1992 Vol. 9 No. 3
1992, 9(3): 1-7.
Abstract：
This paper presents the stochastic finite element--maximum entropy method for reliability analysis of structures subjected to random loads with stochastically varying properties. In this method, firstly the structure response of uncertainties is evaluated using Neumann SFEM, and then the probability of failure of the structure is obtained using maximum entropy method. The numerical examples show that the method has higher accuracy for structural reliability analysis and the computational efforts are reduced.
This paper presents the stochastic finite element--maximum entropy method for reliability analysis of structures subjected to random loads with stochastically varying properties. In this method, firstly the structure response of uncertainties is evaluated using Neumann SFEM, and then the probability of failure of the structure is obtained using maximum entropy method. The numerical examples show that the method has higher accuracy for structural reliability analysis and the computational efforts are reduced.
1992, 9(3): 8-22.
Abstract：
Based on the potential energy method, a semi-discrete method is presented in this paper. The method can save much computing time than the standard finite element method and even the finite strip method. By abandoning one of the Vlasov's assumption in analysis of thin-walled members, which assumption is that the shear strains along the central line of the cross section are zero or constant, the present method can show the shear lag effect well. In this paper, the longitudinal displacement along the cross section is interpolated by a new proposed sectional spline function. By using the variational principle, a group of ordinary differential equations and natural boundary conditions can be derived. By solving these equations an analytical solution for the longitudinal displacement along the member axis can be obtained. The present method is suitable for any shape of cross section thin-walled members. Some typical numerical examples in this paper demonstrate the versatility, efficiency, accuracy and convergency of the proposed method.
Based on the potential energy method, a semi-discrete method is presented in this paper. The method can save much computing time than the standard finite element method and even the finite strip method. By abandoning one of the Vlasov's assumption in analysis of thin-walled members, which assumption is that the shear strains along the central line of the cross section are zero or constant, the present method can show the shear lag effect well. In this paper, the longitudinal displacement along the cross section is interpolated by a new proposed sectional spline function. By using the variational principle, a group of ordinary differential equations and natural boundary conditions can be derived. By solving these equations an analytical solution for the longitudinal displacement along the member axis can be obtained. The present method is suitable for any shape of cross section thin-walled members. Some typical numerical examples in this paper demonstrate the versatility, efficiency, accuracy and convergency of the proposed method.
1992, 9(3): 23-31.
Abstract：
This paper studies the stress and strain fields of the plastic axisymmetrical body, all the fundamental equations are considered. The equation controlling the stress distribution is deduced, and general method used to solve axisymmetrical problem are proposed.
This paper studies the stress and strain fields of the plastic axisymmetrical body, all the fundamental equations are considered. The equation controlling the stress distribution is deduced, and general method used to solve axisymmetrical problem are proposed.
1992, 9(3): 32-40.
Abstract：
The nonlinear response of the locally buckled box beam-column is reported in this paper. The locally buckled beam-column is treated as a model where the beam-column is divided into a series of the locally buckled cells that are connected at their ends. A large-deflection elastic-plastic finite strip analysis is implemented to calculate the elastic-plastic M-P-Φ curves of the cells. Then a finite integration approach is employed to evaluate the ultimate load of the beam-column.A test program on the cold-formed box columns loaded eccentrically was performed. The comparison of experimental and numerical results shows a good agreement. A design formula is proposed to estimate the ultimate strength of the cold-formed box columns subjected to a concentric loads.
The nonlinear response of the locally buckled box beam-column is reported in this paper. The locally buckled beam-column is treated as a model where the beam-column is divided into a series of the locally buckled cells that are connected at their ends. A large-deflection elastic-plastic finite strip analysis is implemented to calculate the elastic-plastic M-P-Φ curves of the cells. Then a finite integration approach is employed to evaluate the ultimate load of the beam-column.A test program on the cold-formed box columns loaded eccentrically was performed. The comparison of experimental and numerical results shows a good agreement. A design formula is proposed to estimate the ultimate strength of the cold-formed box columns subjected to a concentric loads.
1992, 9(3): 41-46.
Abstract：
According to the results calculated by the finite element program TJSDAP the parameter formulas of local flexibility of T and X type tubular joints are presented by using function fitting method. Numerical results of T type tubular joints are quite coincident with the results of the experiments and that obtained by some other parameter formulas presented aboard.
According to the results calculated by the finite element program TJSDAP the parameter formulas of local flexibility of T and X type tubular joints are presented by using function fitting method. Numerical results of T type tubular joints are quite coincident with the results of the experiments and that obtained by some other parameter formulas presented aboard.
Abstract：
In this paper, firstly using the commonly used piecewise continuum technique for the structures, a new continuous parallel model is taken as the computational model for the overall stability analysis of tall building frame-shear wall structures with N-stepped parameters and with floor slab and base soil deformation considered. The stability differentialequations for the model have been derived.Secondly, using the ODE (Ordinary Differential Equation) Solver, which is a general purpose program developed for solving various ODE problems, including eigen-value problems numerically, the critical loads and buckling modes have been obtained.Finally, a Computer program "OSATBS" is developed for engineering design, and some calculation examples are given.
In this paper, firstly using the commonly used piecewise continuum technique for the structures, a new continuous parallel model is taken as the computational model for the overall stability analysis of tall building frame-shear wall structures with N-stepped parameters and with floor slab and base soil deformation considered. The stability differentialequations for the model have been derived.Secondly, using the ODE (Ordinary Differential Equation) Solver, which is a general purpose program developed for solving various ODE problems, including eigen-value problems numerically, the critical loads and buckling modes have been obtained.Finally, a Computer program "OSATBS" is developed for engineering design, and some calculation examples are given.
1992, 9(3): 63-72.
Abstract：
In this paper, an extended form of generalized variational principle with given initial conditions and boundary conditions explicitly expressed and two arbitrary constants contained for linear elastodynamics is developed and various special cases of the principle are also derived.
In this paper, an extended form of generalized variational principle with given initial conditions and boundary conditions explicitly expressed and two arbitrary constants contained for linear elastodynamics is developed and various special cases of the principle are also derived.
1992, 9(3): 73-80.
Abstract：
On the basis of the early days experiments, the relation of the increments of stress to the increments of strain is proposed for the two way forced elements of masonry by assuming orthogonal anisotropy materail in the directions of mortar joint and the matrix of stiffness is presented which represents the increment relation in the directions of main stress by translation of elastic constants of orthegonal anisotropy mater-ail. The nonlinear analysis model of the elements of masonry is used to calculating the wall beams. It has been shown that the calculated results are approximately to those agree with the tested.
On the basis of the early days experiments, the relation of the increments of stress to the increments of strain is proposed for the two way forced elements of masonry by assuming orthogonal anisotropy materail in the directions of mortar joint and the matrix of stiffness is presented which represents the increment relation in the directions of main stress by translation of elastic constants of orthegonal anisotropy mater-ail. The nonlinear analysis model of the elements of masonry is used to calculating the wall beams. It has been shown that the calculated results are approximately to those agree with the tested.
1992, 9(3): 81-94.
Abstract：
In this paper, the initial functions of Biot consolidation of one layer subsoil and multilayer subsoils are established by using the Laplace and Hankel transformation and the recurrence law of matrices respectively. Then the equations for evaluate the vertical deformation due to the Biot consolidation are derived. These equations may be solved easily by using numerical integration. Only three simultaneous linear algebric equations are to be solved no matter how many layers are involved in the subsoils.
In this paper, the initial functions of Biot consolidation of one layer subsoil and multilayer subsoils are established by using the Laplace and Hankel transformation and the recurrence law of matrices respectively. Then the equations for evaluate the vertical deformation due to the Biot consolidation are derived. These equations may be solved easily by using numerical integration. Only three simultaneous linear algebric equations are to be solved no matter how many layers are involved in the subsoils.
1992, 9(3): 95-100.
Abstract：
Silos are special structures widely used in practice. But dynamic pressure problems of silos wall have not been solved till now. In this paper these problems are studied. The variations of momentum of storing material are considered in analysis and calculating formulas are given. These formulas have simple form. The calculating results are comparedwith some experimental results and they are close to each other.
Silos are special structures widely used in practice. But dynamic pressure problems of silos wall have not been solved till now. In this paper these problems are studied. The variations of momentum of storing material are considered in analysis and calculating formulas are given. These formulas have simple form. The calculating results are comparedwith some experimental results and they are close to each other.
Abstract：
The calculation of the bending load of the normal section for the common reinforced concrete component which has a ring-shaped cross-section and is evenly disposd with reinforcing bars along its periphery,involves solving the equation with triangular functions. Generally, the iteration method is used for this and therefore the calculation is very combersome. In this paper, a sectional quadra-tic approximation method is proposed which is fast, simple to use and with very high accuracy.
The calculation of the bending load of the normal section for the common reinforced concrete component which has a ring-shaped cross-section and is evenly disposd with reinforcing bars along its periphery,involves solving the equation with triangular functions. Generally, the iteration method is used for this and therefore the calculation is very combersome. In this paper, a sectional quadra-tic approximation method is proposed which is fast, simple to use and with very high accuracy.
1992, 9(3): 107-116.
Abstract：
In this paper, deriving directly from the stress-strain relationsfor concrete and steel, the relation among moment, normal force and curvature (M-N-φ) on PC (Prestressed Concrete) sections is determined by the method presented. The methods of determing the M-N-φ relation for the beam section and the column section of a frame are unified. The displacement mode and stiffness matrix for element with variable stiffnessare proposed. The nonlinear analysis for PPC (Partially Prestres sed Concrete) frames is implemented. The obtained results by the non-linearanalysis are in good agreement with the results of the four PPC frame tests conducted by the author.
In this paper, deriving directly from the stress-strain relationsfor concrete and steel, the relation among moment, normal force and curvature (M-N-φ) on PC (Prestressed Concrete) sections is determined by the method presented. The methods of determing the M-N-φ relation for the beam section and the column section of a frame are unified. The displacement mode and stiffness matrix for element with variable stiffnessare proposed. The nonlinear analysis for PPC (Partially Prestres sed Concrete) frames is implemented. The obtained results by the non-linearanalysis are in good agreement with the results of the four PPC frame tests conducted by the author.
1992, 9(3): 117-123.
Abstract：
Moment and shear geometric sequence method is based on moment and shear distribution method. By using some appropriate way of transmiting moments and shears, the moments and shears at the ends of structure elements can be convergently calculated with a geometricsequence to have the exact solution, and the problem of slow convergence rate of moments and shears at the ends of structure elements can be also solved when moment and shear distribution method are used in the analysis.
Moment and shear geometric sequence method is based on moment and shear distribution method. By using some appropriate way of transmiting moments and shears, the moments and shears at the ends of structure elements can be convergently calculated with a geometricsequence to have the exact solution, and the problem of slow convergence rate of moments and shears at the ends of structure elements can be also solved when moment and shear distribution method are used in the analysis.
1992, 9(3): 124-130.
Abstract：
This paper presents an approximate method of solving natural characteristics of transverse vibration of continuous rectangular plates with arbitrary intermediate supports parallel to the boundaries in one or two ways. Ritz method is used approximately to calculate natural frequencies by selecting the combination of vibrating beam functions and polynomial as the basis functions. Formulae are easy to program. Finally, some numerical results are given and compared with those obtained by other methods, which shows this method has good accuracy.
This paper presents an approximate method of solving natural characteristics of transverse vibration of continuous rectangular plates with arbitrary intermediate supports parallel to the boundaries in one or two ways. Ritz method is used approximately to calculate natural frequencies by selecting the combination of vibrating beam functions and polynomial as the basis functions. Formulae are easy to program. Finally, some numerical results are given and compared with those obtained by other methods, which shows this method has good accuracy.
1992, 9(3): 131-135.
Abstract：
Results of analysis reveal that the solution for the displacement at zero-mass point by using the conventional method of mode-superposition is not correct. In order to solve the problem, this paper presents the modified method of mode-superposition, by which we can reach the accurate solution. Furthermore, the strict theoretical testification of the modified mothod is given in this paper.
Results of analysis reveal that the solution for the displacement at zero-mass point by using the conventional method of mode-superposition is not correct. In order to solve the problem, this paper presents the modified method of mode-superposition, by which we can reach the accurate solution. Furthermore, the strict theoretical testification of the modified mothod is given in this paper.
1992, 9(3): 136-142.
Abstract：
Abstract In this paper, a new method named elastoplastic contact boundary element method developed by the authors of this paper is used in the simulation of rolling process. As the roller is an elastic body, and workpiece is made of a kind of elastoplastic material, the rolling process can be taken as an elastoplastic contact problem. The least assumption is considered in simulating the rolling process, therefore, an effective and accurate numerical approach is presented for analysing the rolling process.
Abstract In this paper, a new method named elastoplastic contact boundary element method developed by the authors of this paper is used in the simulation of rolling process. As the roller is an elastic body, and workpiece is made of a kind of elastoplastic material, the rolling process can be taken as an elastoplastic contact problem. The least assumption is considered in simulating the rolling process, therefore, an effective and accurate numerical approach is presented for analysing the rolling process.
