Article
Thermodynamics
R. M. Farias, P. R. F. Teixeira, L. O. Vilarinho
Summary: This study proposes two variable profile heat sources with only two geometrical parameters to achieve accurate simulation results for various welding processes, base materials, and plate thicknesses. Numerical simulations in ANSYS (R) APDL software show that the new heat sources have equal or better agreement with experimental weld profiles and thermal cycles compared to classical heat sources, namely the 3D Conical and the Double Ellipsoid.
INTERNATIONAL JOURNAL OF THERMAL SCIENCES
(2022)
Article
Materials Science, Multidisciplinary
Peihao Geng, Guoliang Qin, Hong Ma, Jun Zhou, Ninshu Ma
Summary: Linear friction welding technology is increasingly used in aerospace industries for manufacturing bimetallic turbine blisks. The microstructural evolution and microhardness of FGH4096 and GH4169 superalloys during LFW were investigated. The study found distinct behavior in different areas of the joint, emphasizing the need for further numerical simulations and analysis.
JOURNAL OF MATERIALS RESEARCH AND TECHNOLOGY-JMR&T
(2021)
Article
Multidisciplinary Sciences
M. Soleimani, B. Dashtbozorg, M. Mirkhalaf, S. M. Mirkhalaf
Summary: Atherosclerosis is a medical condition characterized by the hardening and thickening of arteries' walls. This study applied machine learning, particularly artificial neural networks (ANN), to improve the computational efficiency of atherosclerosis models. By training and validating an ANN model using a database of multi-physics Finite Element Method (FEM) simulations, the model achieved quick and accurate predictions of atherosclerosis development. The use of the ANN model led to significant computational gains compared to the original FEM simulations.
Article
Engineering, Multidisciplinary
Lucas Nogueira Garpelli, Diogo Stuani Alves, Katia Lucchesi Cavalca, Helio Fiori de Castro
Summary: Rotary systems are crucial for industrial production, but they are prone to unbalance faults. This paper proposes a method using Physics-Guided Neural Networks to identify and prevent rotor unbalance faults. The results show that the Physics-Guided Neural Networks have smaller errors and better performance compared to standard Artificial Neural Networks.
STRUCTURAL HEALTH MONITORING-AN INTERNATIONAL JOURNAL
(2023)
Article
Materials Science, Multidisciplinary
Bharat Singh, Piyush Singhal, Kuldeep K. Saxena, Ravindra K. Saxena
Summary: This study evaluates the significance of considering the latent heat effect during phase change in a three-dimensional finite element numerical simulation of arc welding. The developed mathematical model incorporating latent heat effect provides accurate predictions compared to experimental values. It is crucial to account for the latent heat of fusion, solidification, and vaporization in numerical simulations involving phase change analysis.
METALS AND MATERIALS INTERNATIONAL
(2021)
Article
Engineering, Manufacturing
Francisco Werley Cipriano Farias, Joao da Cruz Payao Filho, Victor Hugo Pereira Moraes e Oliveira
Summary: Finite element method (FEM) simulations are used to analyze the effect of idle time on interpass temperature in wire arc additive manufacturing, and an artificial neural network (ANN) is developed for predictive modeling to help manufacturers achieve a balance between productivity and part behavior.
ADDITIVE MANUFACTURING
(2021)
Article
Construction & Building Technology
Taher Ghalandari, David Hernando, Navid Hasheminejad, Mahesh Moenielal, Cedric Vuye
Summary: This study assesses the feasibility of harvesting heat from asphalt pavement and evaluates the structural performance of a Pavement Solar Collector (PSC) in Utrecht, The Netherlands. Experimental tests and numerical simulations reveal that the PSC system can reduce damage and has a high heat harvesting capacity.
CASE STUDIES IN CONSTRUCTION MATERIALS
(2023)
Article
Chemistry, Multidisciplinary
Md Arifuzzaman, Md Alhaz Uddin, Mohammed Jameel, Mohammad Towhidur Rahman Bhuiyan
Summary: This paper investigates the application of artificial neural networks (ANNs) to predict the response of a spar platform. The study finds that ANNs can effectively predict platform responses and significantly reduce the required time compared to conventional finite element methods.
APPLIED SCIENCES-BASEL
(2022)
Article
Engineering, Chemical
Jin Zhang, Qiong Xu
Summary: This study investigates the element transfer behavior of Cr2O3-bearing agglomerated fluxes in submerged arc welding. The gas-slag-metal equilibrium consideration is shown to qualitatively analyze the transfer behaviors, and a thermodynamic approach to predict element transfer levels is proposed.
Article
Construction & Building Technology
Zi-Qing Yuan, Yu Xin, Zuo-Cai Wang, Ya-Jie Ding, Jun Wang, Dong-Hui Wang
Summary: This study proposes a novel nonlinear model updating approach based on an improved generative adversarial network (GAN). The improved GAN incorporates a convolutional neural network (CNN) surrogate model into the discriminator network to enhance its learning capability. The trained network model based on measured acceleration amplitudes can accurately estimate the nonlinear model parameters. Numerical simulations and experimental tests confirm the reliability and effectiveness of the improved GAN model for structural nonlinear model updating under seismic excitations.
STRUCTURAL CONTROL & HEALTH MONITORING
(2023)
Article
Engineering, Manufacturing
Zhijie Zhou, Shupeng Cai, Yuanqing Chi, Lintao Wei, Yongkang Zhang
Summary: A new technique called Weld-LSF has been proposed for low-stress manufacturing of T-joints through in-situ synchronous assisted arc welding using laser shock forging. This innovative process reduces tensile residual stresses associated with welding by introducing plastic deformation in the deposited weld seam at elevated temperatures. Experimental investigations have shown that the in-situ LSF treatment can significantly reduce tensile residual stresses and improve the mechanical properties of the specimens.
JOURNAL OF MANUFACTURING PROCESSES
(2023)
Article
Mechanics
Vinyas Mahesh
Summary: In this article, a novel finite element method - artificial neural network approach is used to study the coupled static parameters of a smart sandwich plate with agglomerated Carbon Nanotubes porous nanocomposite core and piezo-magneto-thermo-electric facings. The effects of CNTs agglomeration, porosity, and pyro-coupling of the PMTE materials are investigated. Data collected from a Finite Element computational tool are used to train the ANN model. Different agglomeration states and forms of porosity are considered for investigation. Numerical examples are solved to understand the interrelated effects of these material properties on the static parameters.
INTERNATIONAL JOURNAL OF NON-LINEAR MECHANICS
(2023)
Article
Mathematics, Applied
Karthik K. K. Vasudeva, S. M. Mallikarjunaiah, Hyun Chul Yoon
Summary: In this study, a specific finite element model is investigated to study the thermoelastic behavior within the context of a nonlinear strain-limiting constitutive relation. The study reveals that the proposed model exhibits slower growth of strains near crack tips compared to the growth of stresses, which differs from the predictions based on classical linearized description of elastic bodies. This research provides a theoretical and computational framework for developing physically meaningful models and examining coupled multi-physics phenomena, such as the evolution of complex crack networks induced by thermal shocks.
COMMUNICATIONS IN NONLINEAR SCIENCE AND NUMERICAL SIMULATION
(2022)
Article
Chemistry, Physical
Debtanay Das, Swarup Bag, Sukhomay Pal, Abhay Sharma
Summary: Despite the limited research on numerical simulation of friction stir welding (FSW) of dissimilar materials, we conducted a numerical study to predict defects and tool wear. Our model successfully predicted various defects and revealed the importance of plate position in defect formation. By considering wear-induced changes in tool dimensions, the model captured the formation of defects like surface tunnels.
Article
Materials Science, Multidisciplinary
Guangjie Feng, Yifeng Wang, Wenze Luo, Long Hu, Dean Deng
Summary: This study investigated the welding residual stress (WRS) and deformation induced by LVEBW and MAG in SUS310S thick-plate joint through numerical simulation and experiment. Results show significant differences in the distribution of WRS induced by the two methods, with LVEBW exhibiting narrower longitudinal RS region and smaller maximum transverse RS compared to MAG. Additionally, LVEBW demonstrates lower transverse shrinkage and out-of-plane deformation compared to MAG, showcasing a great efficiency advantage.
JOURNAL OF MATERIALS RESEARCH AND TECHNOLOGY-JMR&T
(2021)
Article
Mathematics, Applied
Guo Zheng, Zengqiang Cao, Yuehaoxuan Wang, Reza Talemi
Summary: This study introduces two novel methods for predicting the fatigue response of Dynamic Cold Expansion (DCE) and Static Cold Expansion (SCE) open-hole plates. The accuracy of the prediction is enhanced by considering stress distributions and improving existing methods. The study also discusses the mechanisms behind fatigue life enhancement and fatigue crack propagation modes in cold expansion specimens.
FINITE ELEMENTS IN ANALYSIS AND DESIGN
(2024)
Article
Mathematics, Applied
Eric Heppner, Tomohiro Sasaki, Frank Trommer, Elmar Woschke
Summary: This paper presents a modeling approach for estimating the bonding strength of friction-welded lightweight structures. Through experiments and simulations, a method for evaluating the bonding strength of friction-welded lightweight structures is developed, and the plausibility and applicability of the model are discussed.
FINITE ELEMENTS IN ANALYSIS AND DESIGN
(2024)
Article
Mathematics, Applied
Piermario Vitullo, Alessio Colombo, Nicola Rares Franco, Andrea Manzoni, Paolo Zunino
Summary: Many applications in computational physics involve approximating problems with microstructure, characterized by multiple spatial scales in their data. However, these numerical solutions are often computationally expensive due to the need to capture fine details at small scales. Traditional projection based reduced order models (ROMs) fail to resolve these issues, even for second-order elliptic PDEs commonly found in engineering applications. To address this, we propose an alternative nonintrusive strategy to build a ROM, that combines classical proper orthogonal decomposition (POD) with a suitable neural network (NN) model to account for the small scales.
FINITE ELEMENTS IN ANALYSIS AND DESIGN
(2024)
Article
Mathematics, Applied
Chanh Dinh Vuong, Xiaofei Hu, Tinh Quoc Bui
Summary: In this paper, we present a dynamic description of the smoothing gradient-enhanced damage model for the simulation of quasi-brittle failure localization under time-dependent loading conditions. We introduce two efficient rate-dependent damage laws and various equivalent strain formulations to analyze the complicated stress states and inertia effects of the dynamic regime, enhancing the capability of the adopted approach in modeling dynamic fracture and branching.
FINITE ELEMENTS IN ANALYSIS AND DESIGN
(2024)
Article
Mathematics, Applied
Alexandre D. C. Amaro, A. Francisca Carvalho Alves, F. M. Andrade Pires
Summary: This study focuses on analyzing various deformation mechanisms that affect the behavior of PC/ABS blends using computational homogenization. By establishing a representative microstructural volume element, defining the constitutive description of the material phases, and modeling the interfaces and matrix damage, accurate predictions can be achieved. The findings have important implications for broader applications beyond PC/ABS blends.
FINITE ELEMENTS IN ANALYSIS AND DESIGN
(2024)
Article
Mathematics, Applied
David Hoffmeyer, A. R. Damanpack
Summary: This paper introduces a method for determining all six stress components for a cantilever-type beam that is subjected to concentrated end loads. The method considers an inhomogeneous cross-section and employs cylindrically orthotropic material properties. The efficacy of the method is validated by numerical examples and a benchmark example, and the analysis on a real sawn timber cross-section reveals significant disparities in the maximum stresses compared to conventional engineering approaches.
FINITE ELEMENTS IN ANALYSIS AND DESIGN
(2024)
Article
Mathematics, Applied
Vladimir Stojanovic, Jian Deng, Dunja Milic, Marko D. Petkovic
Summary: The present paper investigates the dynamic analysis of a coupled Timoshenko beam-beam or beam-arch mechanical system with geometric nonlinearities. A modified p-version finite element method is developed for the vibrations of a shear deformable coupled beam system with a discontinuity in an elastic layer. The main contribution of this work is the discovery of coupled effects and phenomena in the simultaneous vibration analysis of varying discontinuity and varying curvature of the newly modelled coupled mechanical system. The analysis results are valuable and have broader applications in the field of solids and structures.
FINITE ELEMENTS IN ANALYSIS AND DESIGN
(2024)
Article
Mathematics, Applied
Gihwan Kim, Phill-Seung Lee
Summary: The phantom-node method is applied in the phase field model for mesh coarsening to improve computational efficiency. By recovering the fine mesh in the crack path domain into a coarse mesh, this method significantly reduces the number of degrees of freedom involved in the computation.
FINITE ELEMENTS IN ANALYSIS AND DESIGN
(2024)
Article
Mathematics, Applied
Souhail Chaouch, Julien Yvonnet
Summary: In this study, an unsupervised machine learning-based clustering approach is developed to reduce the computational cost of nonlinear multiscale methods. The approach clusters macro Gauss points based on their mechanical states, reducing the problem from macro scale to micro scale. A single micro nonlinear Representative Volume Element (RVE) calculation is performed for each cluster, using a linear approximation of the macro stress. Anelastic macro strains are used to handle internal variables. The technique is applied to nonlinear hyperelastic, viscoelastic and elastoplastic composites.
FINITE ELEMENTS IN ANALYSIS AND DESIGN
(2024)
Article
Mathematics, Applied
Hoang-Giang Bui, Jelena Ninic, Christian Koch, Klaus Hackl, Guenther Meschke
Summary: With the increasing demand for underground transport infrastructures, it is crucial to develop methods and tools that efficiently explore design options and minimize risks to the environment. This study proposes a BIM-based approach that connects user-friendly software with effective simulation tools to analyze complex tunnel structures. The results show that modeling efforts and computational time can be significantly reduced while maintaining high accuracy.
FINITE ELEMENTS IN ANALYSIS AND DESIGN
(2024)
Article
Mathematics, Applied
Aslan Nasirov, Xiaoyu Zhang, David Wagner, Saikumar R. Yeratapally, Caglar Oskay
Summary: This manuscript presents an efficient model construction strategy for the eigenstrain homogenization method (EHM) for the reduced order models of the nonlinear response of heterogeneous microstructures. The strategy relies on a parallel, element-by-element, conjugate gradient solver, achieving near linear scaling with respect to the number of degrees of freedom used to resolve the microstructure. The linear scaling in the number of pre-analyses required to construct the reduced order model (ROM) follows from the EHM formulation. The developed framework has been verified using an additively manufactured polycrystalline microstructure of Inconel 625.
FINITE ELEMENTS IN ANALYSIS AND DESIGN
(2024)
Article
Mathematics, Applied
Atticus Beachy, Harok Bae, Jose A. Camberos, Ramana V. Grandhi
Summary: Emulator embedded neural networks leverage multi-fidelity data sources for efficient design exploration of aerospace engineering systems. However, training the ensemble models can be costly and pose computational challenges. This work presents a new type of emulator embedded neural network using the rapid neural network paradigm, which trains near-instantaneously without loss of prediction accuracy.
FINITE ELEMENTS IN ANALYSIS AND DESIGN
(2024)
Article
Mathematics, Applied
Arash Hajisharifi, Michele Girfoglio, Annalisa Quaini, Gianluigi Rozza
Summary: This paper introduces three reduced order models for reducing computational time in atmospheric flow simulation while preserving accuracy. Among them, the PODI method, which uses interpolation with radial basis functions, maintains accuracy at any time interval.
FINITE ELEMENTS IN ANALYSIS AND DESIGN
(2024)
Article
Mathematics, Applied
D. Munoz, S. Torregrosa, O. Allix, F. Chinesta
Summary: The Proper Generalized Decomposition (PGD) is a Model Order Reduction framework used for parametric analysis of physical problems. It allows for offline computation and real-time simulation in various situations. However, its efficiency may decrease when the domain itself is considered as a parameter. Optimal transport techniques have shown exceptional performance in interpolating fields over geometric domains with varying shapes. Therefore, combining these two techniques is a natural choice. PGD handles the parametric solution while the optimal transport-based methodology transports the solution for a family of domains defined by geometric parameters.
FINITE ELEMENTS IN ANALYSIS AND DESIGN
(2024)
Article
Mathematics, Applied
Jothi Mani Thondiraj, Akhshaya Paranikumar, Devesh Tiwari, Daniel Paquet, Pritam Chakraborty
Summary: This study develops a diffused interface CPFEM framework, which reduces computational cost by using biased mesh and provides accurate results using non-conformal elements in the mesh size transiting regions. The accuracy of the framework is confirmed through comparisons with sharp and stepped interface results.
FINITE ELEMENTS IN ANALYSIS AND DESIGN
(2024)