February 2025
·
7 Reads
Computer-Aided Design
This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.
February 2025
·
7 Reads
Computer-Aided Design
January 2025
·
7 Reads
Materials
The structure of thermoset composite laminated plates is made by stacking layers of plies with different fiber orientations. Similarly, the stiffened panel structure is assembled from components with varying ply configurations, resulting in thermal residual stresses and processing-induced deformations (PIDs) during manufacturing. To mitigate the residual stresses caused by the geometric features of corner structures and the mismatch between the stiffener-skin ply orientations, which lead to PIDs in composite-stiffened panels, this study proposes a multi-objective stacking optimization strategy based on an improved adaptive genetic algorithm (IAGA). The viscoelastic constitutive model was employed to describe the modulus variation during the curing process to ensure computational accuracy. In this study, the IAGA was proposed to optimize the ply-stacking sequence of L-shaped stiffeners in composite laminated structures. The results demonstrate a reduction in the spring-in angle to 0.12°, a 50% improvement compared to symmetric balanced stacking designs, while the buckling eigenvalues were improved by 20%. Additionally, the IAGA outperformed the traditional non-dominated sorting genetic algorithm (NSGA), achieving a threefold increase in the Pareto solution diversity under identical constraints and reducing the convergence time by 70%. These findings validate the effectiveness of asymmetric ply design and provide a robust framework for enhancing the structural performance and manufacturability of composite laminates.
November 2024
·
14 Reads
Journal of Composites Science
The process-induced deformation (PID) during the manufacturing of thermosetting composite materials can significantly compromise manufacturing precision. This paper introduces an innovative method that combines a finite element analysis (FEA), feature classification algorithms, and an Artificial Neural Network (ANN) framework to rapidly predict the PID of a typical L-shaped structure. Initially, a comprehensive range of parameters that influence PID are compiled in this research, followed by the generation of a dataset through FEA considering viscoelastic constitutive models, validated by experimental results. Influential parameters are classified using Random Forest and LASSO regression methods, with each parameter rated according to its impact on PID, delineating their varying degrees of importance. Subsequently, through a hyperparameter analysis, an ANN framework is developed to rapidly predict the PID, while also refining the assessment of the parameters’ significance. This innovative approach achieves a computational time reduction of 98% with less than a 5% loss in accuracy, and highlights that under limited computational conditions, considering only a subset or all of the parameters—the peak temperature, corner angle, coefficient of chemical shrinkage, coefficient of thermal expansion, curing pressure, and E1—minimizes accuracy loss. The study demonstrates that machine learning algorithms can effectively address the challenge of predicting composite material PID, providing valuable insights for practical manufacturing applications.
November 2024
·
9 Reads
Advances in Engineering Software
October 2024
·
9 Reads
·
2 Citations
Mesh deformation technology is widely used in aerodynamic applications like unsteady flow, aeroelasticity, and aerodynamic shape optimization because of its low computational costs and consistent mesh connectivity. In order to raise deformed mesh quality and improve efficiency, a new mesh deformation method based on quaternion and displacement normal propagation (named QN method) is introduced in this paper. The boundary points propagate their displacements composed of translational vectors and quaternions to corresponding volume points along the normal direction under the control of the damping function, which preserves the mesh shape and guarantees the quality near boundaries, including orthogonality and normal size. It can also prevent the volume points from being interfered by other less relevant boundary points, so dealing with complex displacement fields effectively. In addition, it avoids complicated matrix and interpolation operations, thus saving lots of computational costs. For mesh with complex topology, a hybrid method combining the QN method and the radial basis function method (RBF method) is investigated to broaden application scenarios, which are applied to the mesh with normal correspondence inside the boundary layer and the mesh outside, respectively. Benefitting from effective handling for the near-wall elements by the QN method, the RBF interpolation part in the hybrid method requires minor support points to carry out valid large deformation, improving the deformation efficiency greatly compared to the individual RBF method. Five typical test cases with different deformation modes and mesh characteristics are implemented, showing better performance of the proposed method in deformed mesh quality and deformation efficiency.
October 2024
·
7 Reads
Composites Science and Technology
April 2024
·
38 Reads
·
11 Citations
Computer Aided Geometric Design
January 2024
·
284 Reads
·
5 Citations
Journal of Computational Design and Engineering
3D reconstruction is a significant research topic in the field of Computer-Aided Design (CAD), which is used to recover editable CAD models from original shapes, including point clouds, voxels, meshes, and boundary representations (B-rep). Recently, there has been considerable research interest in deep model generation due to the increasing potential of deep learning methods. To address the challenges of 3D reconstruction and generation, we propose Brep2Seq, a novel deep neural network designed to transform the B-rep model into a sequence of editable parametrized feature-based modeling operations comprising principal primitives and detailed features. Brep2Seq employs an encoder-decoder architecture based on the Transformer, leveraging geometry and topological information within B-rep models to extract the feature representation of the original 3D shape. Due to its hierarchical network architecture and training strategy, Brep2Seq achieved improved model reconstruction and controllable model generation by distinguishing between the primary shape and detailed features of CAD models. To train Brep2Seq, a large-scale dataset comprising one million CAD designs is established through an automatic geometry synthesis method. Extensive experiments on both DeepCAD and Fusion 360 datasets demonstrate the effectiveness of Brep2Seq, and show its applicability to simple mechanical components in real-world scenarios. We further apply Brep2Seq to various downstream applications, including point cloud reconstruction, model interpolation, shape constraint generation and CAD feature recognition.
January 2024
·
3 Reads
January 2024
·
1 Read
... [1][2][3][4] By allowing refinement in complex regions for enhanced computational accuracy, while maintaining coarser meshes in simpler areas to conserve computational resources, unstructured meshes exhibit remarkable adaptability. 5 They excel in handling intricate geometries, irregular boundaries, internal structures, and local details, guaranteeing the accuracy and reliability of simulation results. Consequently, they provide indispensable data support for optimization design and performance analysis in aerospace, automotive manufacturing, energy, and other related fields. ...
October 2024
... AFR frameworks based on GNN, such as AAGNet and CADNet, face inherent limitations, including restricted receptive fields and network depths due to the foundational structure of their underlying message-passing neural network architecture. To overcome these limitations, Zhang et al. introduced BrepMFR, which leverages recent advances in graph transformers to enhance the neural network's feature extraction capabilities for more complex STEP CAD models [28]. ...
April 2024
Computer Aided Geometric Design
... In reverse engineering, a prevalent CAD modeling approach employs the sketch-extrude [Uy et al. 2022;] paradigm, where a sketch [Para et al. 2021;Xu et al. 2022] is defined as a series of parametric curves forming a closed loop, resulting in flexible and complex primitives, which is the focus of this work. Some recent works have borrowed ideas from NLP to interpret CAD models as sequences of tokens [Ganin et al. 2021;Wu et al. 2021;Zhang et al. 2024]. Point2Cyl [Uy et al. 2022] takes a different approach and redefines the CAD reverse engineering task as a geometry-aware decomposition problem. ...
January 2024
Journal of Computational Design and Engineering
... As shown in Table IV, the QN method takes only 153.16 s due to its simplicity and directness, while the RBF method involves complex operations such as selecting control points, constructing large matrices, and solving large linear equations. Even with acceleration measures such as parallel point selection strategies of 96 and the incremental LDLT (ILDLT) method, 53 it still takes 1918.63 s for such a large mesh scale. It can be conducted that the QN method has superior efficiency in solving large-scale mesh deformation problems. ...
July 2023
Aerospace Science and Technology
... TT fp , and sl fp is the fracture angle under pure transverse tension, compression, and in-plane shear [34], respectively. The potential fracture angle refers to the angle between the crack and the corresponding axis, as seen in the literature [32]. ...
December 2022
Composite Structures
... They conducted fatigue experiments at two load levels to investigate the initiation, and propagation of delamination. They discussed the fatigue shear behavior, and damage mechanisms based on experimental results [11]. Zhou, and Gao investigated the effect of adhesive interface properties on the post-buckling response of composite I-profile reinforced panels with lateral beam support under axial pressure [12]. ...
December 2022
Thin-Walled Structures
... The out-of-plane load-bearing capacity of hexagonal structure has been studied [57,58]. However, when the load is off axis, it often causes the honeycomb structure to tilt, eventually leading to catastrophic failure. ...
August 2022
... Such shear instability occurs when the shear stress reaches the shear yielding strength of the matrix and the shear damage mode transforms into the fibre kinking mode, contributing to the kink-band formation [11]. Other studies indicated that with a larger waviness angle, the fibre kinking is more likely to occur in composite materials, which contributes to the reduction of compressive failure strength drastically [14,15]. In addition, void formation in the thermoset composites is inevitable during the curing process of composites due to the infiltration between resin and fibre, the generation of volatile gas, curing pressure and other factors [16,17]. ...
July 2022
Composites Part A Applied Science and Manufacturing
... At the same time, Lu-ana Pereira Dalla Mariga et al. investigated the impact of geometrical tolerance on single-lap composite joints riveted with two fasteners [30,31]. Countersunk bolts have been investigated for the correction of drilling-induced faults, as well as the influence of stacking sequences on the bearing strength of composite bolted joints [32][33][34]. Manasi Palwankar et al. demonstrated a revised fixture for bolted composite joint testing that can support countersunk fasteners [35]. Studies have also looked at the environmental impact of bolted joints, corrosion-resistant alternatives, bearing performance in harsh conditions, and the effects of thermal exposure and seawater aging on bearing strength [36][37][38][39][40][41][42]. ...
July 2022
Materials
... Although it is still a linear contact model which saves run-time, modification of the geometry may be required to include the pressure cone surface, which does not exist in the first place. Another approach is to model the bolted connections as a frictional contact pair [38][39][40], but as it is a non-linear contact case, the computational load will increase. Another approach is to model the bolts and nuts as beams with pre-tension load [41,42]. ...
October 2021
Composite Structures