Introduction to quality engineering : designing quality into products and processes

This chapter contains sections titled: Learning Machine Statistical Learning Theory Types of Learning Methods Common Learning Tasks Model Estimation Review Questions and Problems References for further study

Learning from Data

https://repository.tudelft.nl/islandora/object/uuid%3Ae22dcf36-b9be-4b7a-b7fa-4626e8d5f393/datastream/OBJ/download

Methods for the prediction of fatigue delamination growth in composites and adhesive bonds: A critical review

Machine learning models are increasingly used in many engineering fields thanks to the widespread digital data, growing computing power, and advanced algorithms. The most popular machine learning model in recent years is artificial neural networks (ANN). Although many ANN models are used in the constitutive modeling of composite materials, there are still some unsolved issues that hinder the acceptance of ANN models in the practical design and analysis of composite materials and structures. Moreover, the emerging machine learning techniques are posing new opportunities and challenges in the data-based design paradigm. This paper aims to give a state-of-the-art literature review of ANN models in the constitutive modeling of composite materials, focusing on discovering unknown constitutive laws and accelerating multiscale modeling. This review focuses on the general frameworks, benefits, and challenges and opportunities of ANN models to the constitutive modeling of composite materials. Moreover, potential applications of ANN-based constitutive models in composite materials and structures are also discussed. This review is intended to initiate discussion of future research scope and new directions to enable efficient, robust, and accurate data-driven design and analysis of composite materials and structures.

A review of artificial neural networks in the constitutive modeling of composite materials

Effect of stress ratio or mean stress on fatigue delamination growth in composites: Critical review

Multilevel optimization including progressive failure analysis and robust design optimization for composite stiffened panels, in which the ultimate load that a post-buckled panel can bear is maximized for a chosen weight, is presented for the first time. This method is a novel robust multiobjective approach for structural sizing of composite stiffened panels at different design stages. The approach is integrated at two design stages labelled as preliminary design and detailed design. The robust multilevel design methodology integrates the structural sizing to minimize the variance of the structural response. This method improves the product quality by minimizing variability of the output performance function. This innovative approach simulates the sequence of actions taken during design and structural sizing in industry where the manufacture of the final product uses an industrial organization that goes from the material characterization up to trade constraints, through preliminary analysis and detailed design. The developed methodology is validated with an example in which the initial architecture is conceived at the preliminary design stage by generating a Pareto front for competing objectives that is used to choose a design with a required weight. Then a robust solution is sought in the neighbourhood of this solution to finally find the layup for the panel capable of bearing the highest load for the given geometry and boundary conditions.

/pdf/robust-design-and-optimization-of-composite-stiffened-panels-45qyw7cq6d.pdf

Robust design and optimization of composite stiffened panels in post-buckling

This paper focuses on Deterministic and Reliability Based Design Optimization (DO and RBDO) of composite stiffened panels considering post-buckling regime and progressive failure analysis. The ultimate load that a post-buckled panel can hold is to be maximised by changing the stacking sequence of both skin and stringers composite layups. The RBDO problem looks for a design that collapses beyond the shortening of failure obtained in the DO phase with a target reliability while considering uncertainty in the elastic properties of the composite material. The RBDO algorithm proposed is decoupled and hence separates the Reliability Analysis (RA) from the deterministic optimization. The main code to drive both the DO and RBDO approaches is written in MATLAB and employs Genetic Algorithms (GA) to solve the DO loops because discrete design variables and highly nonlinear response functions are expected. The code is linked with Abaqus to perform parallel explicit nonlinear finite element analyses in order to obtain the structural responses at each generation. The RA is solved through an inverse Most Probable failure Point (MPP) search algorithm that benefits from a Polynomial Chaos Expansion with Latin Hypercube Sampling (PCE-LHS) metamodel when the structural responses are required. The results led to small reductions in the maximum load that the panels can bear but otherwise assure that they will collapse beyond the shortening of failure imposed with a high reliability.

/pdf/reliability-based-design-optimization-of-composite-stiffened-3mmx2la1ai.pdf

Reliability-based design optimization of composite stiffened panels in post-buckling regime

A new multi-fidelity modelling-based probabilistic optimisation framework for composite structures is presented in this paper. The multi-fidelity formulation developed herein significantly reduces the required computational time, allowing for more design variables to be considered early in the design stage. Multi-fidelity models are created by the use of finite element models, surrogate models and response correction surfaces. The accuracy and computational efficiency of the proposed optimisation methodology are demonstrated in two engineering examples of composite structures: a reliability analysis, and a reliability-based design optimisation. In these two benchmark examples, each random design variable is assigned an expected level of uncertainty. Monte Carlo Simulation (MCS), the First-Order Reliability Method (FORM) and the Second-Order Reliability Method (SORM) are used within the multi-fidelity framework to calculate the probability of failure. The reliability optimisation is a multi-objective problem that finds the optimal front, which provides both the maximum linear buckling load and minimum mass. The results show that multi-fidelity models provide high levels of accuracy while reducing computation time drastically.

/pdf/a-novel-multi-fidelity-modelling-based-framework-for-3mt3lbieaa.pdf

A novel multi-fidelity modelling-based framework for reliability-based design optimisation of composite structures

A numerical model capable of dealing with progressive degradation of plain woven composites in a computationally efficient manner is presented in this article. A semi-analytical homogenization method is used to derive effective properties of the composite from the material properties of the constituents. The progressive failure is described using nonlocal continuum damage mechanics where the driving internal variable for the damage is the nonlocal strain. The model was implemented into Abaqus/Explicit, where the failure of a longitudinal tension and an open hole tension specimens were simulated in a multi-scale manner and verified experimentally.

Multi-scale failure analysis of plain-woven composites:

Advanced composites are increasingly being used as a structural material because of their balanced properties, higher impact resistance, and easier handling and fabrication compared with unidirectional composites. However, complex architecture of these composites leads to difficulties in predicting the mechanical response necessary for product design. Different methods for micromechanical analysis for the evaluation of effective mechanical properties of advanced composites are compared. Difficulties in modeling are highlighted and recommendations are given for micromechanical analysis using the finite element method.

Omar Bacarreza

Papers

Robust design and optimization of composite stiffened panels in post-buckling

Reliability-based design optimization of composite stiffened panels in post-buckling regime

A novel multi-fidelity modelling-based framework for reliability-based design optimisation of composite structures

Multi-scale failure analysis of plain-woven composites:

Micromechanical modeling of advanced composites