1993 Vol. 10 No. 4
1993, 10(4): 1-8.
Abstract：
A triangular flat shell element GST18 with three nodes and eighteen degrees of freedom is constructed. The membrane and bendingparts of the shell eloment are the generalized triangular membrane element with drilling freedoms GT9 and the generalized conforming plate bending element GPL-T9, respectively. By using the generalized conformingmethod the convergence of the element is guaranteed. By introducingthe drilling freedoms the difficulty that the stiffness matrix is singularwhen the neighbouring elements around a common node are coplanar is overcome. The results of numerical tests demonstrate that the present element exhibits excellent performance.
A triangular flat shell element GST18 with three nodes and eighteen degrees of freedom is constructed. The membrane and bendingparts of the shell eloment are the generalized triangular membrane element with drilling freedoms GT9 and the generalized conforming plate bending element GPL-T9, respectively. By using the generalized conformingmethod the convergence of the element is guaranteed. By introducingthe drilling freedoms the difficulty that the stiffness matrix is singularwhen the neighbouring elements around a common node are coplanar is overcome. The results of numerical tests demonstrate that the present element exhibits excellent performance.
Abstract：
As the final paper in this series of nonlinear application of the finite element method of lines (PEMOL), the present paper applies this method to material nonlinear problems by presenting a FEMOL analysis of the elastic-plastic torsion problem. Firstly, the elastic region of the cross-section is discretized by FEMOL elements. Next the changes of nodal lines at end-points on the elatic-plastic interface are taken as shape variables which are incorporated into the ODE system by using the trivial ODE technique with the supplementary BCs provided by the yield condition of a weakform. As a result, the elastic-plastic torsion problem is transformed into a standard nonlinear ODE problem and then solved by standard ODE solvers. Numerical examples are given to show the good performance of the present approach.
As the final paper in this series of nonlinear application of the finite element method of lines (PEMOL), the present paper applies this method to material nonlinear problems by presenting a FEMOL analysis of the elastic-plastic torsion problem. Firstly, the elastic region of the cross-section is discretized by FEMOL elements. Next the changes of nodal lines at end-points on the elatic-plastic interface are taken as shape variables which are incorporated into the ODE system by using the trivial ODE technique with the supplementary BCs provided by the yield condition of a weakform. As a result, the elastic-plastic torsion problem is transformed into a standard nonlinear ODE problem and then solved by standard ODE solvers. Numerical examples are given to show the good performance of the present approach.
Abstract：
In this paper, some improvements for the trial function of two and three dimentional problems, a sparse algorithm by block based on the subregions for solving the equation system of residual value and a basic frame of general purpose Program of weighted residual method are given.The results of numerical examples show that its precision for using less located points is excellent.
In this paper, some improvements for the trial function of two and three dimentional problems, a sparse algorithm by block based on the subregions for solving the equation system of residual value and a basic frame of general purpose Program of weighted residual method are given.The results of numerical examples show that its precision for using less located points is excellent.
1993, 10(4): 26-31.
Abstract：
In this paper, the experimental surface roughness of pavementis regarded as the random input and the vehicle is simplified as a five degree-freedom system. The interaction force of Man-Vehicle-Road system is investigated by means of random vibration. The response spectrumand standard deviation of the force is calculated with different cases of pavement roughness. In the end of the paper, the compared results between theory and experiment show that the values of calculation and experiment are very similar.
In this paper, the experimental surface roughness of pavementis regarded as the random input and the vehicle is simplified as a five degree-freedom system. The interaction force of Man-Vehicle-Road system is investigated by means of random vibration. The response spectrumand standard deviation of the force is calculated with different cases of pavement roughness. In the end of the paper, the compared results between theory and experiment show that the values of calculation and experiment are very similar.
1993, 10(4): 32-41.
Abstract：
Experimental research and nonlinear theoretical analysis of the ultimate strength of steel frames are presented in this paper. 4 steel frames in which 2 are two-story onebay frame and other 2 aretwo-story two-bay frame were tested. A suitable nonlinear analysis method, which is based on nonlinear continuum mechanics, stress-resultant yield surface and displacement increment Newton-Raphson subincrementation scheme, is successfully applied to the large deformation elasto-plastic analysis of the ultimate strength of steel frames.The comparison of the theoretical predictions with the experimental results tested by the author verifies that the theoretical method derived in this paper has very high precision and efficiency.
Experimental research and nonlinear theoretical analysis of the ultimate strength of steel frames are presented in this paper. 4 steel frames in which 2 are two-story onebay frame and other 2 aretwo-story two-bay frame were tested. A suitable nonlinear analysis method, which is based on nonlinear continuum mechanics, stress-resultant yield surface and displacement increment Newton-Raphson subincrementation scheme, is successfully applied to the large deformation elasto-plastic analysis of the ultimate strength of steel frames.The comparison of the theoretical predictions with the experimental results tested by the author verifies that the theoretical method derived in this paper has very high precision and efficiency.
1993, 10(4): 42-51.
Abstract：
A method for the analysis of steel frames considering the geometrical and material nonlinearities is presented. Large deformation of members including large rotation, large translation and large strain, the thrust effect on tha flexural stiffness are considered. Equilibrium and compatibility equations based on deformed configuration of structuresare derived for the analysis model, the geometric meaning of the nonlinear terms in the tangent stiffness matrix is discussed. A computer program for the whole range analysis of steel frames considering the geometrical and material nonlinearities is implemented, the load carrying capacity can also be determined.
A method for the analysis of steel frames considering the geometrical and material nonlinearities is presented. Large deformation of members including large rotation, large translation and large strain, the thrust effect on tha flexural stiffness are considered. Equilibrium and compatibility equations based on deformed configuration of structuresare derived for the analysis model, the geometric meaning of the nonlinear terms in the tangent stiffness matrix is discussed. A computer program for the whole range analysis of steel frames considering the geometrical and material nonlinearities is implemented, the load carrying capacity can also be determined.
Abstract：
In this paper the theory of viscoplasticity is applied for the analysis of sea ice loads, and a mathematical creep model under uniaxial stress condition is generalized to a triaxial formulation. The maximum static ice forces exerted on a single rectangular pile are claculatedaccording to this model by using the nonlinear finite element method. The results are compared with those from other sources. The model seems to be more realistic and reasonable than those in present use such as the plastic limit analysis method and the reference stress method.
In this paper the theory of viscoplasticity is applied for the analysis of sea ice loads, and a mathematical creep model under uniaxial stress condition is generalized to a triaxial formulation. The maximum static ice forces exerted on a single rectangular pile are claculatedaccording to this model by using the nonlinear finite element method. The results are compared with those from other sources. The model seems to be more realistic and reasonable than those in present use such as the plastic limit analysis method and the reference stress method.
1993, 10(4): 58-61.
Abstract：
It is quite notorious that the finite element solutions of transient heat conduction equation are very likely oscillating and/or over-bounding ones In order to eliminate the oscillation and/or over-bounding several criteria have been proposed by the author. In this paper a technique for smoothing the oscillation in the numerical results is presented. A high-order discretization scheme in time dimension is also discussed.
It is quite notorious that the finite element solutions of transient heat conduction equation are very likely oscillating and/or over-bounding ones In order to eliminate the oscillation and/or over-bounding several criteria have been proposed by the author. In this paper a technique for smoothing the oscillation in the numerical results is presented. A high-order discretization scheme in time dimension is also discussed.
Abstract：
The Bayes method is one of the basis of statistical decision methods. Through sampling, a prior distribution can be modified, hence the uncertainty of a parameter can be reduced. In harbor engineering, the numbers of the actual trial piles is nuall. It is difficult to obtain the exact estimation of Load-bearing capacity parameter for a single pile. In this paper, the Bayes method is applied considering the influence of subjective uncertainty on the load-bearing capacity as result is considered. The example given in this pap3r shows that the method is very useful in practice.
The Bayes method is one of the basis of statistical decision methods. Through sampling, a prior distribution can be modified, hence the uncertainty of a parameter can be reduced. In harbor engineering, the numbers of the actual trial piles is nuall. It is difficult to obtain the exact estimation of Load-bearing capacity parameter for a single pile. In this paper, the Bayes method is applied considering the influence of subjective uncertainty on the load-bearing capacity as result is considered. The example given in this pap3r shows that the method is very useful in practice.
Abstract：
A continuous velocity field referred to rectangular coordinatesis proposed to plane strain drawing through dies with a circular profile. A parametric integration and integral as a function of the upper limit are applied. And an upper bound analytical solution was obtained.
A continuous velocity field referred to rectangular coordinatesis proposed to plane strain drawing through dies with a circular profile. A parametric integration and integral as a function of the upper limit are applied. And an upper bound analytical solution was obtained.
1993, 10(4): 77-82.
Abstract：
In this paper, according to the theory of Bessel function, the dynamic ring fundamental solution of a plate on foundation of two-parameters are obtained for five cases of ω2>k/ph、ω2=h/ph、k/ph>ω2>k/ph-Gp2/(4Dph)、ω2=k/ph-Gp2/(4Dph) and ω2p2/(4Dph).
In this paper, according to the theory of Bessel function, the dynamic ring fundamental solution of a plate on foundation of two-parameters are obtained for five cases of ω2>k/ph、ω2=h/ph、k/ph>ω2>k/ph-Gp2/(4Dph)、ω2=k/ph-Gp2/(4Dph) and ω2
1993, 10(4): 83-92.
Abstract：
This paper presents the methods for automatic generating of the element node topology matrix (ENTM) and automatic assembling the whole node topology matrix (WNTM) on the basis of the topology analysisof finite element mesh. The algorithms of automatic generating and automatic assembling are given. The methods implement the full automatic generation procedure of 2D and 3D finite element mesh andprovide the basis for analyzing relativity of elements and nodes.
This paper presents the methods for automatic generating of the element node topology matrix (ENTM) and automatic assembling the whole node topology matrix (WNTM) on the basis of the topology analysisof finite element mesh. The algorithms of automatic generating and automatic assembling are given. The methods implement the full automatic generation procedure of 2D and 3D finite element mesh andprovide the basis for analyzing relativity of elements and nodes.
1993, 10(4): 93-101.
Abstract：
In this paper ths cables in prestressed structures are treated as two-node elements. The expression of the total distributed load on cables is deduced, the incremental stiffness matrix of cable elements is presented, and a consistent mass matix of cable elements is derived. The static and dynamic equilibrium equations of prestressed structures are established and the tracing stratage for whole prestres-sing procedures of structures is provided. Some numerical examples are given and the results show that the cable element theory of this paper has excellent accuracy.
In this paper ths cables in prestressed structures are treated as two-node elements. The expression of the total distributed load on cables is deduced, the incremental stiffness matrix of cable elements is presented, and a consistent mass matix of cable elements is derived. The static and dynamic equilibrium equations of prestressed structures are established and the tracing stratage for whole prestres-sing procedures of structures is provided. Some numerical examples are given and the results show that the cable element theory of this paper has excellent accuracy.
Abstract：
The integrally hydroforming technology of spherical tanksis a newly invented technology for manufacturing spherical tanks. And the 32-petal spherical tanks are one kind of earliest spherical tanks developed by this new technology. In the present paper the hydroforming process of this kind of spherical tanks is simulated usiug a nonlinear FEM program, the variation process .of the shell configuration and displacementsand the springback regularities are analysed, the deformation features of this kind of shells and the process measures taken are also compared with the experimental results. Three-dimensional Kirchhoff thin shell elements with 8-noded quadrilaterals are used, the displacements of the intersection points of weld seams and- the bonding effects of the shell are taken into accout. The hydrobulging process of the shell are simulated very well using FEM, and the FEM results are in good agreement with those of the experiments.
The integrally hydroforming technology of spherical tanksis a newly invented technology for manufacturing spherical tanks. And the 32-petal spherical tanks are one kind of earliest spherical tanks developed by this new technology. In the present paper the hydroforming process of this kind of spherical tanks is simulated usiug a nonlinear FEM program, the variation process .of the shell configuration and displacementsand the springback regularities are analysed, the deformation features of this kind of shells and the process measures taken are also compared with the experimental results. Three-dimensional Kirchhoff thin shell elements with 8-noded quadrilaterals are used, the displacements of the intersection points of weld seams and- the bonding effects of the shell are taken into accout. The hydrobulging process of the shell are simulated very well using FEM, and the FEM results are in good agreement with those of the experiments.
1993, 10(4): 110-116.
Abstract：
According to the design experiences accumulated in engineeringover a long period, making resonable and valid simplifications, a corresponding mathematical model is obtained in this paper. Based on the sequential entropy iterative method of structural optimization improvedby the author, the optimal design of thislarge scale stnutur is possible to be performed. The rules about the influence of the major factors on the price of structure are profoundly investigated in the paper. The optimal range of major factors is given. The results give a reference to the design in engineering.
According to the design experiences accumulated in engineeringover a long period, making resonable and valid simplifications, a corresponding mathematical model is obtained in this paper. Based on the sequential entropy iterative method of structural optimization improvedby the author, the optimal design of thislarge scale stnutur is possible to be performed. The rules about the influence of the major factors on the price of structure are profoundly investigated in the paper. The optimal range of major factors is given. The results give a reference to the design in engineering.
1993, 10(4): 117-124.
Abstract：
Wind vibration often has more effectes on the design of tall buildings, high-tower structures and big-span structures, for the periods of the structures are all long. Having exploried a great number of records of pulsating wind in the spot, this paper studies the power spectrum of pulsating wind velocity by using parameter identification technique, and the comparison with the spectrum of level wind velocity of Davenport. Also, using Auto-regressive Moving Average Model and parameter identification technique, the paper presents a method to simulate pulsating wind according to the spectra of level wind velocity of Davenport, provided by Xi’an Institute of Heat Engineering and presentedin this paper, respectively.
Wind vibration often has more effectes on the design of tall buildings, high-tower structures and big-span structures, for the periods of the structures are all long. Having exploried a great number of records of pulsating wind in the spot, this paper studies the power spectrum of pulsating wind velocity by using parameter identification technique, and the comparison with the spectrum of level wind velocity of Davenport. Also, using Auto-regressive Moving Average Model and parameter identification technique, the paper presents a method to simulate pulsating wind according to the spectra of level wind velocity of Davenport, provided by Xi’an Institute of Heat Engineering and presentedin this paper, respectively.
1993, 10(4): 125-131.
Abstract：
In this paper, the optimal designs of maximing axisymmetricalbuckling load of the polarity orthogonal anisotropic annular plate under volume and geometry constaints are studied. Some examples are given.
In this paper, the optimal designs of maximing axisymmetricalbuckling load of the polarity orthogonal anisotropic annular plate under volume and geometry constaints are studied. Some examples are given.
1993, 10(4): 132-138.
Abstract：
Considering the large-scale deformation of rings, the transientresponses of rings under several types of lateral impact loads are simulatedwith a code which is based on a UL FEM central-difference algorithmin the present paper. In this code developed by authors, the one-orderself-adaptive idea is employed in dealing with Mesh instability and contact-impact interfaces.
Considering the large-scale deformation of rings, the transientresponses of rings under several types of lateral impact loads are simulatedwith a code which is based on a UL FEM central-difference algorithmin the present paper. In this code developed by authors, the one-orderself-adaptive idea is employed in dealing with Mesh instability and contact-impact interfaces.
1993, 10(4): 139-149.
Abstract：
In this paper, the zoning-direct method (ZDM) of automatic mesh generation, a efficient and convenient method, is presented. With ZDM, the 3-D FE mesh can be generated to meet the requirement of specialnode distribution, and the element information, node coordinates and otherdata for FE computation can be automatically formed.
In this paper, the zoning-direct method (ZDM) of automatic mesh generation, a efficient and convenient method, is presented. With ZDM, the 3-D FE mesh can be generated to meet the requirement of specialnode distribution, and the element information, node coordinates and otherdata for FE computation can be automatically formed.
