Free Mesh Method, which is a kind of meshless method, is simple, accurate and suitable for parallel computing. In this method, the global stiffness matrix is obtained by adding the temporary element matrices, so this method does not require connectivity among nodes and elements as an input infomation. This paper describes an application of Free Mesh Method to 3-dimensional solid model. In this study, new definition of analysis model by using CAD modeling data is introduced for Free Mesh Method, and treatment of location test in which the element is interior or exterior for boundary in local mesh is described. Besides, this approach is applied to 3-dimension solid model and it is clarified that this treatment is efficient for Free Mesh Method.
We present estimation methodology for market information with Bid-Ask spread. The Bid-Ask spread has been explained as the result from overall market players’ intention in market microstructure theory. We focus on Bid-Ask spread transition after market opening, and interpret it caused by information effect. Our approach is based on modeling methodology for viscoelastic material that can consider memory effect by information in market system. Under the modeling process, we suggest new parameter that related to response effect and relaxation time effect of information in financial market. Lastly we show the model explain Bid-Ask spread transition and try to estimate market information in Japanese equity market empirically.
The slack variable method was introduced into Receding Horizon Control(1)(2)(4) with equal constraint, and formulization of a legged type robot’s swing leg was described. In formula Receding Horizon Control with equal constraint, such as having been proposed with bibliography(2), ZMP(3) conditions were described as equal constraint, and the whole robot center of gravity was formulized. Then, formulization of swing leg was added to this in this paper. Generally, the conditions on which swing leg does not fall from the floor altitude can be described as inequality state constraint. It is difficult to treat this amount states constraint, because input variable for this state constraint can not be analyzed. In this paper, it proposed dealing this with a problem using slack variable. This simple formulization can be applied to varied legged robot with any number of legs.
It is rational to discretize the microstructures that consist of thin walls like cellular solids by thin-walled structural elements such as frame or shell elements from computational viewpoint. In this context, authors tried to apply the structural elements to the micro-scale problems and pointed out the problem due to inconsistency between assumed kinematic field of the structural elements and the two-scale displacement field. In order to avoid this inconsistency, we propose the formulation that employs not the micro-scale displacement but the total ones for the freedom of micro-scale problems. Some numerical examples of linear problems show not only the advantage on the computational cost-efficiency of the present method but also the possibility of independent modeling of micro-scale problems.
Infinitely many windows are realized in the parameter interval of a one dimensional iterated map. The parameter values corresponding to the super stable periodic trajectories are calculated by the supertrack. The distributions of the calculated parameters of various periodicities exhibit the scaling relation. We prove this scaling relation from the supertrack function.
The authors presented previously a buckling hypothesis on the etiology of the idiopathic scoliosis by finite element analyses with a finite element model of spine and rib cage. The hypothesis says that the growth of the vertebral bodies induced buckling modes and that the fourth and the sixth buckling modes similar to the clinical modes of the single and the double curves were remained as uncontrollable modes by posture change. This paper investigates a factor of the superior growth of the vertebral bodies in the modeling behavior of the vertebral bodies against the gravity of the head and the trunk. The distribution of the strain energy density under the gravity was concentrated on the front parts of the thoracic vertebrae. The fourth buckling mode induced by the growth in proportion to the strain energy density bore a similar mode to the single curve of clinical idiopathic scoliosis.
In the existing regression model for the fuzzy linear function, fuzzy parameters are derived by solving a linear programming problem so that the fuzzy dependent variable or fuzzy output contains the estimated fuzzy output with more than a certain predetermined degree while the total fuzziness of the estimated outputs becomes minimal in the so-called minimization formulation. It is easily shown that the above mentioned formulation of the problem falls through when the width of the output shrinks away, in other words, the output data is non-fuzzy. Thus, it is clear that the existing model does not include the non-fuzzy linear regression model as its extreme case. In this study proposed is a new fuzzy linear regression model based on the minimization criterion of the sum of squared distance(Euclidean distance over the center and width space) between the observed fuzzy output and its estimate The model is reduced to the usual non-fuzzy regression model when the width of the output variable vanishes. Numerical test results demonstrate that the proposed model substantially outperforms the existing model in both estimating output center and output width.
In many mechanical structures, structural design for noise reduction is becoming increasingly important. Noise reduction is often achieved through structural modifications. However, it is hard to predict the effectiveness of noise reduction by typical approaches. This paper presents an optimal design approach for reducing sound power from a vibrating plate by its curvature design. The method couples an optimization technique based on a genetic algorithm (GA) with the shape representation technique, vibration analysis and acoustic radiation analysis. It is shown that the curvature design of the plate obtained by using this method can achieve effective reductions in radiated sound power. The robustness of the optimized design candidates obtained is needed to study by using stochastic simulations. We present a method that applies GA to the choice of robust design candidates. It is shown that the presented method is successfully implemented for this study.
This report describes a newly developed cascade impacter pollen collection device. We improved on the design of an existing cascade impacter using a numerical simulation method and performed an adhesion experiment using alder pollen. The results are shown to be in close agreement with the results predicted by a numerical analysis based on the Navier-Stokes equation and the Euler-Lagrange method. Based on the Euler-Lagrange method, we analysed the separation of four typical Japanese pollens under various conditions of air suction velocity and nozzle diameter. Pollen grains are collected in four stages of the device. To separate pollens adhered to the same plate, we developed an automatic pollen measuring system, which allowed pollen grains to be completely separated.
Fully developed three-dimensional turbulent flow over sand ribbons in an open-channel is computed to investigate the coherent structure of streamwise vortices by direct numerical simulation (DNS) using a regular grid under a generalized curvilinear coordinate system. Coherent structure are educed from the simulated instantaneous flow field using a conditional sampling technique developed by Hunt et al, Jeong & Hussain which extracts the entire extent of dominant vortical structures. The results showed that stable secondary currents and turbulent characteristics were reasonably reproduced, highly elongated quasi-streamwise vortices were identified and the streamwise vortices of alternating sign were recognized to stagger over the sand ridge.
The conjugate projected gradient method is applied for solving large-scale contact problems. The finite element discretization of contact problems results in linear algebraic equations with a set of linear multipoint constraints. When the penalty method is adopted for incorporating the constraints, the linear equations without the constraints are derived. However, they are ill-conditioned and difficult to be solved by using the conjugate gradient method with a simple preconditioner such as the diagonal scaling. In the present study, the constraints due to the contact are formulated by using the Lagrange multiplier method, and the linear equations are solved by the conjugate projected gradient method. The projection matrices for the node-to-node and master-slave contact strategies are derived. The performance study of the algorithm is conducted by solving linear contact problems with 0.2-0.7 million degrees of freedom. The latter problem can be solved within 768 seconds by using one Alpha 21164 (433MHz) processor with 448MB memory.
Numerical analysis of natural convection flows in a 8:1 differentially heated square cavity have been carried out. Objectives of this study is to investigate the flows precisely with the Rayleigh number equals to 3.5×105 and to find out the critical Rayleigh number at which the steady flow changes to the flow with oscillation by means of parameter study for Rayleigh numbers. The GSMAC finite element method was applied to calculate the flow field and the CIP finite element method was used to solve the advection phase of the energy equation. Various time history data were obtained and periods and amplitudes of the data are calculated using FFT. Obtained data are good agreement with other calculated data. Parameter study for Rayleigh numbers was carried out and relationship between amplitude of velocity and temperature and Rayleigh numbers were calculated. The critical Rayleigh number with this scheme is approximated 2.9×105.
Taylor vortex flow between two-concentric accelerated cylinders with finite length is investigated numerically. The outer cylinder wall and the upper and lower end walls of the cylinders are stationary, and the inner cylinder begins to rotate with constant accelerations. One of the characters of Taylor vortex flow is its non-uniqueness. The non-uniqueness means that various modes appear, even when the aspect ratio and the Reynolds number are fixed. In this paper, the non-uniqueness of Taylor vortex flow is investigated at wide range of the aspect ratio, Reynolds number and acceleration time of the inner cylinder. To analyze the mode formation processes, the expert system that detects the mode of the flow quantitatively is established. Each mode formation process is not simple but multiple. Some flows develop to final modes via intermediate modes. The dependency of the mode formation processes on the acceleration time of the inner cylinder is complex.
This paper describes the application of Trefftz method to the steady-state heat conduction problem on the functionally gradient materials. Since the governing equation is given as the Poisson equation, it is difficult to apply the ordinary Trefftz method to this problem. For overcoming this difficulty, we will present the combination scheme of the Trefftz method with the computing point analysis method. The non-homogeneous term of the Poisson equation is approximated by the polynomials of the Cartesian coordinates to determine the particular solution related to the non-homogeneous term. The solution of the problem is approximated with the linear combination of the particular solution and the T-complete functions of the Laplace equation. The unknown parameters are determined so that the approximate solutions will satisfy the boundary conditions by means of the collocation method. Finally, the scheme is applied to some numerical examples.
We have developed a new program named Parallel Molecular Dynamics Stencil, which is an integrated assembly of simulation techniques for executing molecular dynamics in a parallel computing environment. The Stencil is designed in such a way to conceal parts of the programs for parallelization. With the Stencil’s framework, parallel programming for force calculation can be done in the same way as sequential programming. So, it can be easily adapted to employed physical models. It is shown that the performance of the Stencil is excellent by choosing the combination of the spatial decomposition and the cell partitioning methods.
Simulators based on agent oriented system have been developed in the field of engineering. However, multiagent system (numerous agents exist in an analytical region) needs a lot of computer resources to simulate in real time. This paper proposes one of parallel and distributed platforms for multiagent oriented system by Java language. Our platform consists of distributed multiagent platform implementation and it provides API (Application Programming Interface), so users easily develop multiagent system using this API. Moreover, users do not need to consider how to use parallel and distributed computing environment because our platform controls these processors by automatic load balancing agent. The effectiveness of proposed platform is confirmed in two cases of application, namely a calculation of pi compared with MPI system and the automatic mesh generation by voxel element. Our platform is no less rapid than MPI and is suitable for parallel and distributed multiagent system.
This paper describes an application of a genetic programming to local contact search. Contact search process consists of two phases: a global search phase for finding the nearest node-segment pair and a local search phase for finding an exact local coordinate of the contact point within the segment. In this paper, the genetic programming is utilized for finding the function which approximates local coordinate of the contact point in the local search phase. The fundamental formulation of this method is described in detail, and its basic performance is demonstrated through some analyses.
This paper presents an optimization method of production efficiency about the placement time in the chip mounter system under one constraint. This constraint is that an operator specifies a chip mounter which has minimum placement time. Placement time changes greatly by the performance of the chip mounter and the configuration of the system. The problem is how to distribute parts and parts feeders to each chip mounter. We propose a distributing method by GA adopting the penalty method. Several numerical experiments about some systems composed of more than one chip mounter with different performance were done. It was confirmed that our proposed method had validity more than conventional method through the results of numerical experiment. Furthermore, those characteristics and validity are explained about the GA adopting simple penalty method (SPM-GA) and the GA adopting improved penalty method (IPM-GA) that a penalty is changed one after another with the progress of the generation.
In this paper, the authors present a new evolutionary optimization method for three-dimensional structural statics. The method consists of two optimization processes in series. The former process is to deal with the topology optimization. The latter process is to deal with the configuration optimization. The method finds an approximate optimum topology and configuration of the structures from any initial design of them: That is, an optimum structure design can be obtained from a simple and small initial structure. Therefore, less computational cost and time will be required to obtain results by the method than those by conventional ones.
There are methodologies that use a physical scale and methodologies that use a psychological scale to evaluate urban landscapes. Summary data has been gained from a questionnaire, which asked whether views should be regarded as objects and therefore measured directly. This method does not provide a direct subjective evaluation but it is effective in providing objective evaluations. For the whole system to be understood it is important to know how the view is seen through the eye of the observer. This is because it minimizes information loss from summarized data error. This report determines the most appropriate system for each observed city view, by using alpha wave and beta wave data, gained from taking an electroencephalogram from each person tested. It shows the method that best measures whether the stated condition of Self-Organized Criticality is relevant, based on the electroencephalogram data generated and also if there is Fractal structure in the system.
One of the major concerns in the structural integrity of aged-welded structures is to evaluate residual stresses nondestructively. Residual stresses are generated by eigenstrains in thevicinity of weld bead. It means that eigenstrains are the source of residual stress generation. If eigenstrains are known experimentally, residual stresses can be estimated by a simple elastic FEM analysis. Based on this concept, authors have proposed a new method, the bead flush method. The purpose of the present study is to simplify the bead flush method by partially removing weld reinforcement. To improve accuracy of the conventional method, one of the regularization methods, SVD and L-curve are adopted to regularize the solution in the inverse analysis. The proposed partial removal method has following advantages comparing to the conventional one. The proposed method saves inspection times and costs because only a part of weld reinforcement is removed to measure the released strains. Secondly, residual stresses of structures at the same position can be evaluated periodically.
A structural performance evaluation rule was settled in the Japanese Building Standard Law. The new rule evaluates structural seismic safeties by importance factor. Structural designers must explain costs and performances of buildings to customers, and prepare a new method for such new requests of customers. This report proposes a support system for planning reinforced concrete (RC) structures according to structural performances by neural network and spline function. Neural network is modeled after the neuron of human brains, and has been applied to structural control and planning by a couple of co-authors of this paper and others. In this paper, it is shown that at the stage of planning, we can obtain useful information on life cycle cost of RC structural skeletons by this system.
International exchanges among engineers in Asian countries or the world are expected to increase by information communication technology. However they may meet a problem of communication language in order to understand each other: one engineer/scientist has obtained knowledge in some specific areas in one language, for example, in his/her own mother tongue, and he/she wants to access information described in another language. In this case he/she meets difficulties of find correct words in the other language corresponding to the words that he/she knows in one language. Therefore a multi-language capable translation system is expected for the communication. We propose a term-based translation support system BABEL that has a multilingual dictionary and a hypertext formation system in order to overcome this barrier. Most translation support systems translate sentences and have difficulties to generate natural sentences after translation into a target language. In fact, this brings a serious obstacle to users. In BABEL, technical terms in sentences are extracted automatically and translated into the intermediate language, and are linked with the BABEL dictionary technical terms. At the same time the technical terms linked to BABEL’s dictionary are attached with HTML tags. Finally the BABEL presents the translated words in the target language.
Ingot annealing is indispensable process for InP single crystal to improve its electric characteristics. One of the technical problems of InP ingot annealing is the increase of dislocation density during its annealing process that affects the performance of electronic devices. A computer code was developed for dislocation density evaluation of a single crystal ingot during annealing process. A dislocation kinetics model called the Haasen-Sumino model was used as a constitutive equation. In this model, creep strain rate is related to the dislocation density, and this model was extended to the multiaxial stress state based on the theory of crystal plasticity. Three-dimensional finite element model was used to take account of crystal anisotropy in elastic constants and specific slip directions. Dislocation density analyses were performed for an InP ingot with 4-inch diameter, and time variations of dislocation density and stress were obtained from this computer code.
A numerical simulation of an electrohydrodynamic flow is performed. The purpose of this analysis is to establish a scheme for accurate and robust analyses of electrohydrodynamic problems. The method used here is a level set method based on the finite element method. In our problem we have two material derivatives, one in the Navier-Stokes equations, the other in the advection equation for the interface of two fluids, air and a liquid. We apply a first order characteristic method to approximate these material derivatives, where an integral including upstream information is calculated with an accurate computation. Some results including the surface tension effect are presented comparing with the Lagrangian method that the author previously obtained.
I proposed the finite element formulation using SO(3) and Linear stress rate before. The tangent stiffness using this stress rate is much simpler and easier to get for the beam and the shell elements than any others. Then, the new membrane element is developed according to this formulation. It has the strong point that there are 3 rotational degrees of freedom at each grid point. So, the elimination of the in-plane rotational degree of freedom is not necessary when 3-dimensional analysis is executed. In this paper, the theory limited to linear stiffness is shown and the results of benchmark tests are shown.
It is well known that the vector finite element method is one of the powerful tools for solving electromagnetic problems. The vector shape functions which consist of the facet and the edge functions have a lot of characteristics. One of them is automatic conservation of magnetic flux density in analyzing the Induction equations without iterative correction. In this present paper the vector finite element method is applied to the problems of magnetohydrodynamics. The computational scheme is verified by analyzing the flow between parallel plates under a constant magnetic field. Two-dimensional natural convection in a cavity under a constant magnetic field is analyzed numerically. computational results are in good agreement with other numerical results.
The purpose of this study is to develop a procedure for performing a dynamic analysis in the case that a geometrically nonlinear shell-structure in three dimensions moves along a nonlinear trajectory at variable velocity. The finite-element equations of motion were derived in the case that the shell-structure undergoes large translational and rotational displacements in three dimensions. A computational code for time history response analysis of shell structures was then developed. The validity of the formulations and the developed computational code was verified by the comparing the numerical solutions obtained using the developed computational code with our experimental results of the static bending and free vibration of the cantilever plate. Then a comparison was made between the calculated results and the experimental ones in the case that the clamped end of the cantilever plate moved along the trajectory of an arc.