A. Khosravifard¹, M.R. Hematiyan^1,2

CMES-Computer Modeling in Engineering & Sciences, Vol.78, No.3&4, pp. 185-208, 2011, DOI:10.3970/cmes.2011.078.185

Abstract An inverse method for optimal control of the freezing front motion in the solidification of pure materials is presented. The inverse technique utilizes the idea of a pseudo heat source to account for the latent heat effects. The numerical formulation of this inverse method is based on a formerly introduced meshless technique. In this method, the flux and the velocity of the liquid-solid interface are treated as secondary variables and the liquid and solid domains are modeled simultaneously. Some numerical examples are provided to demonstrate the efficiency of the presented method. The effects of regularization and the number of nodes… More >

Eduardo T Lima Junior¹, Wilson S Venturini², Ahmed Benallal³

CMES-Computer Modeling in Engineering & Sciences, Vol.78, No.3&4, pp. 169-184, 2011, DOI:10.3970/cmes.2011.078.169

Abstract Some considerations on the numerical analysis of brittle rocks are presented in this paper. The rock is taken as a poro-elastic domain, in full-saturated condition, based on the Biot's Theory. The solid matrix of this porous medium is considered to be susceptible to isotropic damage occurrence. An implicit boundary element method (BEM) formulation, based on time-independent fundamental solutions, is developed and implemented to couple the fluid flow and two-dimensional elastostatics problems. The integration over boundary elements is evaluated by using a numerical Gauss procedure. A semi-analytical scheme for the case of triangular domain cells is followed to carry out the… More >

C.T.M. Anflor¹, R.J. Marczak²

CMES-Computer Modeling in Engineering & Sciences, Vol.78, No.3&4, pp. 151-168, 2011, DOI:10.3970/cmes.2011.078.151

Abstract This paper aims to demonstrate the final result of an optimization process when a smooth technique is introduced between intermediary iterations of a topological optimization. In a topological optimization process is usual irregular boundary results as the final shape. This boundary irregularity occurs when the way of the material is removed is not very suitable. Avoiding an optimization post-processing procedure some techniques of smooth are implemented in the original optimization code. In order to attain a regular boundary a smoothness technique is employed, which is, Bezier curves. An algorithm was also developed to detect during the optimization process which curve… More >

N.A. Dumont¹, E.Y. Mamani¹

CMES-Computer Modeling in Engineering & Sciences, Vol.78, No.2, pp. 109-150, 2011, DOI:10.3970/cmes.2011.078.109

Abstract A particular implementation of the hybrid boundary element method is presented for the two-dimensional analysis of potential and elasticity problems, which, although general in concept, is suited for fracture mechanics applications. Generalized Westergaard stress functions, as proposed by Tada, Ernst and Paris in 1993, are used as the problem's fundamental solution. The proposed formulation leads to displacement-based concepts that resemble those presented by Crouch and Starfield, although in a variational framework that leads to matrix equations with clear mechanical meanings. Problems of general topology, such as in the case of unbounded and multiply-connected domains, may be modeled. The formulation, which… More >

Y.C. Shiah¹, C. L. Tan²

CMES-Computer Modeling in Engineering & Sciences, Vol.78, No.2, pp. 95-108, 2011, DOI:10.3970/cmes.2011.078.095

Abstract By differentiating the Green function of Ting and Lee (1997) for 3D general anisotropic elastotatics in a spherical coordinate system as an intermediate step, and then using the chain rule, derivatives of up to the second order of this fundamental solution are obtained in exact, explicit, algebraic forms. No tensors of order higher than two are present in these derivatives, thereby allowing these quantities to be numerically evaluated quite expeditiously. These derivatives are required for the computation of the internal point displacements and stresses via Somigliana's identity in BEM analysis. Some examples are presented to demonstrate their successful implementation to… More >

T. Matsumoto¹, T. Yamada¹, T. Takahashi¹, C.J. Zheng², S. Harada¹

CMES-Computer Modeling in Engineering & Sciences, Vol.78, No.2, pp. 77-94, 2011, DOI:10.3970/cmes.2011.078.077

Abstract Shape design and topology sensitivity formulations for acoustic problems based on adjoint method and the boundary element method are presented and are applied to shape sensitivity analysis and topology optimization of acoustic field. The objective function is assumed to consist only of boundary integrals and quantities defined at certain number of discrete points. The adjoint field is defined so that the sensitivity of the objective function does not include the unknown sensitivity coefficients of the sound pressures and particle velocities on the boundary and in the domain. Since the final sensitivity expression does not have the sensitivity coefficients of the… More >

P. Maghoul¹, B. Gatmiri^1,2, D. Duhamel¹

CMES-Computer Modeling in Engineering & Sciences, Vol.78, No.1, pp. 51-76, 2011, DOI:10.3970/cmes.2011.078.051

Abstract This paper aims at obtaining boundary integral formulations as well as three dimensional(3D) fundamental solutions for unsaturated soils under dynamic loadings for the first time. The boundary integral equations are derived via the use of the weighted residuals method in a way that permits an easy discretization and implementation in a Boundary Element code. Also, the associated 3D fundamental solutions for such deformable porous medium are derived in Laplace transform domain using the method of Hérmander. The derived results are verified analytically by comparison with the previously introduced corresponding fundamental solutions in elastodynamic limiting case. These solutions can be used,… More >

A. Sellier, N. Ghalia

CMES-Computer Modeling in Engineering & Sciences, Vol.78, No.1, pp. 25-50, 2011, DOI:10.3970/cmes.2011.078.025

Abstract The Green tensor complying with anisotropic slip conditions at the surface of a plane, impermeable, motionless and slipping wall is theoretically obtained and an efficient numerical method is proposed to accurately compute at a very reasonable cpu time cost each of its Cartesian component. The accuracy of the advocated numerical strategy is tested against the Maple Software and the employed procedure makes it possible to calculate the Green tensor for a non-isotropic slip condition at a cpu time cost comparable with the one needed for the less complicated isotropic Navier condition. More >

A. Brancati¹, M. H. Aliabadi¹, A. Milazzo²

CMES-Computer Modeling in Engineering & Sciences, Vol.78, No.1, pp. 1-24, 2011, DOI:10.3970/cmes.2011.078.001

Abstract This paper presents an improved adaptive cross approximation (ACA) approach developed in conjunction with the Hierarchical format matrix and the GMRES solver. A novel scheme to generate the cluster tree (based upon preliminary considerations of the prescribed boundary conditions) and an improved ACA algorithm (approximating the system matrix for mixed Robin conditions) are described. The asymptotic smoothness property of a kernel generated by a linear combination of two asymptotic smooth kernels is demonstrated. Numerical results show the new approach to be up to 50% faster than the conventional ACA approach. More >

H.F.Qiang¹, F.Z.Chen¹, W.R. Gao¹

CMES-Computer Modeling in Engineering & Sciences, Vol.77, No.3&4, pp. 239-262, 2011, DOI:10.3970/cmes.2011.077.239

Abstract Based on smoothed particle hydrodynamics (SPH) method with surface tension proposed by Morris, this paper is intended to modify equations for surface tension by modifying normal and curvature with corrective smoothing particle method (CSPM). Compared with the continuum surface force (CSF) model for surface tension employed in the traditional SPH method, the accuracy in the present paper is much higher in terms of handling the problems with large deformation and surface tension. The reason is that in the traditional SPH method the deficiency of particles is near the boundary and sharp-angled areas, and it causes gross errors of curvature calculation.… More >