|
|
DESIGN, STRESS ANALYSES AND LIMIT LOAD OF SANDWICH STRUCTURES
Löffelmann, František ; Pavlínek,, Vladimír (oponent) ; Klement, Josef (oponent) ; Píštěk, Antonín (vedoucí práce)
The thesis starts with a review of design calculations of sandwich beams, plates, and complicated structures, where FEM plays an important role. Next, optimization methods are reviewed to shed light on the wide area of mathematical programming and basic topology optimization principles up to its implementation by other authors in composite design, including representative examples of analytical and numerical optimization of sandwiches. The thesis objective is defined as an implementation of mass minimization with failure constraints aiming to make the sandwich design process easier. This is done by own implementation of gradient optimization based on topology optimization principles, known as Discrete Material Optimization (DMO), which helps to find optimal layup. Approach to material interpolation and failure constraints interpolation is developed and programmed in Python, using First Order Shear Deformation Theory (FSDT) to evaluate stresses on elements, based on element loads given by the Nastran FE solver. Gradient optimizer searches for optimal materials for each layer of the sandwich face-sheet and core from the user-defined candidates. The program is tested on examples of sequential complexity from one-element beams where the true optimum is known up to a practical task of the sandwich galley from an airliner. Results have shown that the algorithm can reach a discrete solution without (significant) violation of constraints and thus can be practically used to make conceptual sandwich design more efficient.
|
|
|
DESIGN, STRESS ANALYSES AND LIMIT LOAD OF SANDWICH STRUCTURES
Löffelmann, František ; Pavlínek,, Vladimír (oponent) ; Klement, Josef (oponent) ; Píštěk, Antonín (vedoucí práce)
The thesis starts with a review of design calculations of sandwich beams, plates, and complicated structures, where FEM plays an important role. Next, optimization methods are reviewed to shed light on the wide area of mathematical programming and basic topology optimization principles up to its implementation by other authors in composite design, including representative examples of analytical and numerical optimization of sandwiches. The thesis objective is defined as an implementation of mass minimization with failure constraints aiming to make the sandwich design process easier. This is done by own implementation of gradient optimization based on topology optimization principles, known as Discrete Material Optimization (DMO), which helps to find optimal layup. Approach to material interpolation and failure constraints interpolation is developed and programmed in Python, using First Order Shear Deformation Theory (FSDT) to evaluate stresses on elements, based on element loads given by the Nastran FE solver. Gradient optimizer searches for optimal materials for each layer of the sandwich face-sheet and core from the user-defined candidates. The program is tested on examples of sequential complexity from one-element beams where the true optimum is known up to a practical task of the sandwich galley from an airliner. Results have shown that the algorithm can reach a discrete solution without (significant) violation of constraints and thus can be practically used to make conceptual sandwich design more efficient.
|
host ::
RSS.