In this study, we present numerical results for the defect topology identification based on the adjoint variable and the finite element methods. The performance function is defined by the time integration of square value of difference between the computed and the observed displacements, and we solve the problem that unknown defect topology is identified so as to minimize the performance function. As for the constraint condition for the performance function, the equation of motion is introduced, and the performance function is extended by the adjoint variable method. We can obtain the Lagrange function, the first variation of the Lagrange function is calculated. Consequently, the gradient of the Lagrange function with respect to the level-set function is obtained, and the defect topology is finally obtained by solving the reaction diffusion equation using this gradient. Some numerical results are shown in this study.

In this study, we introduce the idea, the present research state, and the prospects for the optimal shape/structure design of fishing lures based on CAE. Since a fishing lure appeals its underwater motion or behavior, e.g., fluid-induced vibration, rotation. sound, flashing light, as baits to catch fish, we mainly investigate the vibration behavior of fishing lures and perform their shape/structure optimization for lure design. Up to now, only considering water pressure as external force on vibration lures, we have developed a gradient-based shape optimization method for enhancing their vibration behavior. Recently, considering the real analytical condition, we reproduced flow-induced vibration of fishing lures by weakly coupled fluid-structural interaction analysis and optimized their optimal structures using the Box-Behnken design. In the future, we aim to develop a gradient-based shape/structure optimization method for lure design that can make lure fishing more efficient and more interesting.

A method that combines time-window particle swarm optimization (TWPSO) and covariance matrix adaptation evolution strategy (CMA-ES) is applied to the simultaneous optimization problem of trusses in which the cross-sectional areas and the nodal positions are discrete variables. To accelerate the convergence of TWPSO, the change rate of the member cross-sectional area is modified. In addition, to improve the solution search performance of CMA-ES, the variance of the variable distribution is modified. The solution obtained by TWPSO, which shares global information in the solution space with the entire swarm, is utilized as the initial solution for CMA-ES, which has solution improvement performance against local optimal solutions. Rational optimal solutions can be obtained even for three-dimensional trusses with hundreds of design variables using the proposed method.

Inthispaper, weattempttospeeduptopologyoptimizationbasedonCMA-ESbyusingthespectrallevelsetmethod (SLSM). Although CMA-ES provides robust topology optimization for design problems with intractable properties, it suffers from the curse of dimensionality. SLSM uses Fourier series expansion to generate level set functions to represent the boundary between structures and voids, which significantly reduces the number of design variables for topology optimization. We aim to use SLSM to speed up topology optimization with CMA-ES and evaluate it on several design problems.

With the progress of the shift to EVs in automobiles, the improvement of "electricity cost" has become an important issue. In order to improve electricity cost, it is especially necessary to reduce the weight of the vehicle body structure other than the battery since the battery itself have large portion of the vehicle weight. Therefore, the possibility of using a lightweight and high strength cylindrical honeycomb structure proposed by Origami Engineering as an impact energy absorbing member that accounts for a relatively large proportion of the vehicle body weight was investigated. To evaluate the impact energy absorption performance, the cylindrical honeycomb structure, the conventional tubular honeycomb member and the box beam type member were compared, and tried to optimize the thickness of the members for weight reduction while maintaining the current impact energy absorption performance. As the result, it was shown that the cylindrical honeycomb structure had excellent energy absorption performance than other members.

Multiobjective topology optimization was performed for innovative design of airfoils at takeoff. An evolutionary algorithm employing a quadtree structure for solution encoding was coupled with a numerical fluid dynamics solver. The objective functions were to maximize lift and minimize drag, and constraint was imposed on the cross-sectional area to ensure design feasibility. Twenty-three non-dominated solutions were obtained through optimization. The number of airfoil elements and shapes of the solutions were confirmed to be diverse. Also, it was confirmed that a shape similar to flaps found in conventional designs was also generated.

Fiber shape optimization was performed to improve the damping properties of CFRP fabricated by electrodeposition resin molding (EDRM) method. Damping properties of composites were modeled by the specific damping capacity (SDC), and the material damping parameters required for numerical estimation were identified by an inverse analysis using results of experimental modal analysis and finite element analysis. In the optimization, a curvilinear fiber shape that maximizes the first mode SDC was explored. The thickness distribution due to the curvilinear fiber shape was estimated and applied to the finite element analysis. The fiber shapes obtained by the optimization improved the first natural frequencies and mode SDCs compared to several straight fiber orientations, indicating that the curved fiber orientation is effective in improving damping.

Perceptible floor vibration occurs infrequently due to human activities such as walking in steel framed buildings. A general strategy to reduce the floor vibration is to install Tuned Mass Dampers (TMDs) between a floor slab and a raised floor. Two parameters, resonance frequency and damping ratio, of a TMD should be designed properly depending on the dynamic characteristics of a target floor so that the TMD acts effectively. In a conventional method, even when installing multiple TMDs to a floor, they have been regarded as one large TMD and the fixed theory, which is a method to design a single TMD, has been applied. We applied differential evolution, which is one of global optimization methods, to design of TMDs to reduce floor vibration and verified its effectiveness in an actual building. Vibration due to human walking of a floor to which TMDs has been installed were measured. As a result, 6 TMDs designed by differential evolution achieved vibration suppression performance equivalent to 10 TMDs designed by the fixed theory.

This study proposes a shape optimization method to minimize the maximum thermal stress of the microstructures in a multiscale structure caused by thermal loading. Using the homogenization method, the homogenized elastic tensors and the homogenized thermal expansion coefficient tensors for the microstructures are calculated to link the microstructures to the macrostructure. The KS function is used to avoid discontinuities and singularities in the maximum thermal stress. Based on the problem formulation, the shape gradient functions are derived for shape optimization.

In this study, we propose a multi-scale shape optimization method to find the optimal macrostructure and microstructures that achieves the target displacement for a porous laminated shell structure. The displacement at any point of the laminated shell structure is controlled to the target value under the total volume constraint by minimizing the squared error. The equilibrium equation for the macrostructure and the homogenization equations for the unit cells are used as constraints. The shape optimization problem is formulated, and the shape gradient functions theoretically derived are applied to the H¹ gradient method. The homogenization method is used for connecting the macrostructure and the microstructures. With the H¹ gradient method for shape optimization, a smooth macrostructure and microstructure can be obtained.

In this paper, high-temperature-resistant ultrasonic transducers for nondestructive inspection of chemical plants and power plants were fabricated. And the transducers were tested for high-temperature fatigue. Nondestructive pipe wall thickness measurements are typically conducted as piezoelectric ultrasonic transducers with acoustic bonding material like couplant material and backing material. However, the heat resistance of the couplant material and backing material cannot withstand nondestructive thickness measurement of piping in high-temperature environments. Therefore, in this method, piezoelectric devices were directly fabricated to the piping by applying piezoelectric ceramics with high temperature resistance and flexibility. The abilities are generated by a porous constitution of the film using a sol-gel composite spray coating method. The fatigue tests were conducted to the devices by repeatedly exposing them to 600°C. The devise could be available 6 times temperature increase and the decrease of ability acceptable. This paper describes sufficient availability of the so-gel composite piezoelectric film on high temperature environment.

In this paper, we propose new type of particle swarm optimization named forked divergence particle swarm optimization (PSO). PSO is one of the powerful heuristic optimization methods due to its simple algorithm, good convergence, and ease of implementation for various problems. However, its performance is highly dependent on the velocity term in the gradient equation, the value of which is the same for all individuals, making it difficult to have diversity in the individuals. Therefore, when given an inappropriate velocity term parameter or when the problem is complex, premature convergence occurs and the search stalls. In proposed method, each individual has its own parameter values and diffuses at the individual and global level depending on each convergence status. By using forked divergence PSO, we can avoid premature convergence to the locality and continue the search. In this study, we demonstrate the proposed method by numerical examples and demonstrate its effectiveness and characteristics.

In the multi-stage hot forging, flash should always be minimized for material saving, and the uniform deformation is also preferable for high strength of a product. In this paper, the billet shape and the process parameters in multi-stage hot forging is optimized so as to minimize the flash and the equivalent strain using sequential approximate optimization. It is found from the numerical result that the trade-off between the flash and the equivalent strain is clarified, and the flash is drastically reduced compared to the conventional billet shape.

In general, structural optimization emphasizes the functionality of a structure through maximizing or minimizing a specific objective function. Therefore, it is interesting to realize the optimization that takes into account the design characteristics, as in humain design. To achieve this, we focused on the structural optimization method proposed by Guo et al. that incorporates dimensionality reduction by VAE, and applied it to an acoustic problem. In this article, we present an example of the sound pressure maximization problem at an observation point based on the experiments of Guo et al. We succeeded in obtaining optimized structures that takes into account the design constraints of the given design data and still performs the function.

In this paper, we develop a topology optimization to control thermal waves which allow wave-like profile of temperature by conduction and convection. To calculate the convection velocity for low Reynolds flow in porous media, the Darcy law and the pressure equation are solved, and the thermal wave is analyzed by the finite element method assuming the temperature of harmonic wave in the advection-diffusion equation. The structure for cloaking the temperature distribution of the thermal wave disturbed by obstacles is designed by topology optimization, and the level set function representing the structure is optimized by the covariance matrix adaptation evolution strategy (CMA-ES).

This study proposes a simoultaneous design optimization for a ply drop-off laminated composite to maximize strength at the regin pocket in terms of both the stacking sequence and the ply drop-off placement under several empicical stacking constraints. Genetic algorithms including the gene repair strategy and adaptive mutation is adopted for efficient searching. The strength is evaluated based on the stress index based on FEM analysis using the three-layer model. Through numerical examples, the validity of the proposed simultaneous design method is demonstrated.

Walking includes many variations among humans, from one step to another. Understanding what kind of walking variations exist is therefore important to determine how normal gait behavior is. Ideal walking characteristics by considering human dynamics can be found by using optimization. For this purpose, a multibody dynamic model of human walking was developed. After creating multibody dynamic model, the motion patterns were optimized by minimizing total joint energy and moment during one gait cycle. To define the joint angle characteristics during one gait cycle, many design variable should be considered. However, the number of required design variables were reduced significantly by using singular value decomposition. After obtaining optimal walking patterns, a passive walking assistive system was proposed to reduce joint power and moment. The stiffness and location of passive walking assistive system was optimized to reduce joint moment and power further.

When designing a part of machines, it is desired to generate shapes that satisfies performance requirements. For such an aim, deep generative models are used. Generative adversarial network (GAN), variational autoencoders (VAE), and VAEGAN are usually employed. In the present study, we compare those three generative models, and explain the necessity of physics guided generative models.

A determination of cracks using topology optimization in 3-D time domain acoustics is considered. We introduce the cost function which is the misfit function between the obserbed data and the numerical data on the boundary of the domain having the cracks. We determin the cracks as the minimizer of the cost function using the topological derivative. The determination of two cracks having the less forecast information are shown in this paper.

In recent years, the demand for high-performance natural convection heat sinks has been increasing due to the miniaturization of electronic devices and their wide range of applications. It has become important to design a natural convection heat sink that can exhaust the maximum amount of heat with a compact size. In this study, we apply the framework of data-driven multifidelity topology design (MFTD) for the design problem of natural convection heat sinks. In this method, we initially prepare several promising and diverse solutions and iterate generating new solutions from them while preserving diversity and structural features and simultaneously arranging moderate perturbation. At each iteration, better solutions are selected as the seeds for generating the succeeding set of solutions, and they are arranged into new ones through a deep generative model. It is expected that repeating this operation finally reaches superior solutions without gradient information. We applied this framework to the 2D topology optimization problem. The shape of the heat sinks was optimized for several values of the coefficient of volumetric expansion (β), which is a number that determines the strength of convection. Different optimal structures were obtained by changing β, and they were reasonable. For example, the optimal solutions under strong convection had partial structures with rectifying effects.

The trial numbers have been dependent on the row numbers of Design matrices. At layout of three levels factors, we had used to apply Hadamard type L₂₇(3¹³) and ,mix type L₁₈(2¹3⁷). Lower subscripts means the trial numbers.We adapted the conference matrices C with composited by Legendre Symbol as the design matrices for optimizing. Also, we adapt the raw data instead of log transformation with contaminated by non-linear effects. And data analysis was changed from the decomposition of the sum of square style to regression analysis. It was difficult for researcher to treat the interaction effect between factors. Hadamard design matrices will use the linear graph to catch the interaction which was the subunit of design matrices. It will reduce the numbers of layout columns in design matrices. Mix type will confound the interaction to main effects. So, it will contaminate the main effects. Conference matrices has only the linear effects. So, we have adapted the new system to confirm the source of the factors on the interaction. It is the difference levels between factors by comparing the trial best condition(a) of the design matrices with the best combination(b) of the factor effect. The same level factors are relative the main factors and the different level factors are relative the interaction effects. So, after gathering new ways, we could reduce the trial numbers to 1/3-1/2 with the same result to the current optimization. We are going to propose to apply it for SDGs subjects.

The algorithmic design is a structural shape design approach based on mathematical functions that consists of several parameters. On the basis of the approach, optimual structural design problems can also be formulated with respect to a relatively limited number of design variables, generating somewhat elegant and complex resultant designs. In the current study, the approach is applied to optimal design of truss structures. We develop a formulation applicable to various types of truss structures. Some computational examples are also demonstrated.

To calculate the behavior of wind turbines in real-time and highly accurate, we developed a method to assimilate an approximation model with measurement data without the finite element analysis of mechanical systems. There is an advantage that the particle filter is easy to implement as data assimilation which identifies uncertain parameters of simulation with measurement data. On the other hand, it is necessary to prepare many particles, and accordingly the calculation cost of the data assimilation is high when using the finite element analysis. The input and output of the analysis were therefore approximated by using neural networks to realize real-time computation. The proposed method was applied to determine an uncertainty parameter included in the input of simulation for obtaining the behavior of wind turbines, and consequently, the behavior was accurately simulated with the resulting parameter.

This article introduces a shape optimisation system for 3D unsteady acoustic scattering problems. The main ingredients of the system are the adjoint variable method, fast and stable boundary element method, and NURBS representation of surfaces. The applicability is demonstrated through a numerical example.

In various data analysis tasks, modeling patterns such as sets, series, and graphs as features is useful not only for improving prediction performance, but also for knowledge extraction. However, considering all possible combination of patterns exponentially increases the number of model parameters, making it difficult to train models using ordinary optimization algorithms. In this study, we introduce an optimization algorithm that uses convex optimization and pattern mining approaches to efficiently train predictive pattern mining models.

The present paper deals with the aeroelastic characteristics of aft-swept composite plate wings with structural damping. Based on the complex modulus approach, a minimum weight design of cantilevered laminated plates with a sweep-back angle under the flutter and divergence speed constraints is conducted by using a differential evolution in which lamination parameters are used as design variables.

As regards thermal-fluid topology optimization problems, many beneficial results have been reported. However, only a few studies have considered the unsteady state thermal-fluid problem. In actual design problems, there are lots of cases in which temporal changes in temperature distribution are important rather than steady-state. In this study, the framework of a density-based topology optimization method for the unsteady thermal-fluid problem is proposed. The governing equations are unsteady-state incompressible Naiver-Stokes equation and thermal convection-diffusion equation, discretized with stabilized finite element method. The design sensitivity is computed accurately based on the adjoint variable method. The numerical analysis shows that the proposed framework captures the transient effect correctly and the results of optimization are physically reliable.

We have modeled the skull by a circular array of point-like scatterers, and optimized the arrangement of point-like scatterers constructing a transcranial ultrasonic lens by the condition that the sound field intensity within the skull has a peak at a focal point and vanishes at any other points, and shown that an ultrasonic lens with sufficient function is realized. In this paper, we analyze the case of an elliptical skull, and show that even for this case as well, the proposed optimization yields a arrangement of scatterers with equivalent performance to the case of a circular skull.

In order to consider a finite deformation for the compliant morphing flap, this study resolves a numerical instability, which is caused by an artificial weak material under excessively large deformations. The remeshing method which adopts more precise structural model without using such weak materials is integrated into a level set-based topology optimization method. Then, the optimal design problem is formulated for the compliant morphing flap which deforms to the target shape under the applied load. Through numerical examples, validity of the proposed method is demonstrated.

This paper proposes a numerical method to estimate track lateral alignment. The estimation of the track lateral alignmentisformulatedastheproblemoffindinganapproximatesolutiontoafunctionalequation, andtheSincfunction approximation, which has been studied in the field of numerical analysis, and the Fourier transform approximation based on numerical integration methods are utilized to construct the approximate solution. In particular, our newly proposed method directly deals with the singularity in the frequency transfer function involved in the estimation.

In this study, the application of the Gauss-Newton method to several material parameter identification problems was investigated. The parameters of materials with complicated constitutive laws are determined to reproduce known physical quantities such as stiffness and natural frequency. The identification problem is usually formulated as a squared error minimization problem for the physical quantity. Gauss-Newton method is one of the important solution methods. In this method, the Hessian matrix is approximated using the first order differentiation of the physical quantity with respect to the parameters. Although this method obtains good convergence in the initial stage, it is observed that it becomes unstable in the final stage. We introduced a line search technique based on the Marquardt method to avoid the instability. We compared this method with the general gradient method and confirmed good convergence of this method.

Thanks to the development of additive manufacturing technology, it is becoming possible to produce materials with desired mechanical properties defined by their periodic microstructures. Multi-scale topology optimization has been paid attention to in many engineering fields to design optimal microstructures. However, its high computational cost prevents practical use, such as high-resolution 3D analysis for precision modeling and non-linear analysis assuming actual materials. In this study, to solve this problem, we focus on the homogenization approach using fast Fourier transform and develop a new optimization method with fast computing speed and low memory requirement. By performing sample analyses, we demonstrate the validity and efficiency of the proposed method.

In the maintenance of cable structures, such as cable-stayed bridges, it is necessary to estimate the tension acting on the cables. In current Japanese practice, the cable tension is estimated using the vibration method or the higher-order vibration method from the cable’s natural frequencies. However, in recent years, the aerodynamic vibration of cables caused by wind has become a problem. To suppress the aerodynamic vibration, dampers are installed onto the cables. Because the damper changes the cable’s natural frequencies, the vibration method and higher-order vibration method are inappropriate for measuring the tension of a cable with a damper. This paper introduces a tension estimation method for a cable with a damper proposed by the authors. The authors’ method inversely estimates the tension and bending stiffness of the cable and damper parameters, simultaneously, from the natural frequencies. The validity of the method was confirmed by conducting numerical simulations and experiments.

One of the authors has proposed a motion analysis method for multibody systems called nullspace matrix method of differential equation type. In the proposed method, a nullspace matrix for the constraint Jacobian matrix is obtained by solving a differential equation. Then, the term which represents the constraint force in the equation of motion is eliminated with the nullspace matrix. The proposed method is applicable to systems with singular configurations. For efficient design of machines and structures, sensitivity of the behavior of the system with respect to design variables such as lengths of linkages and positions of joints is required. In this research, based on the procedure of the nullspace matrix method of differential equation type, a motion sensitivity analysis method of multibody systems is proposed. An example of motion sensitivity is presented.

In practical design of building structure, the cross-sectional dimensions of the members are selected from a prescribed standard list. Machine learning (ML) is potentially effective for this problem; however, preparing a large dataset in advance is difficult. Even in this situation, Reinforcement Learning (RL) is applicable, can assist to find a better solution, and is expected to reduce computational cost when compared without the RL agents. We propose a new minimum volume design of a building steel frame based on a decomposition and reconstruction framework using a RL agent. In the method, i) the design problem of a space frame is decomposed into that of plane frames, ii) the plane frames are optimized by standard nonlinear programming, iii) and the plane frames are assembled to a space frame by the RL agent. The advantages of the methods are presented through a numerical example.

There has been research on airfoil design using deep generative models such as generative adversarial network (GAN). However, in prior methods, the generated results do not always satisfy the governing equations. This paper reports the results of an attempt to construct a physics guided deep generative model and use it for fine-tuning. Computational software that calculates the aerodynamic performance of shapes was placed on a network. An objective function was expressed in terms of the relationship between the generated and required performance. As a result, it is confirmed that desirable shapes that accurately satisfies requirements were obtained, but on the other hand, a drawback was found in that the variety of shapes was reduced.

In recent years, topology optimization methods have been adopted not only for structural problems but also for fluid flow problems. Topology optimization for unsteady flows requires a fine mesh, which is computationally expensive. Therefore, we propose an transient incompressible flow topology optimization method based on the building cube method (BCM), which is suitable for massively parallel computing. BCM is a type of hierarchical Cartesian mesh method that has been reported to have very good scalability. The transient Navier-Stokes equations in the topology optimization procedure are solved by a finite volume discretization based on the BCM. The sensitivity of the objective function is obtained by continuous sensitivity analysis based on the adjoint variables method.

We propose a choice of stabilization parameters in x-ray computerized tomography by a Cauchy type boundary integral formula. Since most of the inverse problems are ill-posedness in the sense of Hadamard, accurate numerical proceduresalsoapproximateitsinstability. Thereforecomputationalandobservationerrorsinfluencesnumericalresults seriously. In order to relax it, regularization methods are introduced with some artificial stabilization parameters which control accuracy and stability. The proposed criteria gives a posteriori choice of the stabilization parameter, and is computable from only measurement data, and no a priori information are required. We will also show a feasibility of the proposed method by numerical reconstruction from measurement data.

For the planning and conception process of automobiles where a large number of control factors and constraints exist, we will research methods for creating side constraints that satisfy all of them and performing overall optimization in a short period of time．In previous research, we succeeded in maximizing the range of design variables and obtaining the maximum aspect constraints for benchmark problems．I was also satisfied with the accuracy．In this study, we will verify the practicality of these methods using the actual car problem．

Data-driven topology design is a method that enables gradient-free topology optimization like evolutionary algorithms. However, the dimension of design variables that can be handled by this method is limited to about 10⁴ due to the fact that the deep generative model is used as the driving force for the solution search. Therefore, for high-resolution problems with large numbers of design variables, it is necessary to use some method to reduce the dimensionality of the design variables handled by the deep generative model. In this study, we propose a new framework that incorporates an interpolation function-based surrogate representation into the design process and discuss its applicability to high-resolution problems. We demonstrate that the performance of solutions obtained using the proposed method can almost achieve that of the existing method under a lower computational cost by demonstrating the effectiveness of the proposed method.

By regarding the solution updates in extremal optimization as a local search process, its intensification function is characterized by the downward interbasin dynamics observed in the cost landscape, and its functionality as well as the optimization performance is investigated for the solution of the traveling salesman problems. While the parameter dependence of these dynamics is similar to that observed previously for simulated annealing with constant temperature, the operating times needed to show their optimization ability differ significantly for the two methods. The comparative experiment is continued by considering the reciprocal relationship between smoothing the landscape and annealing the system. It is shown that, contrary to the original finding, thermal dynamics can be competitive with extremal dynamics in solution quality under the function-oriented operating condition.

The optimization function of simulated annealing (SA) and temperature parallel SA (parallel tempering (PT)) is studied in solving the quadratic assignment problems (QAPs). The experimental analyses are performed in the same manner as for the traveling salesman problems (TSPs). As similar to the case of TSP, the optimization performance of SA is maximized by the intensive search at some intermediate temperature and at this effective temperature, downward inter basin transition dynamics last until the end of the search time. The performance of PT also depends on the search in some intermediate temperature range; however, the role of the effective temperature is unclear compared to the case of TSP.

It is efficient to use the modal method for shape design of structures considering vibrations in a wide frequency range. However, when there are viscous dampers in the structure, it is difficult to use the modal method because the equation of motion cannot be decoupled. In this paper, we propose an efficient shape optimization analysis method using the modal method for the displacement regulation problem in a wide frequency range with a viscous damper. The shape gradient function is derived analytically, and is given as a modal solution. In addition, the H1 gradient method was used for shape modification analysis. Using a simple numerical model, it was shown that this method can be applied not only to the shape-optimal design of structures, but also to the shape-optimal design of the materials for eardrum reconstruction.

In this study, the shape reconstruction method of a cavity is developed for 2-D pure antiplane anisotropic elastodynamics. Elastic wave scattering by a cavity is implemented by using the convolution quadrature time-domain boundary element method. Scattered waves obtained at receiver points are prepared and utilized for the deep learning. The created deep learning model estimates position and size of a cavity in 2-D pure antiplane anisotropic solids. Numerical example shows that deep learning can be useful for inverse analysis of a defect in anisotropic elastic materials.

To improve the thrust by the O/F shift, which is one of the weak points of the hybrid rocket, a method with a variable oxidizer strength has been proposed. In this study, the superiority of this type of hybrid rocket over the conventional hybrid rocket without variable oxidizer supply was compared through several launch vehicle designs, such as altitude maximization by vertical launch and downrange extension problems with an optimized launch angle. A multi-objective evolutionary algorithm and the multi-disciplinary evaluation method of the launch vehicle were applied.

In aerodynamic multi-objective optimization for object geometry, design variables are often multidimensional and weakly independent, which reduces the efficiency of optimization. In this study, a design space reduction method using ANOVA and AIC was proposed and implemented, then evaluated by applying it to a multi-objective aerodynamic optimal design problem. The results show that the proposed design space reduction method can extract important variables appropriately and enables fast, efficient, and accurate solution search.

For the transformer which is general industrial apparatus, electromagnetic noise and earthquake resistance are also important functions. Especially electromagnetic noise serves as an environmental load, so noise control is called for.

Therefore, in order to predict the oscillation characteristic of an iron core correctly, a suitable material constant is needed. This research considers the process of asking for this material constant efficiently. The oscillation characteristic is the eigenvalue and corresponding vibration mode of the iron core can be obtained by experimental modal analysis. This research shows that change of the oscillation characteristic at the time of changing bolt fastening force and bolt positions can be predicted by setting up a suitable material constant using steel sheets which imitated the iron core.

The design theory of dynamic absorbers is a classic research area in the vibration engineering. In contrast, there remain many problem settings where rigorous formulations are not widely known. This paper deals with a minimization of a maximum transmissibility when multiple dynamic absorbers are installed in a multi-layered structure. For the optimal solution, the author has found that the necessary conditions can be expressed in terms of equations via the submatrix of Jacobian, and optimization can be achieved using the Newton-Raphson method. This is suitable for the case that the numerical solution confirms that the heights of all resonance points coincide at the optimal solution. In this paper, it is shown that this formulation can be explained in the context of Lagrange's undetermined multiplier method.

Multi-material structures consisting of several different material properties are promising for achieving ideal functionality over single-material structures. In particular, increasing the value of the fundamental eigenfrequency is important because it can enhance dynamic stability. Since the beginning, topology optimization (TO) for maximization of the fundamental eigenfrequencies of single-material structures has been studied. However, there have been few research papers on multi-material topology optimization (MMTO) for eigenfrequency maximization problems. Therefore, we propose a new framework for MMTO of eigenfrequency problems. In the proposed method, a multi-material level set (MMLS) method is developed, which represents the interfaces between different material phases by isosurfaces of multiple level set functions. The proposed design methodology uses the reaction-diffusion equation (RDE) to update the level set functions based on topological derivatives, allowing the nucleation of new holes during the optimization process. In addition, the eigenfrequency analysis with ersatz material and that without ersatz material using body-fitted mesh are compared and discussed.

Topology optimization is a structural optimization method in which design variables are updated based on the design sensitivity of the objective function and constraints to approach the optimal structure. However, the solution obtained by optimization based on design sensitivity is a local optimum, which is not suitable for nonlinear and multimodal problems. In this study, we propose a topology optimization method based on differential evolution, which is an excellent computational method for global search. While differential evolution has high global search capability, it is computationally expensive, so it is effective to reduce the number of design variables. Therefore, the Karhunen-Loève (KL) expansion is applied to the density function and the terms with small eigenvalues are truncated to express the structural density with a limited number of design variables. We have also proposed a method of binarization by gradually increasing the penalty parameter. In order to construct the optimization algorithm, the optimization is performed for the stiffness maximization problem of a cantilever beam, and the validity of the proposed method is verified by comparing it with the optimization results based on the design sensitivity.