Source Code and Data for "A Parametric Level Set Method for Dynamic Optimization of Laminates with Thickness Tapering"

This dataset contains the complete source code and core computational results used in the research "A Parametric Level Set Method for Dynamic Optimization of Laminates with Thickness Tapering." The dataset is generated primarily through a structural dynamics calculation and optimization analysis process based on the Finite Element Method (FEM). Methodologically, this research employs the Parametric Level Set Method (PLSM) to conduct concurrent optimization design of the topological configuration and thickness tapering features of composite laminates. The optimization iteration process utilizes the Method of Moving Asymptotes (MMA). In each iteration step, a finite element solver calculates the dynamic response and sensitivities of the structure, updating the design variables accordingly. All data is generated via numerical simulation computation, involving no physical measurements. The primary environment for development, calculation, and post-processing is MATLAB.The files in the dataset mainly consist of two categories. The first category includes MATLAB source code files (.m format), covering the main control program, finite element solving modules, sensitivity analysis, level set filtering, and post-processing plotting scripts. For example, A_cantilever_beam_Thickness_Opt_Level_set_filter_Dynamic_MMA.m is the core execution program for the dynamic optimization of thickness and topology targeting a cantilever beam structure. The second category comprises workspace data files (.mat format) exported during the calculation process, which are used to comprehensively record the optimization iteration history, objective function convergence status, structural mass, natural frequencies, and final structural morphological feature data. For instance, the file L型梁[0_10_50].mat records the key matrices and intermediate variables of an L-shaped beam under specific boundary conditions or parameter settings. All relevant spatial and physical information, such as model geometric dimensions, material properties, and spatial meshing, are explicitly defined as parameters within the corresponding initialization .m files.Regarding data integrity and error descriptions, since this dataset is purely the product of computer numerical simulations, there is no missing data typically found in traditional observational experiments. In terms of calculation accuracy, constrained by the resolution of the finite element mesh (e.g., element size division) and computer floating-point truncation errors, conventional numerical errors may exist at boundaries or in high-frequency dynamic response calculations, which are inherent characteristics of numerical analysis methods. For data usage and reproduction, users only need a local installation of the MATLAB environment to run the code. If the main program calls third-party plain-text subroutines, such as the classic MMA optimizer, users simply need to ensure that the relevant dependency files are located in the same running directory. By directly running the corresponding main program scripts, users can reproduce the entire structural optimization process presented in the paper and automatically generate the corresponding post-processing contour plots.