摘要:

AbstractThis paper describes an optimization procedure for the minimum weight design of complex structures. The procedure is based on a new cubic extended interior penalty function (CEIPF) used with the sequence of unconstrained minimization technique (SUMT) and Newton's method. The Hessian matrix of the penalty function is approximated using only constraints and their derivatives. The CEIPF is designed to minimize the error in the approximation of the Hessian matrix, and as a result the number of structural analyses required is small and independent of the number of design variables. Three example problems are reported. The number of structural analyses is reduced by as much as 50 per cent below previously reported results.

ISSN:0029-5981    

DOI:10.1002/nme.1620140802

出版商:John Wiley&Sons, Ltd    

年代:1979

数据来源: WILEY

摘要:

AbstractIn Reference 1 a class of first‐order factorization methods for the solution of large, sparse systems of equations, of which the coefficient matrix is a symmetricM‐matrix, was introduced. Asymptotic results for the computational complexity was proved in the case of systems arising from finite difference approximation of second‐order self‐adjoint elliptic partial differential equation problems in two dimensions with Dirichlet boundary conditions. In this paper the result is extended to cover even mixed boundary conditions and problems with discontinuous material coefficients. Results from numerical experiments with various finite difference and finite element approximations are presented and comparisons with direct and other interative methods a

ISSN:0029-5981    

DOI:10.1002/nme.1620140803

出版商:John Wiley&Sons, Ltd    

年代:1979

数据来源: WILEY

摘要:

AbstractThis paper describes an explicit numerical method for solving transient combined heat conduction and convection problems. Applications include the cooling of many types of engineering plant; for example, stator and rotor windings of turbogenerators and high voltage underground cables. The Du Fort–Frankel and the fully implicit finite difference schemes have been used to solve the conduction and convection equations, respectively. It is shown that, with a suitable order of calculation, the overall method becomes explicit. Computational procedures are outlined and stability, accuracy and convergence are considered. Numerical examples are given to illustrate the use of the method and to validate some of the theoretical points. Results have also been obtained using existing numerical methods and have been compared with those from the proposed method. For certain problems, it is shown that the proposed method uses less overall computing time than other methods such as that devised by Crank and Nicolso

ISSN:0029-5981    

DOI:10.1002/nme.1620140804

出版商:John Wiley&Sons, Ltd    

年代:1979

数据来源: WILEY

摘要:

AbstractThis paper describes a numerical procedure for the solution of metalworking problems of a strain‐hardening material. The overall metalworking process is split up into a number of stages and the governing equations for a typical stage are set up in terms of incremental displacements. The incremental deformation occurring in each successive stage is found by an interative numerical procedure based on the concept of piecewise linearization. The ‘composite’ solution yields the distribution of stress and strain in the work material at various stages of the process. Finally, application of the method to finite axisymmetric compression is desc

ISSN:0029-5981    

DOI:10.1002/nme.1620140805

出版商:John Wiley&Sons, Ltd    

年代:1979

数据来源: WILEY

ISSN:0029-5981    

DOI:10.1002/nme.1620140806

出版商:John Wiley&Sons, Ltd    

年代:1979

数据来源: WILEY

ISSN:0029-5981    

DOI:10.1002/nme.1620140807

出版商:John Wiley&Sons, Ltd    

年代:1979

数据来源: WILEY

ISSN:0029-5981    

DOI:10.1002/nme.1620140808

出版商:John Wiley&Sons, Ltd    

年代:1979

数据来源: WILEY

摘要:

AbstractA new dynamic substructuring method is presented. The substructures are identified by their fixed‐interface modes and the condensed stiffness and mass matrices at zero vibration frequency. Physical co‐ordinates are used instead of the modal co‐ordinates to make the condition of compatibility between elements applicable. The dimension of the fundamental matrices is equal to the number of interface co‐ordinates for each substructure. All information needed to formulate the fundamental matrices may be obtained experimentally. Both discrete and continuous parameter models are considered.A number of numerical examples are given and the method is compared with the other methods of dynamic substructure. It is shown that the numerical accuracy lost in the ordinary eigenvalue economization process may be recovered by the present

ISSN:0029-5981    

DOI:10.1002/nme.1620140809

出版商:John Wiley&Sons, Ltd    

年代:1979

数据来源: WILEY

摘要:

AbstractThis paper presents and results obtained from the stabilitystudies of plane Poiseuille flow and magnetohydrodynamic flow by the finite element method. Applying Galerkin's weighted residual method and introducing the interpolation function in the exponetial form with respect to time, the governing flow equations are reduced to an eigenvalue problem. This formulation is much simpler than that of the asymptotic expansion method. Solutions are obtained directly in terms of velocity and pressure. Results of the critical Reynolds number obtained by this method compare well with those of other methods for plane Poiseuille flow and magnetohydrodynamic flow.

ISSN:0029-5981    

DOI:10.1002/nme.1620140810

出版商:John Wiley&Sons, Ltd    

年代:1979

数据来源: WILEY

ISSN:0029-5981    

DOI:10.1002/nme.1620140811

出版商:John Wiley&Sons, Ltd    

年代:1979

数据来源: WILEY