The Free Method (hereinafter referred to as "FMM") has a good compatibility with the parallel computing. But, the conventional FMM has a problem in accuracy. Recently, the Enriched Free Mesh Method (hereinafter referred to as "EFMM") was proposed by Yagawa et al. This paper describes the parallel EFMM, intending the large scale analysis with high parallel performance. Saving of times is attained by parallel Conjugate Gradient method. The parallel EFMM code is developed and employed on a cluster system.

View full abstract

In this paper, the effect of radial processing corrugation on the axial crushing behavior of square tubes is studied by using the finite element method. The corrugations are one of the dominant factors in influencing the deformation mode and they improve the energy absorbing capacity of the square tube. The numerical results show that the average stresses of square tubes with corner (STC) are affected by the shape of corrugations and those intervals. The average stresses of STC, however, are same with no dependence of the width when the shape and interval of corrugations are identical. Furthermore, the numerical results of the model predict two different modes, termed the Extensible and the Inextensible modes corresponding to the aspect ratio. In addition, the average stress predicted by the proposed method using a master curve and approximate equations for the simple square tube is found to be good agreement with results by FEM.

View full abstract

In this paper, we derive the solution for two circular holes or rigid inclusions perfectly bonded to an elastic medium (matrix) of infinite extent, under In-Plane deformation. The two holes or rigid inclusions have different radii and different central points. The matrix is subjected to arbitrary loading like uniform stresses as well as a concentrated force at an arbitrary point. The solution is obtained, via iterations of Mobius transformation as a series with an explicit general term involving the complex potential of the corresponding homogeneous problem. This procedure has been termed "heterogenization". Using these solutions, several numerical examples are shown by the graphical representation.

View full abstract

This study was conducted to investigate damage condition of glass-fiber reinforced plastic (GFRP) which was immersed in hot water for a long time. Furthermore, we aimed to establish a non-destructive method based on ultra-sonic echo data for estimating the damage condition and strength of GFRP. For those purposes, we made three kind of GFRPs in which base resins (isophthalic-type UP, bisphenol-type VE and novolak-type VE) possessed different corrosion resistance, and then immersed them in hot water maintained at 368K for 2.6-31.1Ms (1-12 months). The GFRP, whose base resin was isophthalic-type UP possessing relatively low corrosion resistance, suffered to remarkable damage such as interfacial debonding between glass-fibers and the base resin, delamination of glass-fiber cloths and cracking of glass-fibers, so that the bending strength was rapidly reduced. On the other hand, the damage condition in the other GFRPs was not striking because of the base resins having higher corrosion resistance, and their bending strengths were maintained at relatively high levels. Moreover, since the damage condition of the GFRPs was reflected to the ultra-sonic echo data, the bending strengths could be roughly estimated by such data.

View full abstract

Shape memory alloy has been attracting because of its unique characteristics such as the shape memory effect and the superelasticity. These behaviors arise from reversible movement of atoms during the martensitic transformation driven by thermomechanical loadings. In the present paper, the molecular dynamics simulation of stress-induced martensitic transformation in a NiAl shape memory alloy is carried out focusing on the relation between the evolution of martensite variant and resulting stress-strain behavior. In order to simulate martensitic transformations and to observe stress-strain relation under combined stress state the Parrinello-Rahman method with periodic boundary conditions is used. Microscopic martensite variants are classified by the relative displacement and the angle between transformation shear planes which is identified by change in lattice parameters. In the present simulation, there exist three types of martensite variant depending on the loading direction. The transformation stress varies with the loading direction. Martensitic transformation under simple shear starts and finishes at higher stress level compared to tension and compression. Martensite finish stress is higher in compression than in tension while there is no significant difference in martensite start stress.

View full abstract

This paper discusses the influence of irreversible hydrogen and low temperature annealing on the fatigue strength in a cold drawn high strength steel. Fatigue tests were conducted for as-received and annealed samples. Samples were cathodically hydrogen charged, and internal hydrogen states were changed as follows: (a) virgin sample and (b) the sample that contained only irreversible hydrogen. In the case of an as-received sample, irreversible hydrogen did not influence the. fatigue strength. However, the fatigue strength of an annealed sample was decreased by irreversible hydrogen. The activation energy which irreversible hydrogen requires to desorb from its trap site was unchanged by low temperature annealing. Slow strain rate tensile tests showed that the sensitivity to hydrogen embrittlement was increased by low temperature annealing. In the annealed sample, the fine carbide precipitates in ferrite lamella, which was observed by transmission electron microscope, were considered to induce sensitivity to hydrogen embrittlement, and thereby caused a decrease in the fatigue strength of the annealed sample by irreversible hydrogen.

View full abstract

Isothermal low-cycle fatigue tests and thermal-mechanical fatigue tests with linear out-of phase cycles were carried out using cylindrical specimens made of Ni-base superalloy and coated with several kinds of corrosion-resistant coatings. In some tests, fatigue loading was interrupted several times and coating cracks were observed by means of replica technique. As a result, three types of coating cracks were observed, namely, multiple parallel long cracks, multiple parallel short cracks and non-parallel cracks. Multiple parallel long cracks are caused from lack of ductility of coating materials. This type of coating cracks initiate in an early stage of fatigue loading and has a strong effect to reduce fatigue life. Multiple parallel short cracks are caused by fatigue of coating layer. These cracks propagate into the substrate before cracks initiate in the substrate and reduce fatigue life of the coated specimen. Non-parallel coating cracks are caused by propagation of cracks initiated in the substrate and have little effect to reduce fatigue life.

View full abstract

In order to investigate the importance of pore size and hardness on the fatigue strength of a pre-alloyed powder metal (Fe-3.0 Cr-0.5 Mo-0.5 C mass%) with the density of ρ=7.0g/cm^3, ultrasonic fatigue tests were conducted at a test frequency of 20kHz in air at room temperature. The fatigue limit of the powder metal was the maximum critical stress under which fatigue cracks initiating from the pore stop propagating. Therefore, the fatigue limit was accurately predicted by the √<area> parameter model, σ'_ω=1.43(HV+120)/(√<area>)^<1/6>, for the surface defect. Here, as √<area> was employed the effective maximum pore size,√<area_<eff, max>>, in the control volume of the ultrasonic fatigue specimen which was estimated from the statistics of extremes distribution of the pore sizes at the polished surface. As HV was employed Vickers hardness, HV_<98.1N>, of the microstructure which included both the matrix and pores. The usage of √<area_<eff, max>> and HV_<98.1N> in the √<area> parameter model is practical use and useful for predicting the fatigue limit of the powder metal. In addition, the usage of √<area_<eff, max>> indicates that the dominating cause of fatigue strength for the powder metal is the maximum pore size rather than the porosity or density.

View full abstract

In order to clarify the effect of hydrogen on Mode II fatigue behavior, fatigue tests were conducted using hydrogen-precharged and uncharged specimens of bearing steel (SUJ 2) tempered at 438K loaded under cyclic torsion with and without a compressive axial mean stress (σ_m=0, -750MPa). The fatigue lives in the hydrogen-precharged specimens were shorter than those in the uncharged specimens under cyclic torsion with and without a compressive axial mean stress. In both hydrogen-precharged and uncharged specimens, Mode II fatigue cracks initiated and propagated along the axial direction of specimen due to the longitudinal microstructure texture formed in the rolling direction, and finally branched into two Mode I cracks. In the case of hydrogen-precharged specimens under cyclic torsion without a compressive axial mean stress, however, Mode I crack initiated and propagated directly from subsurface inclusion in two specimens and Mode II crack initiated and propagated from the matrix in one specimen. Mode II fatigue crack growth rate and Mode I fatigue crack growth rate of the hydrogen-precharged specimen were higher than those of uncharged specimen under cyclic torsion with a compressive axial mean stress. More mode II fatigue cracks were observed at the final stage (N/N_f=0.97) in the hydrogen-precharged specimen than in the uncharged specimen. Microstructural changes caused by the shear slip deformation were formed in the hydrogen-precharged specimens. From these observations, it is presumed that hydrogen enhances slip deformation and localized slip bands, and eventually produces more mode II fatigue cracks along slip bands in the hydrogen-precharged specimens.

View full abstract

We have already proposed a generation laser scanning method for visualizing ultrasonic waves propagating on 3-D objects. This method has the excellent feature that enables a visible and quick flaw inspection for actual object. However it is disadvantageous in that defect echoes are often hidden in the forward traveling wave because this method visualizes all waves propagating on the object. In this research we investigated the synchronizing differential method to extract defect echoes from the visualized images. The ultrasonic tests with aluminum plate and pipe having slits on the rear surface proved the validity of the synchronizing differential method, by which, we could visualize the slit echoes hidden in the forward traveling wave. These results demonstrated the usefulness of the laser-based ultrasonic visualization technique in flaw inspection.

View full abstract

Three methods of an area detector type multiaxial stress measurement were studied in this paper. The first method, which was proposed by Sasaki and Hirose, gives us correct results when precise diffraction data are obtained and are used in the stress calculation. The error of the stress determination sometimes occurs in this method when the diffraction data contains the error of measurement. In order to avoid such an error, the second method was developed in this study by improving the first method. The second method was also found to be correct in this paper and to be effective within practical use. The third method was also proposed to decrease the number of diffraction data needed in the stress calculation. The evaluation on the accuracy and the comparison of these three methods were carried out by conducting a numerical simulation and an experiment. It was found that the two methods proposed in this paper can contribute to make progress in the reliability of the multiaxial stress evaluation by the area detector type X-ray method.

View full abstract

Friction-stir spot welding (FSSW) has been performed using a newly designed scrolled groove shoulder tool without probe in aluminium alloy. The effect of welding parameter on the welding structure and tensile-shear strength of welds was evaluated and the possibility of the scrolled groove shoulder tool in FSSW was discussed. When the scrolled groove shoulder tool was used, the stir zone attained a depth of 2mm, while when the plain tool without scroll groove was employed, no stirring occurred in the depth direction, thus it was confirmed that the scroll groove formed on the shoulder surface was very useful in FSSW. Tensile-shear strength was found to be dependent on the welding parameters such as tool rotational speed, shoulder plunge depth and tool holding time and was considerably higher than in the welds joined by the plain tool. The maximum strength of 4.6kN was obtained under the condition of a tool rotational speed of 3000rpm, a shoulder plunge depth of 0.7mm and a tool holding time of 4sec. Two different fracture modes, shear fracture and plug fracture, were observed depending on shoulder plunge depth and tool holding time at a fixed tool rotational speed. Shear fracture occurred at a less shoulder plunge depth or at short tool holding times, while plug fracture took place at long tool holding times or at large shoulder plunge depths.

View full abstract

From a traffic safety point-of-view, there is an urgent need for intelligent tires equipped with strain sensors as a warning system for road conditions and for optimized braking control on poor road surfaces. However, since a conventional foil strain gage has high stiffness, it causes the analyzed region to behave unnaturally. The present study proposes a novel rubber-based strain sensor fabricated using photolithography. The rubber base has the same mechanical properties as the tire surface; thereby the sensor does not interfere with the tire deformation and can accurately monitor the behavior of the tire. This investigation details the design and manufacture of the rubber-based sensor. For verification, the sensor was attached to a beam specimen and found to have a fatigue life of over 10^6 and a measurable strain range up to 14%. The sensor was applied to a commercially available automobile tire and found to successfully measure the strain of the tire.

View full abstract

Fiber waviness is one of initial defects during processing such as filament winding. The fiber waviness causes reduction of compressive strength of the composite structure. The mechanism of growing of the fiber waviness is, however, not completely clear. In the present study, a new model of generating fiber waviness is proposed. It is assumed that the local fiber micro buckling causes the fiber waviness. The micro buckling is based on the compression load caused by shrinkage of a metal jig. Three processing faults are considered as causes of this micro buckling: bonding between a metal jig and a composite (poor release), insufficient cure of the resin (non-uniform temperature) and initial deflection of fibers (poor tension during FW process). The model predicts fiber micro buckling during cooling process after cure. In the present study, to simplify the experimental conditions, prepreg sheet is adopted instead of FW process and the fiber micro buckling predicted using the model is compared to the experimental results. As a result, the new model successfuly predicts the initiation of the fiber micro buckling during cooling process.

View full abstract

Due to the nonlinear viscoelasticity observed in moisture solids or polymer materials, it is generally difficult to analyze their deformation problems because the analysis requires complex constitutive equations. A typical analysis technique that employs a generalized viscoelastic model involves considerable amount of effort in the evaluation of material parameters, and it is difficult to apply the analysis technique to finite element simulation. On the other hand, fundamental constitutive models such as those proposed by R. Hooke, C. Maxwell and W. Voigt are useful in the simulation of various industrial problems. In this study, the technique for the evaluation of the nonlinear parameters in the three-element solid model, one of the fundamental models, is presented and used to analyze viscoelastic problems. The technique uses the simple tensile tests for 3 different strain rates. In one case, the strain rate is very slow and viscosity can be neglected and, while in the other two cases viscocity is not negligible and is evaluated from the strain rates. By the sequential application of the technique from a low stress level to a higher level, the nonlinear parameters of the three-element solid model, which can be defined to depend on state quantities, can be evaluated. The reliability of the technique is also examined by studying its application to the evaluation of the mechanical behavior of biological soft tissues, which are of moisture solids. In addition to the fundamental procedure of the evaluation presented in former section, the modified procedure is also shown to evaluate more precise parameters by improving the experimental condition.

View full abstract

Dia-Core is a newly devised core panel formed by gluing/welding of two same shaped panel pieces which have periodically indents. The basic model of Dia-core is based on Octet-Truss developed by Fuller, and it consists of tetrahedra and octahedra. It has good cost performance due to easy press forming. Therefore, it is expected to become new core materials which can compete with honeycomb panels from the total points of view. In this paper, by using experiments and numerical simulations, the bending rigidity, a representative propertie of Dia-Core is clarified. The results suggest that the bending rigidity of Dia-core is proportional to the initial thickness of panel pieces and the square of the panel height. These trends can be explained by the application of the theory of sandwich structures. And by using these results, the specific rigidity of Dia-core is calculated and compared with that of two typical honeycomb panels.

View full abstract

Volumetric examination range is defined to cover the wall thickness for welds of category B-F and B-J in current JSME Fitness-for-Service (FFS) Code. Since it was not reasonable for examiners to have higher radiation dose due to meaningless data such as configuration echoes, a rule change of the volumetric examination range to 1/3 thickness of inner surface side from the whole thickness was proposed. By this modification in the examination range, advanced inspection techniques such as phased-array ultrasonic test would be applicable. Through the survery of various report, it was shown that the welds of these categories had not been suffered from any trouble which initiated from embedded flaw. Furthermore, the validity of the modification was also confirmed by showing the allowable flaw size for plant fabrication was smaller than that of the JSME FFS Code. Namely, any serious flaw is not initiated from inside the body during the plant operation, or, even if a flaw exists in plant fabrication, the size of the flaw is small enough from view point of the integrity assessment.

View full abstract