The purpose of this paper is to use a surface height function method and a Moving-Grid Finite Volume Method to perform a hydrodynamic analysis when a plurality of fish-shaped submerged objects moves with rotation. Two fish were rotated and the fish were arranged longitudinally and horizontally. As a result of the analysis, the free surface shape above the fish also changed. When the fish were placed longitudinally, the drag coefficient in the rear fish was larger and the lift coefficient was positive in the backward fish. When the fish were placed horizontally, the drag coefficient was larger in the outer fish and the lift coefficient was positive in the outer fish. It shows the possibility to recreate a swimming environment that is close to the actual one than cannot be calculated by the conventional method.

The particle methods such as the SPH method and the MPS method have been widely used for the analysis of free-surface flows. This paper focuses on the initial placement of particles for the particle method. The initial particle placement on the regular Cartesian grid is generally adopted. However, it is difficult to represent slopes and curved surfaces of the wall boundary and then has negative effects on the precision of the numerical calculation. To solve this issue, we propose a new initial particle placement determined by the centroidal Voronoi tessellation (CVT), which is a special type of Voronoi tessellation. Using CVT in the two dimensional case, Voronoi cells asymptotically agree to regular hexagon grid. In this paper, we propose a CVT-based particle placement algorithm and then demonstrates some numerical examples with curved wall boundaries in the MPS method.

Handwork by highly-skilled workers is an indispensable technique in manufacturing process. However, highly-skilled handwork is usually tacit knowledge which is difficult to express in words, and it takes a long time to learn. The authors have been trying to express the handwork’s motion as multi-dimensional orbit vectors by low-order spline functions to convert the tacit knowledge to explicit knowledge, and ultimately to automate the handwork by robots. In order to clarify the characteristics of the handwork that appear in the orbit vectors, it is necessary to establish a new numerical computational method to compute higher-order derivatives that exceed the order of its component function. In this computation, simultaneous equations that consist of Taylor expansions at more than ten different points of increment need to be solved. The coefficient matrix of those simultaneous equations is asymmetric, dense and severely ill-conditioned. Since the simultaneous equations with such a coefficient matrix cannot be solved by normal method such as LU decomposition, the authors propose a new computational method called “Odd-Even decomposition of coefficient matrix.” Then, it is confirmed that this method accurately computes approximation values of up to at least the 20^th order of derivatives.

The present study proposes a topology optimization method considering finite deformation for loading uncertainty. The loading angle is assumed to be uncertain as a condition. The objective is to minimize expectation and standard deviation of end-compliance obtained by means of a Total Lagrangian finite element formulation. In this case, an analytical estimation of the expectation and the standard deviation is not allowed. In order to solve this problem, we approximate the end-compliance by a Taylor series expansion and derive the mathematical formulation. In this approach, the second derivative of the objective function is necessary to keep the accuracy in sensitivity. This phenomenon is investigated in terms of numerical validations. Finally, some numerical examples demonstrate the usefulness of the proposed method.

The absorbing boundary conditions is developed for analysis of two-dimensional elastic waves using MPS method. Absorption boundary based on PML absorbing boundaries is applied to MPS method. Performance is evaluated by giving longitudinal wave or transverse wave as input and calculating energy reflectance. When the thickness of the absorption boundary layer is 100 layers or more, the reflectance is less than -20 dB. In the case of P waves, it peaked at about -22 dB, and the performance don’t improve even when the thickness is 200 layers. In the case of S waves, it peaked at about -32 dB.

In this study, an optimal design of lattice sandwich panels based on the ground structure method has been investigated. In our calculation, the sandwich panels were assumed to be subjected to three-point bending load. The lattices discussed in this paper can be manufactured by using the selective laser metal melting machine. Based on the former experimental study on the manufacturable diameter of each strand, a new equation for controlling the cross-section of each strand was introduced, which is based on the multiphase material optimization approach. In this paper, three kinds of unit-cell models (BCC, f₂BCC and half-BCC) were applied as an initial shape, and the effect of the micro-architecture on the optimized result was discussed. It is found that the lattice composed of f₂BCC unit behaves the lowest compliance among three geometries. The optimized shape was manufactured with a good accuracy by SLM.

This paper presents a computational modeling of concrete cracking due to steel-corrosion using a damage model. The damage model, which is based on fracture mechanics for concrete, is applied to the constitutive law of concrete with cracking. The constitutive relation of concrete adjacent to reinforcements is modified to simulate the crack propagation behavior due to steel expansion using the damage model. Several numerical examples in 2D and 3D are presented to demonstrate the validity of the proposed model. First, the performance of the modified damage model is verified in 2D in comparison with the isotropic model. Second, the 3D crack behavior provided by the proposed model is compared with that obtained experimentally. Finally, the applicability is verified to a detailed 3D simulation with tetrahedral finite elements.

This paper presents a numerical advection method for a three-phase flow which is composed of a solid phase and two liquid phases. Two interfaces of the three materials are reconstructed by nested dissections scheme using three-dimensional PLIC(piecewise linear interface calculation) method successively twice for a single computational cubic cell. An algorithm to apply PLIC method successively twice for a cubic cell volume is developed by extending an existing PLIC method which uses convergent calculation to determine the position of an interface plane.

Effectiveness of the proposed PLIC algorithm is examined by demonstrating cell division with various combinations of volume ratios and normal vectors. The results of the above demonstration show satisfactory iteration times required up to convergence.

The proposed PLIC is applied to a time-dependent incompressible shear flow field in a parallelepiped domain. A series of numerical experiments to track deformation of a liquid column in the domain were performed at various numbers of cells. The result shows an excellent convergence ratio in accordance with cell numbers, total mass error and shape errors.

According to a recent experimental measurement, the limiting Reynolds number Re_c for the appearance of vortex breakdown bubbles confined in a cylindrical container is shifted toward higher values of Re than that reported in the classical experimental study by Escudier. In order to study this point, numerical simulation is performed using finite-difference codes. Grid-convergence-index (GCI) is estimated and grid-independent-solution (GIS) is derived by applying Richardson extrapolation. Whether the numerical solution shows higher values of limiting Re_c as in the recent PIV measurement or lower values of Re as in the classical dye-injection observation by Escudier is assessed.

The voxel-based FEM is an efficient numerical procedure to deal with complex geometries. It is difficult to analyze models containing contact conditions by this procedure, in which surfaces of objects are represented as stepwise shapes. In the voxel modeling based on the finite cover method (FCM), geometry is defined by using surface modeling procedures such as polyhedrons and the level set method instead of voxel representation and hence contact problems can be set more easily. In this paper, a contact algorithm for the FCM is proposed. The penalty method is widely used for contact constraint, and then a large penalty parameter, which needs to be employed to assure accuracy, may cause numerical instability. Moreover, in the FCM, it is difficult to manage positions of integration points for contact constraints and inaccurate distribution of contact stress may be obtained due to an over-constraint with concentration of integration points. In this paper, the Nitsche’s method is applied to contact constraint in the FCM in order to circumvent difficulties arising in the penalty method. To show effectiveness of our approach, we calculate several 2D frictionless contact problems.

Piezoelectric bimorph for actuators and sensors is usually thin and includes conductor layers such as the electrode and the shim plate. Therefore, this study deals with the coupled piezoelectric and inverse-piezoelectric analysis of thin piezoelectric bimorph with the conductor layers. Solid elements can describe the various types of distributions of electric potential along the thickness, while shell elements are more appropriate to analyze the thin structures. Therefore, instead of using piezoelectric solid elements or piezoelectric shell elements for both the piezoelectric and inverse-piezoelectric analyses, the solid and shell elements are used to simulate the electrical and structural fields, respectively. The coupled algorithm is based on the block Gauss-Seidel method and the transformation method between the solid and shell variables. In the structural analysis, the rule of mixtures for the bending rigidity and the mass is used. A pseudo-piezoelectric evaluation method for the conductor is proposed in order to use the existing programs for the piezoelectric and inverse-piezoelectric analyses without any modification. Finally, it is demonstrated that the proposed method shows very accurate potential distributions in the actuator and sensor modes of the thin piezoelectric bimorph with the conductor layers.

In this paper, we proposed a novel numerical treatment for no-slip wall boundary conditions for incompressible fluid flow simulations in the LSMPS (least squares moving particle semi-implicit) method. In the conventional wall boundary treatments, velocity and pressure on a wall particle were treated as computational variables to enforce prescribed boundary conditions. However, such a treatment not only increases computational cost but also causes numerical instability. In the proposed treatment, no variable is defined on a wall particle, and boundary conditions themselves are incorporated into the spatial discretization scheme by means of the weighted least squares approach. To verify consistency of the proposed scheme, convergence studies were carried out. As numerical examples for fluid flow simulation, plane Poiseuille flow, three-dimensional square channel flow, rigid rotation problem and lid-driven cavity flow have been calculated. As a result, high accuracy and validity of the proposed method were confirmed.

The objective of this study is to develop a new cohesive zone model derived from Paris law and identification of its material parameters to realize fatigue crack propagation. A damage variable representing material degradation under cyclic loading is introduced into an existing cohesive zone model within the framework of thermodynamics and the evolution law is determined from Paris law. After simple validation is conducted to investigate the effect of the material parameters on crack growth rate, we present the parameter identification to employ the material parameters of Paris law on our proposed model. Several numerical examples are provided to demonstrate the consistency assessment of the proposed model with previous studies, and the capability for problems of fatigue crack propagation affected with different plastic deformation.

A novel numerical electrodeposition (ED) model for high accuracy automotive ED simulation is proposed. In contrast to the conventional ED models, the newly proposed ED model is based on the new basic ED experiments replicating actual manufacturing line conditions. The new basic ED experiment introduces an external resistor so that the time history of the surface potential on the cathode is similar to that of the actual lines. The proposed ED model consists of two submodels: a film resistance model and a a film growth model to decide the electric resistance and the thickness growth rate of the coated film, respectively. The new resistance model considers the flow rate dependency of the paint as well as the film thickness dependency. Meanwhile, the new growth model considers the concentration diffusion of paint particles before the initial film deposition. Some finite element analyses of the basic experiments including the 4-plate box test show that the results with the proposed model are more accurate than those with the conventional model.

A data-driven approach is developed for micro-macro coupled multiscale analysis of hyperelastic composite materials. The offline process in this approach is to make a database that stores the microscopic stress distributions in response to various patterns of macroscopic deformation gradients. This can be done by carrying out an adequate number of numerical material tests on a periodic microstructures, or equivalently, a unit cell and followed by the proper orthogonal decomposition (POD) to extract the principal components of the data along with the corresponding basis vectors. In order to realize FE²-type two-scale analysis in the online process, we interpolate each of the coefficients with the radial basis functions as a function of a macroscopic deformation gradient and make the resulting continuous function gently varying by means of the L2-regularization followed by the cross-validation and Bayesian optimization techniques. Each of the functions thus obtained is referred to as “data-driven function” of the microscopic stress distribution and can be used to obtain the macroscopic stress by the averaging process in the homogenization method. A representative numerical example is presented to validate the proposed data-driven FE² analyses in comparison with high-fidelity direct FE².

Singly diagonally implicit Runge-Kutta (SDIRK) schemes are applied to the projection method for the incompressible flows. SDIRK-projection method is compared with the second-order pressure correction projection method by Kim and Moin and the third-order velocity correction projection method by Owen and Codina. The result of benchmark computation by SDIRK-projection method shows considerable stability and computational efficincy.