Zhiguo Yong¹, Hongmei Kang¹, Falai Chen^2,*

CMES-Computer Modeling in Engineering & Sciences, Vol.138, No.1, pp. 739-760, 2024, DOI:10.32604/cmes.2023.027171 - 22 September 2023

Abstract PHT-splines are defined as polynomial splines over hierarchical T-meshes with very efficient local refinement properties. The original PHT-spline basis functions constructed by the truncation mechanism have a decay phenomenon, resulting in numerical instability. The non-decay basis functions are constructed as the B-splines that are defined on the 2 × 2 tensor product meshes associated with basis vertices in Kang et al., but at the cost of losing the partition of unity. In the field of finite element analysis and topology optimization, forming the partition of unity is the default ingredient for constructing basis functions of… More > Graphic Abstract

Abdelkarim El Kahoui¹, Mustapha Malek¹, Nouh Izem¹, M. Shadi Mohamed², Mohammed Seaid^3,4,*

CMES-Computer Modeling in Engineering & Sciences, Vol.124, No.1, pp. 61-78, 2020, DOI:10.32604/cmes.2020.010874 - 19 June 2020

Abstract We propose a high-order enriched partition of unity finite element method for linear and nonlinear time-dependent diffusion problems. The solution of this class of problems often exhibits non-smooth features such as steep gradients and boundary layers which can be very challenging to recover using the conventional low-order finite element methods. A class of steady-state exponential functions has been widely used for enrichment and its performance to numerically solve these challenges has been demonstrated. However, these enrichment functions have been used only in context of the standard h-version refinement or the so-called q-version refinement. In this paper… More >

Gang Xu^{1, *}, Ran Ling¹, Lishan Deng¹, Qing Wu¹, Weiyin Ma²

CMC-Computers, Materials & Continua, Vol.62, No.1, pp. 309-319, 2020, DOI:10.32604/cmc.2020.06509

Abstract In this paper, we propose a novel image interpolation method by using Gaussian-Sinc automatic interpolators with partition of unity property. A comprehensive comparison is made with classical image interpolation methods, such as the bicubic interpolation, Lanczos interpolation, cubic Schaum interpolation, cubic B-spline interpolation and cubic Moms interpolation. The experimental results show the effectiveness of the improved image interpolation method via some image quality metrics such as PSNR and SSIM. More >

D. Ngo-Cong^1,2, N. Mai-Duy¹, W. Karunasena², T. Tran-Cong^1,3

CMES-Computer Modeling in Engineering & Sciences, Vol.83, No.3, pp. 275-310, 2012, DOI:10.3970/cmes.2012.083.275

Abstract In this paper, natural convection flows in concentric and eccentric annuli are studied using a new numerical method, namely local moving least square - one dimensional integrated radial basis function networks (LMLS-1D-IRBFN). The partition of unity method is used to incorporate the moving least square (MLS) and one dimensional-integrated radial basis function (1D-IRBFN) techniques in an approach that leads to sparse system matrices and offers a high level of accuracy as in the case of 1D-IRBFN method. The present method possesses a Kronecker-Delta function property which helps impose the essential boundary condition in an exact More >

K. Wang¹, S.J. Zhou^2,3, Z.F. Nie⁴

CMES-Computer Modeling in Engineering & Sciences, Vol.73, No.1, pp. 77-102, 2011, DOI:10.3970/cmes.2011.073.077

Abstract The natural neighbour Galerkin method is tailored to solve boundary value problems of the couple-stress elasticity to model the size dependent behaviour of materials. This method is based on the displacement-based Galerkin approach, and the calculation of the global stiffness matrix is performed using gradient smoothing technique combined with the non-Sibsonian partition of unity approximation scheme. This method possesses the following properties: the complex C1-continuous approximation scheme is avoided without using either Lagrange multipliers or penalty parameters; no domain integrals involved in the assembly of the global stiffness matrix; and the imposition of essential boundary More >

Phong B.H. Le¹, Timon Rabczuk², Nam Mai-Duy¹, Thanh Tran-Cong¹

CMES-Computer Modeling in Engineering & Sciences, Vol.66, No.1, pp. 25-52, 2010, DOI:10.3970/cmes.2010.066.025

Abstract A novel meshless method based on Radial Basis Function networks (RBFN) and variational principle (global weak form) is presented in this paper. In this method, the global integrated RBFN is localized and coupled with the moving least square method via the partition of unity concept. As a result, the system matrix is symmetric, sparse and banded. The trial and test functions satisfy the Kronecker-delta property, i.e. Φ_i(x_j) = δ_ij. Therefore, the essential boundary conditions are imposed in strong form as in the FEMs. Moreover, the proposed method is applicable to scattered nodes and arbitrary domains. The More >

Raju Sethuraman¹, N.R.Rajesh²

CMES-Computer Modeling in Engineering & Sciences, Vol.39, No.1, pp. 67-100, 2009, DOI:10.3970/cmes.2009.039.067

Abstract This paper presents a methodology based on Partition of Unity Finite Element Method (PUFEM) and Pseudo Elastic Analysis for solving material non-linear fracture problems within the scope of total deformation theory of plasticity. Local enrichment base functions are used to represent the asymptotic field near the crack tip and discontinuous field across the crack faces. An iterative linear elastic analysis using PUFEM is carried out for the determination of elastic-plastic crack tip stress fields by treating effective material properties as spatial field variables. The effective material parameters are defined using deformation theory and are updated… More >

Y.C. Cai¹ and H.H. Zhu¹

CMES-Computer Modeling in Engineering & Sciences, Vol.34, No.2, pp. 179-204, 2008, DOI:10.3970/cmes.2008.034.179

Abstract The popular Shepard PU approximations are easy to construct and have many advantages, but they have several limitations, such as the difficulties in handling essential boundary conditions and the known problem of linear dependence regarding PU-based methods, and they are not the good choice for MLPG method. With the objective of alleviating the drawbacks of Shepared PU approximations, a new meshless PU-based Shepard and Least Square (SLS) interpolation is employed here to develop a new type of MLPG method, which is named as Local Meshless Shepard and Least Square (LMSLS) method. The SLS interpolation possesses… More >