Shen-Haw Ju

Resonance characteristics of high-speed trains passing simply supported bridges

This paper investigates the resonant characteristics of three-dimensional bridges when high-speed trains pass them. Multi-span bridges with high piers and simply supported beams were used in the dynamic finite element analysis. The dominated train frequencies proposed in this study can be clearly seen from the finite element result. To avoid resonance, the dominated train frequencies and the bridge natural frequencies should be as different as possible, especially for the first dominated train frequency and the first bridge natural frequency in each direction. If the two first frequencies are similar, the bridge resonance can be serious. This study also indicates that a suitable axial stiffness between two simple beams can reduce vibrations at a near-resonance condition. The axial stiffness of the continuous railway and the friction of the bearing plate should be enough to obtain this axial stiffness.

Numerical investigation of a steel arch bridge and interaction with high-speed trains

Abstract This paper studies the vibration characteristics of a three-dimensional (3D) arch bridge when high-speed trains pass it. An accurate representation of the moving wheel element was developed; therefore, two elements for modeling a simple beam can obtain accurate results. In the dynamic analysis, this study advises two simple criteria to predict the train-bridge resonance effect. The first criterion is that the dominated train frequencies of nVtrain/Ltrain similar to the first several bridge natural frequencies will produce resonance, where Vtrain is the train velocity, Ltrain is the compartment length and n is a positive integer. The second criterion is that the resonance will be produced when the train velocity is higher than 0.5Lf2, where L is the bridge length and f2 is the bridge natural frequency of the anti-symmetric shape along the bridge direction. Finite element analyses indicate that the two criteria can predict the resonance of the arch bridge and the high-speed train precisely.

Comparisons of contact pressures of crowned rollers

Abstract The present work determines the contact pressure and stress concentration between the crowned roller and the raceway by using three-dimensional finite element analysis. A number of crowned profiles with various dimensions were examined. Fine meshes and node-to-Hermit-surface contact elements were used along the contact surface in order to obtain accurate analysis results. A table was generated to show the stress concentration near the roller edge for various crowned profiles and dimensions. This table indicates that the exponential profile is the optimal crowned profile to eliminate stress concentration.

Simulating three-dimensional stress intensity factors by the least-squares method

The main purpose of this paper is to investigate the accuracy of the least-squares method incorporating the finite element method for finding three-dimensional (3-D) Stress Intensity Factors (SIFs). Numerical simulations in this paper indicate that the least-squares method can be used to calculate 3-D SIFs accurately, if three or more than three displacement or stress terms are included. The calculated SIFs of this method are independent of the maximum radius of the area from which data is included; furthermore, a very fine mesh is not necessary.

Simulating stress intensity factors for anisotropic materials by the least-squares method

The main purpose of this paper is to investigate the accuracy of the least-squares method incorporating the finite element method for finding the stress intensity factors of composite materials. Numerical simulations in this paper indicate that the least-squares method can be used to calculate stress intensity factors accurately, if seven or eight displacements terms are included. The calculated stress intensity factors by using the least-squares method can be as accurate as the calculated J-integrals by using the J-integral formulation. If seven or eight displacement terms are included, the calculated stress intensity factors of this method are independent of the maximum radius of the area from which data is included; furthermore, a very fine mesh is also not necessary.

Thermoelastic Determination of KI and KII in an Orthotropic Graphite–Epoxy Composite

Motivated by difficulties in recording reliable data in the immediate vicinity of a crack tip, the general fracture mechanics concept often employed for isotropy of utilizing stresses which are valid away from the crack with distance measurements is extended to orthotropy. Specifically, stress intensity factors, KI and KII, for inclined cracks are determined here in a uniaxially-loaded orthotropic graphite–epoxy composite using measured temperatures and least-squares. An advantageous novel technique for evaluating the thermo-mechanical coefficients is also implemented by commingling numerical and measured information. The through cracks are inclined at either 85 or 45 with respect to the load, the latter also being the strong/stiff composite direction. Retaining higher-order terms in the stress representations enables using measured temperatures away from the crack. The present method of determining the two thermo-mechanical coefficients circumvents the challenges encountered when attempting to evaluate the largest thermo-mechanical coefficient from uniaxial coupons since this coefficient typically occurs in the weakest direction of the composite.

Numerical Simulation of Stress Intensity Factors via the Thermoelastic Technique

The purpose of this study is to investigate the accuracy of the least squares method for finding the in-plane stress intensity factorsKI andKII using thermoelastic data from isotropic materials. To fully understand the idealized condition ofKI andKII calculated from thermoelastic experiments, the total stress field calculated from finite element analysis is used to take the place of data obtained from real thermoelastic experiments. In the finite element analysis, theJ-integral is also calculated to compare with (KI2+KII2)/E evaluated by the least squares method. The stress fields near the crack tip are dominated by the two stress intensity factors; however, the edge effect will cause inaccuracy of the thermoelastic data near the crack tip. Furthermore, the scan area of thermoelastic experiments cannot be too small. Therefore, we suggest that three or four terms of stress function be included in the least squares method for evaluating stress intensity factors via the thermoelastic technique. In the idealized condition, the error can be smaller than 3 percent from our numerical simulations. If only ther−1/2 term (KI andKII) is included in the least squares method, even in the idealized case the error can be up to 20 percent.

Behaviors of a single crack in multiple bolted joints

This paper examines the pin load ratios and the stress intensity factors (SIFs) of a single crack in the multiple bolted joints by using finite element analyses. Cubic-spline contact elements and rigid links were used to model the contact surface between the bolt and the rigid pin. The least-squares method was used to determine the SIFs. The finite element results indicate that the cracked hole can still sustain the major part of the original loading at the uncracked condition. The first hole sustains the largest pin load and mode-I SIF, which are reduced little for crack propagation. This critical condition cannot be reduced by the arrangement of more pins in the plate. In this paper, two simple formulae were also investigated to fit the load ratios and SIFs of the multiple bolted-joints problems.

A Deformation Formula for Circular Crowned Roller Compressed Between Two Flat Plates

The main purpose of this paper is to develop a deformation equation for the circular crowned roller compressed between two plates. First, the roller is divided into three parts, two crowned parts and one cylindrical part. The superposition method is then introduced to obtain the roller stiffness. The stiffness contribution of the crowned parts is calculated by the classical Hertzian contact solution and the stiffness contribution of the noncrowned part is obtained by the Hoeprichs formula. Comparisons with various finite element results indicate that the deformation equation derived in this paper can be a good deformation formula for the circular crowned roller.

A three-dimensional frictional contact element whose stiffness matrix is symmetric

A three-dimensional contact element based on the penalty function method has been developed for contact frictional problems with sticking, sliding, and separation modes in finite element analysis. A major advantage of this contact element is that its stiffness matrix is symmetric, even for frictional contact problems which have extensive sliding. As with other conventional finite elements, such as beam and continuum elements, this new contact element can be added to an existing finite element program without having to modify the main finite element analysis program. One is therefore able to easily implement the element into existing nonlinear finite element analysis codes for static, dynamic, and inelastic analyses. This element, which contains one contact node and four target nodes, can be used to analyze node-to-surface contact problems including those where the contact node slides along one or several target surfaces.