In this research we use locally-coupled analysis to evaluate the damage and fracture behaviors of sprayed coatings. We conduct analysis without damage at first (in this case we use ANSYS). Then the results of strain will be input in the damage analysis program which uses the elasto-viscoplastic damage constitutive equation based on continuum damage mechanics. Both of the timing of damage occurrence and the relations of load-displacement have almost agreed well with the three-point bending experimental results. The locally-coupled analysis employing damage mechanics models can be expected to be used as a computational tool for the evaluation and the design of sprayed coatings.

View full abstract

  To reduce CO₂ emission, control region for automotive engine systems are near the boundary between normal and abnormal engine operations such as knocking and misfiring. In order to realize near boundary operation control, it has become essential to accurately identify the boundary between the operable region and the abnormal operation region with a numerical model.
  In this paper, we propose a dynamic model structure based on the Nonlinear Auto-Regressive with eXogenous inputs (NARX) model as a control model of the automotive engine systems. This model is learned using a deep learning or a Gaussian process regression. Furthermore, we clarify relationships between the number of learning data, prediction/estimation accuracy and learning time by using numerical simulations. This approach is useful for deciding how to select a model learning method with respect to the number of training data. As a result, when the number of learning data is small, the model based on the Gaussian process is effective, and in the case where the number of learning data is large, the model using deep learning is suitable.
  The effectiveness of the modeling strategy is demonstrated by an engine benchmark problem provided by the joint research committee of JSAE and SICE.

View full abstract

  An integral equation method for wave problems in frequency domain has a bottleneck in solving a linear system of equation with completely dense and complex coefficient matrix by discretization. A fast multipole algorithm (FMA) is a fast computation method for wave radiation fields and is used as a ways of coping the bottleneck. In the FMA, multipole expansion coefficients for the wave radiation fields is computed by using an element integration. It needs to precisely evaluate the multipole expansion coefficients for wave analysis with high accuracy. This paper aims to realize an accurate evaluation of the multipole expansion coefficient with the lowest computational costs. The authors previously presented a method to reduce a square element integration (double integration) to a single one. The above idea is applied to the computation of the multipole expansion coefficients. Numerical experiments disclose a relationship between the number of abscissas for the Gauss-Legendre quadrature and the accuracy of the computation for the multipole expansion coefficient.

View full abstract

  A variety of Krylov subspace methods for solving linear systems with symmetric and nonsymmetric coefficient matrix have been proposed. For example, a Conjugate Gradient method, BiCGStab method, BiCGSafe method and Generalized Minimal RESidual method and so on are listed. Moreover, many preconditioning techniques have been also proposed by a lot of researchers. In this paper, we focus on preconditioned Krylov iterative method for efficiently solving linear systems with symmetric coefficient matrix which appear in engineering simulations. As a result, it turned out that MrsR method with DRRIC preconditioning can efficiently solve structural symmetric problems in engineering simulation.

View full abstract