Search

In many cases, it is advantageous to discretize a domain using several finite element meshes instead of a single mesh. For example, in fluid-structure interaction problems, an Eulerian mesh is advantageous for the fluid domain while a Lagrangian mesh is most suited for the structure. However, the interface conditions between different types of meshes often lead to significant errors. A method of treating different meshes by smoothly varying the description from Lagrangian to Eulerian in an interface or blending domain is presented. A Lagrangian mesh is used for the structure while two different types of mesh are used for the fluid. Arbitrary Lagrangian-Eulerian (ALE) meshes are used in the regions of the fluid-structure interfaces while Eulerian meshes are used for the remainder of the fluid domain. A blending function is used to couple the ALE and Eulerian meshes to ensure a smooth transition from one mesh to another. The method is tested on two fluid-structure problems — flow past a hinged plate, and fluid expansion in a closed container. Results are in good agreement with standard finite element and analytical solutions.

A global-local analysis methodology based on fluid-structure coupling is used to investigate the mechanical responses of both composite and steel risers. Since the design of the riser system can be a daunting task, involving hundreds of load cases for global analysis, semi-empirical fluid load models are considered for the reduced order computations of full-scale riser models. The structural performance of composite risers under real sea current conditions is investigated systematically and discussed with regard to the practical concerns in full-scale settings. The failure envelops of internal liners are found to be within that of the composite layers, which reveals that the liner is the weakest link for composite riser design. Results show that the composite risers can be more prone to vortex-induced-vibration (VIV) due to their lower structural frequencies. In the present study, the composite riser yields 25.5% higher RMS strains than the steel riser. Placement of buoyancy modules along the riser may be critical for the design against VIV, and our results show that the modules are not recommended at the top region of the riser, especially if a top-sheared current is expected. Instead, it is preferable to implement them at the bottom-half portion of the riser and as a continuously buoyed region rather than short discrete buoys separated with gap spaces.

An approach for stress analysis of multilayered composite cylinders is proposed for the analysis of new composite risers used in deep-water oil production of offshore petroleum industries. Risers essentially comprise long cylindrical sections connected end-to-end. In the formulation, only stresses and strains that are continuous through the thickness of the multilayered composite risers are taken to be equal to reported solutions for homogenous orthotropic hollow cylinders using homogenized material properties. These stress and strain solutions are then used to calculate the remaining discontinuous stresses and strains from the material properties of individual layers of materials. The homogenized elastic constants of cylindrically orthotropic composite risers are derived from forcedeformation equivalence, taking into account the stress and strain distributions in each layer. Four typical loading conditions are considered in the stress analysis, namely, internal and external pressures, axial loading, bending, and torsion. Examples of homogenized elastic constants and stress analyses of composite cylindrical structures with different layups and materials are presented to demonstrate the application of the proposed method. The results compared very favorably with those from other solutions. This method provides practical benefits for the design and analysis of composite risers. Because there is no requirement to explicitly enforce interfacial continuity in this method, stress analyses of composite cylinders with many layers of different fiber angles or materials can be carried out efficiently. The homogenized elastic constants can greatly expedite the analysis of entire composite riser systems by replacing complex models of riser sections with homogenized riser sections. [DOI: 10.1115/1.4024695]

Molecular dynamics simulations, supported by experimental characterizations, show that the tribology of nano-structured polymer interfaces are largely influenced by the inter-locking of diffusion-induced polymer chains. The degree of interfacial locking was found to determine the mechanisms of tribological behavior. The instantaneous friction coefficient shows regular stick–slip behavior at low sliding speed due to the concurrency of molecular deformation. Surface melting was found at the commencement of slipping. With increasing sliding speed, the regularity of the stick–slip cycles is less obvious and the magnitude of the fluctuations in friction coefficient decreased due to increasing atomic collisions and less durable slip rebound. Stick–slip phenomenon is reduced gradually before finally converging to complete dynamic frictional sliding at high sliding speed. Based on interfacial diffusion conditions, three mechanisms of interface deformation as ‘brushing’, ‘combing’ and ‘scissoring’ were proposed, among which reversible ‘brush’ was considered as the main source of interfacial deformation and dominant mechanism of the tribological behavior. Comparatively, ‘combing’ and ‘scissoring’ only took place in case of significant interfacial diffusion, and were not reversible. The interfacial structure also imposes effects on the influences of other factors to the frictional characteristics. In models with non-diffusive interface, higher pressure from the slider flattens the substrate surface, and thereby reduces the friction coefficient. In conclusion, once the interfacial structure was known, the tribological behaviors become predictable and qualitatively consistent with experimental observations.

Biochemical studies have determined that reoviruses attach to cells by combining attachment protein sigma1 to the binding interface of its receptor protein junctional adhesion molecule A (JAM-A), and the interface normally takes care of the homodimerization of JAM-A. Tighter binding and slower dissociation of for the sigma1-JAM complex than for the JAM-JAM complex have been probed by both biological and atomic force microscopy experiments; however, the mechanism of the binding preference of the attachment protein for JAM-A still remains unclear. With the help of classical and steered molecular dynamics and energy calculations, the unbinding forces and kinetic properties of the complexes are investigated, together with detailed structural information analyses. A multireceptor mechanism is proposed for the binding preference, which can be helpful for future viral infection and vector targeting studies.

A 3-year collaboration program between the Institute of Microelectronics, National University of Singapore, and the University of Cambridge has been established with the following two objectives: (i) to establish a mechanics and physics-of-failure based board-level test methodology; (ii) to establish design guidelines and, ultimately, failure criteria for board level interconnect during drop/impact test. The following accomplishments have been achieved in the first year of the program: (i) the mechanics and physics of failure in a typical board-level test have been established (ii) the drop impact characteristics of 6 commercial portable products have been comprehensively surveyed; (iii) a weakness has been identified in the drop impact strength of SnAgCu solder alloy (acknowledged as the leading Pb-free solder candidate).Copyright © 2003 by ASME

Molecular dynamics simulations are performed on multiwalled carbon nanotubes (MWCNTs) under axial compression to investigate the effects of the number of walls and their van der Waals (vdW) interaction on the buckling behaviors and mechanical properties (Young's modulus and Poisson's ratio). The Brenner second-generation reactive empirical bond order and Lennard-Jones 12-6 potential have been adopted to describe the short-range bonding and long-range vdW atomic interaction within the carbon nanotubes, respectively. In the presence of vdW interaction, the buckling strain and Young's modulus of MWCNTs increase as the number of tubes is increased while keeping the outermost tube diameter constant, whereas Poisson's ratio was observed to decrease. On the other hand, when the MWCNTs are formed by progressively adding outer tubes while keeping the innermost tube diameter constant, Young's modulus and buckling strain were observed to decrease, whereas Poisson's ratio increases. The buckling load increases with increasing the number of walls due to the larger cross-sectional areas. Individual tubes of MWCNTs with a relatively large difference between the diameters of the inner and outer tubes buckle one at a time as opposed to simultaneously for MWCNTs with a relatively small difference in diameters.

Density functional theory calculations on the structural and mechanical properties of two intermetallic compounds, Ni3Sn4 and Ag3Sn, are reported. The first-principles calculations predict the lattice constants and elastic constants of these Sn-based compounds. The results for lattice constants are found to be within 3% error of the experimental values. Bounds on polycrystalline elastic properties were then obtained, and these are close to the range of experimental values reported. The results provide further evidence for the usefulness and applicability of first-principles calculations when experimental data are sparse or unavailable.

In this paper, molecular dynamics simulations (MDS) are performed on single-walled carbon nanotubes (SWCNTs) in order to study the effects of chirality on their buckling behavior under axial compression. In the MDS, the Tersoff-Brenner potential is used to describe the interaction of carbon atoms in the SWCNTs. The sensitivity of the buckling strains and buckling modes with respect to the chirality of SWCNT is investigated by modeling SWCNTs with different chiral angles, varying from 0° to 30°, but keeping the length-to-diameter ratio constant. The carbon nanotubes are also analyzed using a continuum cylindrical shell model based on the theory of nonlocal elasticity so as to assess its validity in predicting the buckling strains when compared with the results that are obtained by MDS. The differences between the buckling strains at the continuum scale and that at the nanoscale are also studied. The present analysis and results are helpful in understanding the buckling behaviors of axially compressed carbon nanotubes. This knowledge is important for the application of carbon nanotubes as building blocks of nanomechanical devices.

The elastic constants of polycrystalline Cu6Sn5—an intermetallic in lead-free alternatives of several material systems—are presented. The results are obtained by applying: (i) Reported crystallographic structure of monoclinic single crystal Cu6Sn5, (ii) structure optimization and determination of single crystal elastic constants from first principle calculations, and (iii) limit analysis of polycrystal stiffness based on single crystal properties. The agreement between the calculated Young’s modulus (120 GPa) and those from nanoindentation experiments (112–125 GPa), and the tight bounds on the predicted polycrystal values give a measure of confidence in other calculated properties for which experimental data are unavailable.