Jung-Sun Park

Optimization of a composite laminated structure by network-based genetic algorithm

Genetic algorithm (GA), compared to the gradient-based optimization, has advantages of convergence to a global optimized solution. The genetic algorithm requires so many number of analyses that may cause high computational cost for genetic search. This paper proposes a personal computer network programming based on TCP/IP protocol and client-server model using socket, to improve processing speed of the genetic algorithm for optimization of composite laminated structures. By distributed processing for the generated population, improvement in processing speed has been obtained. Consequently, usage of network-based genetic algorithm with the faster network communication speed will be a very valuable tool for the discrete optimization of large scale and complex structures requiring high computational cost.

Optimal Design of a High-Agility Satellite with Composite Solar Panels

This paper defines mode shape function of a composite solar panel assumed as Kirchhoff-Love plate for considering a torsional mode of composite solar panel. It then goes on to define dynamic model of a high-agility satellite considering the flexibility of composite solar panel as well as stiffness of a solar panel’s hinge using Lagrange’s theorem, Ritz method and the mode shape function. Furthermore, this paper verifies the validity of dynamic model by comparing numerical results from the finite element analysis. In addition, this paper performs a dynamic response analysis of a rigid satellite which includes only natural modes for solar panel’s hinges and a flexible satellite which includes not only natural modes of solar panel’s hinges, but also structural modes of composite solar panels. According to the results, we confirm that the torsional mode of solar panel should be considered for the structural design of high-agility satellite. Finally, we performed optimization of high-agility satellite for minimizing mass with solar panel’s area limit using the defined dynamic model. Consequently, we observed that the defined dynamic model for a high-agility satellite and result of the optimal design are very useful not only because of their optimal structural design but also because of the dynamic analysis of the satellite.

Mechanical Characteristics of Helically Twisted Composite Strands Using Sub-Unit Cell Geometry

This paper proposes a geometrical modeling for studying mechanical characteristics for twisted textile composites consisting of helically twisted yarns. From a simple mathematical model for sub-unit cells, the elastic behavior of twisted and impregnated composites is predicted. To get the elastic constants, a volume averaging method for sub-unit cells is applied and compared to other models. In the analysis, the compliance matrix of a helically twisted yarn’s sub-unit cell is based on a system of anisotropic elastic materials. The constitutive properties are calculated by the compliance or the stiffness matrix with generalized transformations about the material axis. To implement the volume averaging method, the concept of equivalent volumes is introduced to geometrical volumes of helically twisted yarns within the sub-unit cell. This study also considers the effects of migration and micro-buckling of filaments on the twisted yarns’ elastic properties. From comparisons between the present and other results, a good conformity is found.

Hybrid Simulated Annealing with Particle Swarm Optimization Applied Krigging Meta Model

[Abstract] This paper is concerned with the optimization of composite housing in a multi-spectral camera using KSAPSO (Kriging-Simualted Annealing-Particle Swarm Optimization). The effective use of KSAPSO on physical problems has been expanded to provide global approximations for optimization problems. KSAPSO is a hybrid approximation algorithm. In the KSAPSO, simulated annealing (SA) uses certain probability to avoid being trapped in a local optimum and particle swarm optimization (PSO) combines global search and local search to search for the optimal results. And in order to reduce computational costs in calculating design problems, the Kriging is used in this study. Kriging is related to the experimental designs that provide maximum information by minimum number of design experiments. The proposed methodology is applied to the design of a Multi-Spectral Camera (MSC), as a practical example, which will provide high resolution panchromatic and multi-spectral images and is carried by a satellite designed to fulfill the need for further Earth observation and allowing scientists and communication experts to conduct potentially valuable experiments. When this composite structure is optimized, design constraints are taken for natural frequency and shear stress which should be considered in a launching environment. I. Introduction Today’s engineering design problems become more complex. Therefore, approximate methods such as Kriging and response surface method are used extensively in engineering applications for a variety of tasks because of the flexibility to approximate many different and complex response functions. 1 Consequently, diverse optimization methods using Kriging are offered. In this paper, an improved probabilistic algorithm is proposed for practical optimal designs by KSAPSO. The effective use of KSAPSO on physical problems has been expanded to provide global approximations for optimization problems. KSAPSO is a hybrid approximation algorithm. This hybrid approach makes full use of the strong global search ability of PSO and the strong local search ability of SA and offsets the weaknesses of each other. In order to reduce computational costs in calculating design problems, the Kriging used in this study is related to the experimental designs that provide maximum information by minimum number of design experiments. This paper typically proposes central composite design (CCD) as the classical experimental designs and uniform design (UD) as space filling experimental designs. 2 Moreover, the sequential approximate procedures by a stochastic algorithm is proposed to help the global optimum (chosen from the local optimum values) to move to higher probabilistic design domain. In this paper, the stochastic process is combined with the Chebyshev inequality, probability of success (POS) that simultaneously satisfies all constraints and objective limit. The optimization by KSAPSO is applied to the design of a composite housing of Multi-Spectral Camera (MSC) which is satellite equipment. MSC is designed for further Earth observation, and will provide high resolution panchromatic and multi-spectral images. Satellite components are too expensive and hard to repair on operation. High reliability for the satellite components is required under launching and space environments for steady functioning. In this study, discrete design variables are the ply thicknesses of composite housing. Design constraints are taken for natural frequency and shear stress that should consider on the design of satellite in launching situation.

Nonlinear Behavior of a Curved Composite Panel with Transverse Matrix Cracks

This study is to investigate the stiffness degradation of a curved composite laminated panel due to transverse matrix cracks and the effect of the stiffness degradation on the nonlinear behavior of the composite laminates. Micromechanics theory on the composite material was derived by introducing crack density. Iterative numerical scheme was developed to calculate the degraded composite stiffness which has nonlinear relation with the crack density. A nonlinear finite element program was developed using 3-D degenerated shell element and the theory of the first order shear deformation to analyze the large deformation of the composite panel. Nonlinear structural responses of the curved composite panel were examined for various stacking sequences and crack density under the bending and twisting moments. Also, the effect of crack opening/closed was considered in the examination. It is realized that the transverse matrix crack causes severe stiffness reduction and affects greatly the nonlinear behavior of the composite panel.

Analysis of a Composite Panel with Transverse Matrix Cracks under Bending and Twisting Moments

This study is to investigate the stiffness degradation of a composite laminated panel including transverse matrix cracks subjected to bending and twisting moments. Micromechanics theory on the composite material is derived by introducing crack density. Iterative numerical scheme is developed to calculate the degraded composite stiffness which has nonlinear relation due to the crack density. The finite element method is used for structural analysis of the composite panel. Structural responses of the composite panel are examined for various laminated angles and crack density under the bending and twisting moments. Also, the effect of crack opening and closing is considered in the examination. It is realized that the matrix cracks may cause severe stiffness reduction and should be considered in the composite laminated panel.

Fluid-Structure Coupled Analysis of a Fabric-Covered Airfoil

Extended Abstract Wind turbines (WTs) rapidly become bigger than bigger but conventional WT blade designs have still used fiberglass. Though fiberglass is cheaper than advanced materials like carbon composites it is heavy and its manufacturing process is extremely hands-on time intensive. This conventional WT blade design may not be good for the development of advanced WT blades. A new approach to design, manufacture, and install WT blades will be necessary in the near future. The use of architectural fabrics could change the conventional design of the WT blades [1] and could reduce the manufacturing costs of very large blade manufacturing 25% to 40%. This fabric-covered wind turbine blade is composed of spars, ribs and covering fabrics similar to the truss structure of the aircraft wings. In this concept, the front and rear spars sustain all of the loads like bending moments of a wing. The fabric-skin is mainly subjected to torque and keeps the airfoil shape. This fabric skin acts like a membrane and is deformed by the pressure acting on it. If it deforms the pressure distribution acting on it is also changed and consequently the fabric skin deforms again. Finally, due to the fabric skin deformation, the aerodynamic characteristics of the blade and wing are changed and the aeroelastic characteristics of the fabric skin should be investigated. In the present study, the aerodynamic deformations and aeroelastic characteristics of the fabric skin are studied. The membrane analysis of the fabric skin is performed for its structural modelling and the CFD (computational fluid dynamics) analysis of the airfoil section is performed by using ANSYS. The present membrane analysis results and CFD results are verified by comparing the reference results. Finally, the fluid-structure coupled analsys of the fabric skin are performed and its aerodynamic deformations, deformed pressure distributions, and aeroelastic characteristics are investigated.

Design Optimization of Composite Radar Absorbing Structures to Improve Stealth Performance

In this study, an efficient method of designing laminate composite radar absorbing structures (RAS) is proposed with consideration given to the structural shape so as to improve aircraft stealth performance. The calculation of the radar cross section (RCS) should be decreased to enhance the efficiency of the stochastic optimization when designing an RAS. In the proposed method, RAS are optimized to match up the input impedance of the minimal RCS, which is obtained by using physical optics and the transmission line theory. Single and double layer dielectric RAS for aircraft wings are employed as numerical examples and designed using the proposed method, RCS minimization and reflection coefficient minimization. The availability of the proposed method is assessed by comparing the similarity of the results and computation time with other design methods. According to the results, the proposed method produces the same results as the stochastic optimization, which adopts the RCS as the objective function, and can improve RAS design efficiency by reducing the number of RCS analyses.

Design Optimization of Double-array Bolted Joints in Cylindrical Composite Structures

A design optimization is performed for the double-bolted joint in cylindrical composite structures by using a simplified analytical method. This method uses failure criteria for the major failure modes of the bolted composite joint. For the double-bolted joint with a zigzag arrangement, it is necessary to consider an interaction effect between the bolt arrays. This paper proposes another failure mode which is determined by angle and distance between two bolts in different arrays and define a failure criterion for the failure mode. The optimal design for the double-bolted joint is carried out by considering the interactive net-tension failure mode. The genetic algorithm (GA) is adopted to determine the optimized parameters; bolt spacing, edge distance, and stacking sequence of the composite laminate. A purpose of the design optimization is to maximize the burst pressure of the cylindrical structures by ensuring structural integrity. Also, a progressive failure analysis (PFA) is performed to verify the results of the optimal design for the double-bolted joint. In PFA, Hashin 3D failure criterion is used to determine the ply that would fail. A stiffness reduction model is then used to reduce the stiffness of the failed ply for the corresponding failure mode.

Reliability Based Design Optimization using Moving Least Squares

This study is focused on reliability based design optimization (RBDO) using moving least squares. A response surface is used to derive a limit-state equation for reliability based design optimization. Response surface method (RSM) with least square method (LSM) or Kriging will be used as a response surface. RSM is fast to make the response surface. On the other hand, RSM has disadvantage to make the response surface of nonlinear equation. Kriging can make the response surface in nonlinear equation precisely but needs considerable amount of computations. The moving least square method (MLSM) is made of both methods (RSM with LSM+Kriging). Numerical results by MLSM are compared with those by LMS in Rosenbrock function and six-hump carmel back function. The RBDO of engine duct of smart UAV is pursued in this paper. It is proved that RBDO is useful tool for aerospace structural optimal design problems.