites can be achieved by 2D diamond nanofiller. [Display omitted] [ABSTRACT FROM AUTHOR]

• A new data-driven topology optimization (DDTO) framework is proposed. • Sensitivity analysis for discrete noisy material data sets is developed. • Verification of the new DDTO framework by truss structure designs. The application of structural topology optimization with complex engineering materials is largely hindered due to the complexity in phenomenological or physical constitutive modeling from experimental or computational material data sets. In this paper, we propose a new data-driven topology optimization (DDTO) framework to break through the limitation with the direct usage of discrete material data sets in lieu of constitutive models to describe the material behaviors. This new DDTO framework employs the recently developed data-driven computational mechanics for structural analysis which integrates prescribed material data sets into the computational formulations. Sensitivity analysis is formulated by applying the adjoint method where the tangent modulus of prescribed uniaxial stress-strain data is evaluated by means of moving least square approximation. The validity of the proposed framework is well demonstrated by the truss topology optimization examples. The proposed DDTO framework will provide a great flexibility in structural design for real applications. [ABSTRACT FROM AUTHOR]