A locking free formulation of 4-node bilinear shell element and its application to shell structures is demonstrated. The Enhanced Assumed Strain (EAS) method based on three-field variational principle of Hu-Washizu is used in the formulation. Transverse shear locking and membrane locking are circumvented by means of enhancing the displacement-dependent strain field with extra assumed strain field. Several benchmark shell problems are analyzed.

The formulation of one Timoshenko curved beam on which a pair of piezoelectric segments bonded is presented. The analytical-transfer matrix method is adopted to study structural responses of the surface-mounted actuation curved beam induced by an external force at the tip of the beam and by an applied voltage on the actuator. The effects of length, location and thickness of the piezoelectric pair on the structural responses of the entire curved beam are investigated.

In this paper, a methodology with workable procedures was proposed and successfully transferred the hysteresis loops of BGA solder joint specimen held by loading-directionally aligned grip system under oblique displacement controlled cyclic tests, into hysteresis loops of “representative” solder ball itself under proportional straining and constant strain rate cyclic tests. Under the elastic unloading during cyclic test, value of elastic modulus of bulk solder specimens and the linear behavior of grip system were used. Following the study of Endochronic cyclic viscoplasticity, the kernel function ρ(Z) = ρ0 exp (−KZ)/Zα was found to be independent of the oblique angle (Φ) of straining paths. However, the steady cyclic behavior of material function f(ξ) ≡ f0 in the definition of intrinsic time Z contained two functions: (1) is the effective inelastic strain amplitude, and (2) f(Φ) Φ(rad) between Φ = 0 (uniaxial) and Φ = π/2 (shear). With (ρ0, α, K) = (4MPa, 0.84, 46) and f0 = [0.24(π/2−Φ)2 − 0.018(π2−Φ) + 1.2] , the endochronic theory predicted experimental hysteresis loops of BGA Sn/3.5Ag/0.75Cu solder joint specimens and Φ = 0°, 27°, 45°, 63°, 90° quite well.

Results of ρ(Z) under f0 = 1 were found under independent study based on cyclic tests of bulk specimens under constant uniaxially strain amplitude and various constant strain rates. The values of (ρ0, α, K) = (7.3MPa, 0.84, 30) with f0 = 1.0 were all independent of strain rates. Comparisons of results of both types of specimens revealed that the values of (α, K) were almost the same but (1) ρ0 was smaller for micro-size solder ball, (2) is a material function of Sn/3.5Ag/0.75Cu itself (3) f(Φ) is not a material function, rather it is a special feature to reflect the effects of detail design of solder joint specimen and its connecting methods to the substrates

An improved method of space-time conservation element and solution element (CE/SE) is developed to solve the equations of conservation laws in fluid dynamics. The present method substantially differs in both concept and methodology from the traditional CE/SE method. In this paper the improved second-order CE/SE method is presented in a hexahedral grid. Furthermore, the present CE/SE method was successfully applied to solve the interaction problem of shock waves and detonation. Several numerical examples were also given. Numerical results have compared with the results of experiments and other computational methods. The compared results have shown a good agreement. The improved CE/SE method has higher accuracy and becomes a more prospective scheme.

For improving the performance of a miniature loudspeaker, engineers have been pursuing a goal of lowering its first damped frequency to have a broad and smooth region of its sound-pressure-level response. This article applies the approach of electro-mechano-acoustical analogy to study the first damped frequency of a miniature loudspeaker used for cellular phones. Using the analogy approach, the authors derive an explicit formulation for the sound-pressure-level of a typical miniature loudspeaker of dynamic type. The veracity of the derived formulation is verified by comparing with the experimental result recorded by B and K apparatus. The derived formulation is further simplified by neglecting factors that are either irrelevant or insignificant for the first damped frequency. Through basic algebraic and calculus operations, this article also presents an explicit form of the first damped frequency. This work has laid the platform for engineers to tailor the design in pursuit of improving the performance of miniature loudspeakers.

Due to a lack of closed-form solutions for three dimensional anisotropic bodies, the computational burden of evaluating the fundamental solutions in the boundary element method (BEM) has been a research focus over the years. In engineering practice, transversely isotropic material has gained popularity in the use of composites. As a degenerate case of the generally anisotropic material, transverse isotropy still needs to be treated separately to ease the computations. This paper aims to investigate the computational efficiency of the BEM implementations for 3D anisotropic, transversely isotropic, and isotropic bodies. For evaluating the fundamental solutions of 3D anisotropy, the explicit formulations reported in [1,2] are implemented. For treating transversely isotropic materials, numerous closed form solutions have been reported in the literature. For the present study, the formulations presented by Pan and Chou [3] are particularly employed. At the end, a numerical example is presented to compare the computational efficiency of the three cases and to demonstrate how the CPU time varies with the number of meshes.

We present the development of a MEMS piezoelectric bimorph generator, a cantilever type bimorph which was formed by laminating two PZT piezoelectric layers. Our bimorph generator can scavenge mechanical energy from ambient vibrations and transform it into useful electrical energy. Two poling configurations of the PZT piezoelectric layers of our bimorph MEMS generator were fabricated and tested. A tip proof mass used for adjusting the resonance frequency was also demonstrated. Experimental results confirm that our device possessed a maximum open-circuit output voltage of 1.91VP-P and a 3.42VP-P for a parallel polarization device and a serial polarization device, respectively with a 2g externally applied vibration. At an optimal resistive load, the maximum output power was 1.548μ–W and 1.778μ–W for a parallel polarization device and a serial polarization device, respectively.

This study presents an analytical model for describing propagation of Rayleigh waves along the impermeable surface of an unsaturated poroelastic half-space. This model is based on the existence of the three modes of dilatational waves which employ the poroelastic equations developed for a porous medium containing two immiscible viscous compressible fluids (Lo, Sposito and Majer, [13]). In a two-fluid saturated medium, the three Rayleigh waves induced by the three dilatational waves can be expressed as R1, R2, and R3 waves in descending order of phase speed magnitude. As the excitation frequency and water saturation are given, the dispersion equation of a cubic polynomial can be solved numerically to obtain the phase speeds and attenuation coefficients of the R1, R2, and R3 waves. The numerical results show the phase speed of the R1 wave is frequency independent (non-dispersive). Comparatively, the R1 wave speed is 93 ∼ 95% of the shear wave speed, and 28% to 49% of the first dilatational wave speed for selected frequencies between 50Hz & 200Hz and relative water saturation ranging from 0.01 to 0.99. However, the R2 and R3 waves are dispersive at the frequencies examined. The ratios of R2 and R3 wave phase speeds to the second and third dilatational wave speeds fall between 56% and 90%. The R1 wave attenuates the least while the R3 wave has the highest attenuation coefficient. Furthermore, the phase speed of the R1 wave under an impermeable surface approaches 1.01 ∼ 1.37 times of the R1 wave under a permeable boundary. Impermeability has significant influence on the phase speeds and attenuation coefficients of the R1 and R2 waves at high water saturation due to the existence of confined fluids.

This paper studies a novel approach to optimize trusses and truss-like structures for minimum weight design. It is based on the force method of analysis which is incorporated inside the optimization routine. The design variables in force formulation are the member cross sectional areas and the redundant forces in each load condition. The optimization method used is the method of center points using the inscribed hyperspheres to the feasible-usable design space. By incorporating the analysis step as part of the optimization problem, a separate structural solution phase, which is necessary in all other methods, is avoided resulting in large computational savings. In this article the simplest form of structures i.e. trusses are treated to illustrate the efficacy of the method. Stress limits on the members as well as limitations on their sizes, and linking among them, under several load conditions have been considered. Combination of the method of center points and force formulation results in a viable routine for structural optimization. Comparison of the example results with those obtained by others clearly shows the effectiveness and novelty of the proposed method.

An analysis of the Poiseuille-Couette flow of two immiscible fluids between inclined parallel plates is investigated. One of the fluids is assumed to be electrically conducting while the other fluid and channel walls are assumed to be electrically insulating. The viscous and Ohmic dissipation terms are taken into account in the energy equation. The coupled nonlinear equations are solved both analytically valid for small values of the product of Prandtl number and Eckert number (= ε) and numerically valid for all ε. Solutions for large ε reveal a marked change on the flow and rate of heat transfer. The effects of various parameters such as Hartmann number, Grashof number, angle of inclination, ratios of viscosities, widths and thermal conductivities are presented and discussed in detail.

In this paper, an analysis of the large-amplitude dynamic-plastic behavior of the circular plates with a rigid perfectly plastic material is presented. The plate is subjected to a short-time high-intensity impulsive load uniformly distributed over the surface. Modeling is complemented by using specific convex yield criteria. Corresponding to boundary conditions of the plate, it can be deformed through more than one mechanism, so, the mathematical formulation is based on the principle of calculus of variations in which the transverse displacement fields are assumed as a combination of appropriate paths. Based on the upper bound approach, the different terms of kinetic and consumed plastic energies likewise the applied impulse energy derived to produce an energy functional with unknown coefficients which is minimized through the displacement path. Finally, calculating the constants maximum residual deflection and strain distribution are obtained. Results of present model show satisfactory correlation with the empirical data for the different levels of the pulsed loads.

A distributed Brillouin fiber sensor was used to monitor the health of steel structures. We used this method on two steel beam specimens and longitudinal strain and cracks were detected well under different loads. The difference between the longitudinal strain measured by the distributed fiber sensor and that measured by strain gages is less than 4%. Traditional sensors or transducers that measure the average strain over a small region always miss cracks. The method proposed in this study gives very good results for the detection of cracks and the surrounding strain on a square pipe.

Many unresolved issues remain in the field of modelling of shape memory alloys. In this paper the problem of unequal elastic properties of austenite and martensite is addressed. We propose a modification of the micromechanical material model that enables the application of different Young's modulus for austenite and martensite. The corresponding computational model for the application of the micromechanical approach to modeling of superelasticity in shape memory alloys is demonstrated. Material properties for Ni-Ti alloy (50.8 at.% Ni) obtained from literature and from our own experiments were applied to the model and a sample calculation of a 3D model subjected to uniaxial loading was performed. The results were compared to experimental results obtained from tensile and compressive tests. In general the presented model predicts well the level of the superelastic stress plateau and maximum transformation strain in tension. The agreement in compression is worse but the overall characteristics of the tension-compression asymmetry are predicted correctly.

A numerical study is undertaken to simulate failure of solder joint caused by cyclic shear deformation. A progressive ductile damage model is incorporated into the rate-dependent elastic-viscoplastic finite element analysis, resulting in the capability of simulating damage evolution and eventual failure through crack formation. It is demonstrated that quantitative information of fatigue life, as well as the temporal and spatial evolution of fatigue cracks, can be explicitly obtained.