A concept of a parameter U* has been introduced by the authors to express load transfer paths in a structure. In this paper, a new extended expression of U* is introduced, and it is shcown that the new parameter can be applied to structures with complicated multiple support conditions. The present expression that is derived by matrix formulation of internal stiffness and initial deformation includes the definition of U* in the previous report. Examples of U* calculation for structures with multiple support conditions are demonstrated. Calculation results for a simple structure and an actual truck cab body show the effectiveness of the present method of U* calculation.

View full abstract

The principle of structural analysis by the ε-method was investigated for a two dimensional multi-storied and multi-span rectangular rigid frame loaded vertically or horizontally, in cases where rigid zone and shearing deformation are considered. The members, having such as earthquake-registant wall and wall structure in the rigid frame, can not be successfully analized only by bending deformation theory. When the sectinal form has fairly larger height compared to the length of the member in the rigid frame, shearing deformation must be taken into account. Finally, the results calculated by the digital computing program on the basis of proposal algorithm by the ε-method were shown in several figures.

View full abstract

In past few years, we have resorted to the methodology of special-purpose computer in order to accelerate the Boundary Integral Equation Method (BIEM). Here we applied a special-purpose computer MDGRAPE-2 to the acceleration of the BIEM for three-dimensional elastostatics. The formulation was developed so that MDGRAPE-2 can evaluate the layer-potentials of the BIEM. The implemented BIEM code using MDGRAPE-2 and the iterative solver GMRES was nearly 70 or more times faster than the naive BIEM code in solving a boundary value problem with 3.9 million unknowns.

View full abstract

As the first step of the development of a computational tool to evaluate the impact resistance of multi-layer coatings, the authors have conducted the formulation of elasto-viscoplastic damage constitutive equation based on continuum damage mechanics and extended to high strain-rate behaviors, the identification of the constitutive modeling based on the tensile test results for each layer material, and its applications to the finite element analysis of tensile fracture behaviors of multi-layer coatings. The obtained results of the identification of constitutive modeling and the tensile fracture analysis have agreed well with the corresponding experimental results. The present dynamic damage mechanics models can be expected to be used as a computational tool for the evaluation and the design of multi-layer coatings.

View full abstract

For the development of a computational tool to evaluate the impact resistance of multi-layer coatings, the authors have conducted the finite element formulation for the dynamic damage analysis of multi-layer coatings by using the elasto-viscoplastic damage constitutive equation based on continuum damage mechanics. The proposed method has been applied to the analysis of multi-layer coatings under transverse impact using the material constants determined by the material test results for each coating layer. The identified constitutive modeling and the calculated damage zones under lateral impact have almost agreed well with the corresponding experimental results, although separation of the damaged and fractured zone from the tested specimens has not yet completely simulated because of the use of relatively coarse mesh.

View full abstract

A single-fiber pull-out model composite tor an aramid/epoxy system was specially designed to measure the stress distribution of the aramid fiber embedded in the matrix using micro-Raman spectroscopy. The stress transfer length of the fiber obtained was about 400-500 μm, which was equal to the result of FEM analysis. Just after the initiation and propagation of the fiber/matrix interfacial debonding, the fiber was broken, and the fiber in the matrix had the axial tensile residual stress. The tensile residual fiber axial stress showed the maximum at around the tip of the interfacial debonding. The stress was reduced after the specimen was kept in air at 80°C for 44h, and it became almost equal to zero after being immersed in deionized water at 80°C for 44h. This behavior agreed with the result of FEM analysis, in which the friction coefficient was introduced in the fiber/matrix interface. The axial residual stress was caused by the friction between the fiber and matrix, due to the compressive stress acting between the resin and the fiber, which was caused by the difference of the coefficient of thermal expansion.

View full abstract

This paper suggests the method to create the individual finite element model by the dimensions only that can be measured easily. Individual head FE models are constructed by transformed a basic head FE head model with matched the individual head outer shape to the basic one by using Free Form Deformation (FED) technique. The individual head outer shape is created by weighted composition of the representative shapes of Japanese head calculated by the method combining Multi Dimensional Scaling (MDS) with FFD technique. Weights for the composition are calculated by minimizing the error between measured head dimensions and calculated ones. Furthermore, the construction method was validated by the comparison of the shape and pressure responses in brain of the head FE model created by the present method with those of the head FE model created from CT images. The suggested method can create individual FE model easily even when individual medical images (e.g. CT or MRI) cannot be taken.

View full abstract

We propose new finite element method using numerical database (FEM-ND) concerning stress-strain relationship of solids. The database is defined on six dimensional space spanned by six independent components of strain tensor. The database automatically grown on-the-fly by density functional theory (DFT) based total energy calculation during FEM analysis. Once we attain mature database, the FEM-ND does not need any the heavy DFT calculations more because all the required data by FEM analyses can be restored from the database. Therefore total computational time becomes shorter and shorter significantly with increasing number of FEM analysis cases. We perform test calculations by applying this method to aluminum perfect crystal.

View full abstract

Generally, reinforcement clusters are formed in a composite. For the case that the rapture strain of the reinforcement in such a composite is smaller than that of the matrix, some of the reinforcements may break in the cluster when the composite is subjected to the external force. Thus the broken reinforcements and the unbroken reinforcements coexist within the cluster. Such a cluster can be regarded as a multi-phase inhomogeneity because the cluster contains different inhomogeneities correspond to the broken and the unbroken reinforcements. From another viewpoint, the cluster can be regarded as a triple inhomogeneity because a crack exists in the reinforcement and the reinforcements are gathered to the cluster. Therefore, the reinforcement cluster in the composite can be modeled as a multi-phase triple inhomogeneity. In the present study, the average interaction stress is derived for the material containing many multi-phase triple inclusions by adopting the Mori-Tanaka theorem. The stress-strain relations to every regions in the multi-phase triple inclusion are formulated by using the resultant average interaction stress. Moreover, the stress-strain relations to the multi-phase triple inhomogeneity are derived and they are compared with those to the multi-phase triple inclusion, and the equivalent expressions for the multi-phase triple inhomogeneity are derived to the every regions in it.

View full abstract

The fractal branch and bound method (FBBM) has been developed by authors for optimizing the stacking-sequence of composite structures. The method can directly optimize feasible stacking-sequences comprising of limited fiber angles, 0°, ±45°and 90°layers. The applicability of the method has been confirmed for buckling load maximization and flutter limit maximization problem in the previous studies. In this paper, we focus on the composite wing flutter problem. The relationship between the base fiber angle variance and the optimized flutter limit by means of fractal branch and bound method is discussed. To improve the optimal flutter limit of the composite wing structure obtained from fractal branch and bound method, the base fiber angle optimization method using Kriging method is developed. As a result, the method is successfully applied, and the optimal base fiber angle is obtained using newly proposed base fiber angle optimization method.

View full abstract

This study treats the optimization problem of material distributions for minimizing the stress intensity factor of radial edge crack in thick-walled functionally graded material (FGM) circular pipes under thermomechanical loading. Homogenizing the FGM circular pipes by simulating the nonhomogeneity of thermal conductivity by a distribution of equivalent eigentemperature gradient and the nonhomogeneities of Young's modulus and Poisson's ratio by a distribution of equivalent eigenstrain, we present an approximation method to obtain the stress intensity factor for a radial edge crack in the FGM circular pipes. The optimum material distributions for minimizing the stress intensity factor of edge crack are determined using a nonlinear mathematical programming method. Numerical results obtained for a thick-walled TiC/Al₂O₃ FGM circular pipe reveal that it is possible to minimize the stress intensity factor of edge crack by choosing an optimum material distribution profile.

View full abstract

The influences of microstructure and specimen size on creep properties of Sn-37Pb and Sn-3.5 Ag solders were investigated by nano-indentation tests and tensile creep tests for fine wire solders. The nano-indentation tests enabled us to measure the indentation creep properties of the specific microstructures individually. Tensile creep experiments were performed on our original creep testing machine for fine wire solders with the diameters of 0.2, 0.5 and 1.0 mm. For the Sn-37Pb solder, the creep deformation properties were uniform for both α and β phases and grain boundary sliding was the dominating creep mechanism. For the Sn-3.5Ag solder, creep deformations of α phase were observed under much smaller stress levels than those for ε phase, and thus the creep deformations occurred only in α phase, which resulted in the nucleation of microvoids around the ε phase. The tensile creep tests for wire solders revealed that the significance of specimen size on the creep properties is strongly influenced by the microstructure of the solders.

View full abstract

In order to clarify the complex effects of surface roughness, transition of microstructures from surface to interior, defects and residual stress, tension-compression fatigue tests have been conducted by using shot blasted ferrite-pearlitic ductile cast irons with as-cast surfaces. Regardless of applied stress levels, the fracture origins of specimens were mostly at a defect near an as-cast surface, slag or pinhole, at N_f<10⁷. At a relatively lower stress amplitude, some fatigue fractures occurred from an inner defect far from the surface, shrinkage, mostly at N_f>10⁷. The √area parameter model was applied to the quantitative evaluation of fatigue limit. The equivalent hardness considering transitional layers and the effective defect size with the interaction between surface roughness and a defect were defined. Moreover, the relief of residual stress during fatigue tests was detected. The mean value of the relieved residual stress acting on a defect was employed for the quantitative evaluation. By using these parameters, the complex effects of ductile cast irons have been evaluated successfully for practical use.

View full abstract

In this paper, a quantitative prediction method for the fatigue limit reliability of metal with different sorts of inhomogenities is proposed. The method is based on the stress-strength model which consists of “statistical characteristics of hardness in a small region” and “statistical characteristics of a mechanical condition in the small region containing the defect under a cyclic stress”. The method was applied to the aluminum cast alloy JIS AC 4 B-T6, which has eutectic Si, Fe compound and porosity in the material as inhomogenities. The fatigue strength at 10⁷ stress cycles was defined as the fatigue limit. Rotating bending fatigue tests of plain specimens were also caried out. The validity of the present method was confirmed by comparing analyzed results with experimental results.

View full abstract

Thermal tempered glass is made by rapidly cooling after heating float glass to near the softening point. Compressive residual stresses at the surface layer of thermally tempered glass strengthen the material, and tensile residual stresses are also generated at the same time in the interior of the material to balance with the surface compressive residual stresses. It is considered that the fatigue crack growth characteristics of thermally tempered glass are remarkably different from that of float glass, due to the existence of those residual stresses. In this study, Au film made by ion sputtering method and an optical microscope are used to measure the crack length of the thermally tempered glass and float glass, and these fatigue crack growth characteristics are investigated. As a result, it is confirmed that the crack length measurement with an Au film has high accuracy of measurement. Furthermore, it is found that the higher stress intensity factor is needed to cause the crack growth in thermally tempered glas than in flat glass.

View full abstract

The surface of silicon nitride ceramics was modified by fine particles bombarding (FPB) of high strength steel with 50 μm in diameter and by ultrasonic shot peening (USP) of tungsten carbide shots with 1.1 mm in diameter. The residual stress in the sub-surface region measured by the X-ray diffraction method was compression. For FPB surfaces, the sin² ψ diagram was nonlinear, indicating a steep gradient of the residual stress distribution in the sub-surface region. The distribution of the residual stress was determined by the constant-penetration-depth method. The stress on the surface was a large compression of about -1.5 GPa and diminished 5 μm below the surface. The compressive stress increased with the shot pressure for FPB. For USP surfaces, the compressive stress was nearly constant with in the layer of 10 μm and increased up to -1.5 GPa with the peening time. The half value breadth of diffraction profiles increased nearly proportionally with the magnitude of the compressive residual stress. The fracture toughness measured by the Vickers indentation method also increased with the compressive residual stress.

View full abstract

The misorientation introduced into an austenitic stainless steel (SUS 316) by plastic deformation was masured with the electron back scattering diffraction (EBSD) and the X-ray diffraction (XRD) techniques. Three parameters related to miorientation distribution, i.e. grain orientation spread (GOS), grain average misorientation (GAM) and kernel average misorientation (KAM), were calculated from the orientation measurements by EBSD. The misorienation between neighboring measured points, KAM, was larger close to the grain boundary where the slip propagation was difficult. The average of misorientation between neighboring points within one grain is GAM. The GAM value averaged over the measured reagion increased proportionally with the plastic strain to about 20%, and leveled off at higher strains. The average of the misorientation spread within one grain, GOS, also showed similar increase like GAM. The total misorientation of a grain measured by XRD also showed a proportional increase with the plastic strain to about 10%, and it was difficult to measure the total misorientation above that strain because Debye ring became continuous ring. Both EBSD and XRD can be used as a tool to measure the plastic strain in the local area, and the limit of strain to be measured was 30% for the former and 10% for the latter.

View full abstract

This paper is concerned with the development of a shock tube suited for nondestructive inspection in concrete structures. The developed shock tube produces shock waves by combustion of propane-air mixture gas in a long cylindrical tube at short intervals. The propane and air mixture are injected into the reaction chamber from the mixing chamber and ignited. A detonation wave goes through the expansion chamber and extends into the atmosphere. The performance of the tube was evaluated by measuring the pressure distribution on a specimen surface and experiments of defect detection for concrete specimens. A triangular wave with a pressure of 4 kPa and a duration of 0.4 ms was obtained at 500 mm distance from the shock tube. In the defect detection, we could detect a defect with a diameter of 200 mm and a depth of 50 mm clearly.

View full abstract