COMPOSITE LAMINATE STACKING SEQUENCE FOR WEIGHT AND BUCKLING LOAD OPTIMIZATION OF STIFFENED PANEL STRUCTURE

• Main Author: Domas Indrawijaya, Agus
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/45343
• Institution: Institut Teknologi Bandung

Fiber-reinforced Composite (FRC) has been widely used in various fields, including in aerospace industry. The main advantage of FRC material is the relatively higher specific strength and specific stiffness than conventional metal material. Furthermore, composite material has unique material’s properties that can be tailored as needed, making composite material is highly sought for lightweight structure. One of the classifications of composite is laminate structure; a structure that is formed by stacking numbers of composite lamina at a certain sequence, that each of lamina has its unique fiber direction. This stacking sequence defines the whole laminate material’s properties. In aircraft structure, one of the applications of this laminate structure is for a thin-walled structure, such as wing skin and fuselage. The thin-walled structure is vulnerable to the buckling failure, so generally, it is reinforced by stringer or stiffener, becoming a stiffened panel structure. These stiffeners help by giving an additional stiffness, increasing the critical buckling load of the thin-walled structure. Stiffened panel’s properties that designed using composite laminate will depend on the lamina stacking sequence. The most optimum stacking sequence design is certainly will be desired, but it is hard to be achieved with any conventional method because of the many possibilities of stacking sequence, constraint consideration, and different load cases. FEM can solve this problem, but the optimization using FEM taking relatively more time and computational resources, which is not appealing in the early stage of design because it is likely that the optimization process is needed more than once. One of the other alternatives is optimizing by an analytical approach with Genetic Algorithm (GA). GA is well-known for its simplicity, ability to evaluate a huge amount of sample, and is fast to search the solution with high accuracy. In this research, optimization is done by using an analytical approach for buckling analysis through closed-form solution method and Rayleigh-Ritz method, then the optimum design point is obtained by using GA. The interaction between the stiffener and the skin is fully ignored to simplify the buckling analysis, so the buckling analysis is done separately. The objective of the optimization is to find the lightest and also highest critical buckling load of symmetric, balanced, and approaching specially orthotropic stacking sequence that satisfies the buckling load and strength constraint subject to the uniaxial longitudinal load. The design variable that is used for the optimization is lamination parameters, parameters that represent the actual stacking sequence and define the laminate stiffness matrix directly. There are two phases of this GA optimization. First, the GA is used to search the optimum lamination parameters. Then, the GA is employed again to find the actual stacking sequence that is most fit for that lamination parameters. The stacking sequence that is obtained then validated through FEM analysis, to confirm whether the approach that is used is still able to satisfy the constraint in the real case of the stiffened panel structure. The optimization is successfully done with two different load cases, with the result of the stacking sequence from the optimization has the error up to 9.6% compared to the FEM analysis. This concludes that the approach that has been used in this optimization is passable to be used in the real condition of the stiffened panel structure. There are some possible causes, but the main reason for this difference is the buckling analysis that is used does not account for the interaction between the skin and stiffener. Additionally, the cumulative error from closed-form solution, RR method, and Matlab precision contributes to the number too. The optimization result is also compared directly with a certain reference that uses the exact same material, load case, and geometry but using different objective function and slighty different approach. The result is the optimization from this research is able to find the stacking sequence that is up to 33% lighter with the relatively same critical buckling load compared with the reference. This concludes the objective function and the approach that is used in this research is more efficient and accurate than the reference.