摘要:

AbstractA new general purpose boundary element method for domains with cracks has been recently developed. This technique avoids the use of a multi‐domain decomposition by including an additional integral equation expressing the boundary condition on the crack. The principal requirement of this technique is the analytic determination of certain hypersingular integrals of the Green's function which arise from this equation. In order to establish the applicability of this method for fracture, these integrals are evaluated herein for the Kelvin solution of the three‐dimensional Navier equations of linear elasticity. Numerical results for fracture problems using the single‐domain boundary element analysis are also pres

ISSN:0029-5981    

DOI:10.1002/nme.1620290603

出版商:John Wiley&Sons, Ltd    

年代:1990

数据来源: WILEY

摘要:

AbstractThe treatment of ageing processes in polymers by the finite element (FE) method is described. The development of this analytical method is motivated by the need to assess the effects of ageing on the structural performance of polymers and polymeric composites, particularly when the structures are large and expensive to replace. Also, there is a need to assess the effectiveness of corrective action options when ageing problems do occur, which can be treated by this method. The method describes the treatment of multiple, reactive chemical species in multi‐layered polymeric materials by the finite element method. Example problems featuring the simultaneous diffusion and chemical reaction are illustrated: a simple problem of binary diffusion and reaction with comparison of numerical to exact results, and the staged ageing/structural analysis of a solid rocket moto

ISSN:0029-5981    

DOI:10.1002/nme.1620290604

出版商:John Wiley&Sons, Ltd    

年代:1990

数据来源: WILEY

摘要:

AbstractAn embedding method is considered for the numerical determination of the cathode shape for a given anode shape in a two‐dimensional electrolytic cell. Reference is made to an analytic solution to illustrate the difficulties which have to be overcome by any numerical method applied to this ill‐posed problem. The embedding method searches amongst the direct solutions (i.e. anodes for given cathodes) since these, resulting from a well‐posed problem, are less troublesome to compute. For anodes of the form βG(x), x the transverse direction, the embedding method gives good results for β less than 30–60 percent of the β for which the cathode first contains cusps or becomes triple‐valued. It is shown that the embedding method performs better than some other numerical methods in overcoming the error growth difficulties associated with the illposedness of

ISSN:0029-5981    

DOI:10.1002/nme.1620290605

出版商:John Wiley&Sons, Ltd    

年代:1990

数据来源: WILEY

摘要:

AbstractA general formulation is presented for the tangent constitutive matrices of rate‐independent material models which assume that the strain increment consists of two components. One component is related to the stress increment through a known constitutive matrix, while the other component results from deformation mechanisms not reflected in the constitutive matrix. The procedure to modify the known matrix for these additional strains is specified. The formulation is a generalization of the well‐known procedure used in plasticity to modify the elastic constitutive matrix for plastic deformation. Application to both the classical and endochronic theories of plasticity and to a smeared crack theory is illustra

ISSN:0029-5981    

DOI:10.1002/nme.1620290606

出版商:John Wiley&Sons, Ltd    

年代:1990

数据来源: WILEY

摘要:

AbstractThe dispersive and spurious oscillation properties of one‐dimensional uniform finite element meshes are investigated using the information provided by the frequency response functions of the finite element meshes. The modal superposition method, in which the effects of temporal discretization are eliminated, is employed to integrate the discretized equations of motion. Compared with the commonly used complex variable method, this approach reveals more clearly the physical essence of the dispersion and spurious oscillation phenomena. A series of interesting results is obtaine

ISSN:0029-5981    

DOI:10.1002/nme.1620290607

出版商:John Wiley&Sons, Ltd    

年代:1990

数据来源: WILEY

摘要:

AbstractThis study develops and illustrates a dynamic model for vibration analysis of rod and beam elements with arbitrary mass distributions. This arbitrary mass model represents the element shape function using a frequency dependent series that becomes exact as the number of terms in the series approaches infinity. The arbitrary mass model yields the natural frequencies and modes of elements with mass distributions which vary as continuous functions of the spanwise co‐ordinate.The arbitrary mass model, implemented using the unsymmetric formulation for vibration analysis, is illustrated in examples and compared with other dynamic models which use stepped mass representations to model the arbitrary mass distributions. Studies show that the arbitrary mass model is more computationally efficient than the stepped mass representation models, particularly when accuracy requirements are hig

ISSN:0029-5981    

DOI:10.1002/nme.1620290608

出版商:John Wiley&Sons, Ltd    

年代:1990

数据来源: WILEY

摘要:

AbstractAnL2Galerkin approximation is obtained by the use of specially chosen spaces of trial and test functions. The method combines the features of the orthogonal collocation and Galerkin finite element methods. This new algorithm can be more efficient by orders of magnitude than conventional discontinuous Galerkin finite element procedures in solving non‐linear problems which require frequent reformulation of the coefficient matrices and right‐hand‐side ve

ISSN:0029-5981    

DOI:10.1002/nme.1620290609

出版商:John Wiley&Sons, Ltd    

年代:1990

数据来源: WILEY

摘要:

AbstractIn this paper a new approach to the development ofa posteriorierror indicators for the direct boundary element method (BEM) is presented. It takes advantage of a unique opportunity offered by the collocation BEM, that is the possibility of obtaining more than one numerical solution from a given boundary element mesh. Unlike other approaches, the error indicator introduced in the present paper does not require any explicit evaluation of the residual of the approximate solution. In order to test the capability of the proposed error indicators to actually control a selective mesh refinement process, anh‐adaptive refinement strategy is also described, together with the criteria for refinement and for the termination of the adaptive process. Full details on the whole numerical implementation for two‐dimensional problems are provided. In particular, there is described a direct method for the numerical evaluation of Cauchy principal value integrals. Numerical examples show that pointwise convergence is always obtained in a few steps and the final meshes look nearly as if they were designed by an expert anal

ISSN:0029-5981    

DOI:10.1002/nme.1620290610

出版商:John Wiley&Sons, Ltd    

年代:1990

数据来源: WILEY

摘要:

AbstractAn effective method is presented for generating three‐dimensional boundary‐fitted grids. The technique is based on a two‐step procedure involving the solution of differential equations and algebraic interpolations. A coarse grid with an arbitrarily given point distribution at all the boundaries is first generated by solving the Poisson equations, with the forcing functions being automatically adjusted during the grid generation to achieve orthogonality at the boundaries. A fine grid with a desired concentration of grid lines is then obtained by fitting tricubic spline functions on the coarse grid. A primary feature of the method is that the mesh spacing can be quickly and explicitly changed while retaining the smoothness over the entire field and the boundary orthogonality. Applications to a variety of geometries demonstrate the performance of the m

ISSN:0029-5981    

DOI:10.1002/nme.1620290611

出版商:John Wiley&Sons, Ltd    

年代:1990

数据来源: WILEY

摘要:

AbstractAn effective and efficient error estimation scheme for finite element plate and shell analysis is presented. The error estimate method proposed is applicable to all the plate and shell elements developed using either classical plate and shell or continuum mechanics theories. The global error in energy norm is computed by summing all the elemental errors. The method is well suited for adaptive mesh refinement as the error contribution for each element is known. Demonstration examples using uniform and adaptive refinement schemes for both the plate and shell analysis show the effectiveness of the method proposed.

ISSN:0029-5981    

DOI:10.1002/nme.1620290612

出版商:John Wiley&Sons, Ltd    

年代:1990

数据来源: WILEY