Due to efficacy and optimization potential of genetic and evolutionary algorithms, they are used in learning and modeling especially with the advent of big data related problems. This book presents the algorithms and strategies specifically associated with pertinent issues in materials science domain. It discusses the procedures for evolutionary multi-objective optimization of objective functions created through these procedures and introduces available codes. Recent applications ranging from primary metal production to materials design are covered. It also describes hybrid modeling strategy, and other common modeling and simulation strategies like molecular dynamics, cellular automata etc. Features: Focuses on data-driven evolutionary modeling and optimization, including evolutionary deep learning. Include details on both algorithms and their applications in materials science and technology. Discusses hybrid data-driven modeling that couples evolutionary algorithms with generic computing strategies. Thoroughly discusses applications of pertinent strategies in metallurgy and materials. Provides overview of the major single and multi-objective evolutionary algorithms. This book aims at Researchers, Professionals, and Graduate students in Materials Science, Data-Driven Engineering, Metallurgical Engineering, Computational Materials Science, Structural Materials, and Functional Materials.

Artificial neural networks (ANNs) and genetic programming (GP) have already emerged as two very effective computing strategies for constructing data-driven models for systems of scientific and engineering interest. However, coming up with accurate models or meta-models from noisy real-life data is often a formidable task due to their frequent association with high degrees of random noise, which might render an ANN or GP model either over- or underfitted. This problem has recently been tackled in two emerging algorithms, Evolutionary Neural Net (EvoNN) and Bi-objective Genetic Programming (BioGP), which utilize the concept of Pareto tradeoff and apply a bi-objective genetic algorithm (GA) in the basic framework of both ANNs and GP. These concepts are elaborated in detail in this chapter.

Due to efficacy and optimization potential of genetic and evolutionary algorithms, they are used in learning and modeling especially with the advent of big data related problems. This book presents the algorithms and strategies specifically associated with pertinent issues in materials science domain. It discusses the procedures for evolutionary multi-objective optimization of objective functions created through these procedures and introduces available codes. Recent applications ranging from primary metal production to materials design are covered. It also describes hybrid modeling strategy, and other common modeling and simulation strategies like molecular dynamics, cellular automata etc. Features: Focuses on data-driven evolutionary modeling and optimization, including evolutionary deep learning. Include details on both algorithms and their applications in materials science and technology. Discusses hybrid data-driven modeling that couples evolutionary algorithms with generic computing strategies. Thoroughly discusses applications of pertinent strategies in metallurgy and materials. Provides overview of the major single and multi-objective evolutionary algorithms. This book aims at Researchers, Professionals, and Graduate students in Materials Science, Data-Driven Engineering, Metallurgical Engineering, Computational Materials Science, Structural Materials, and Functional Materials.

Focusing on techniques developed to evaluate the forming behaviour of tailor welded blanks (TWBs) in sheet metal manufacturing, this edited collection details compensation methods suited to mitigating the effects of springback. Making use of case studies and in-depth accounts of industry experience, this book gives a comprehensive overview of springback and provides essential solutions necessary to modern-day automotive engineers. Sheet metal forming is a major process within the automotive industry, with advancement of the technology including utilization of non-uniform sheet metal in order to produce light or strengthened body structures. This is critical in the reduction of vehicle weight in order to match increased consumer demand for better driving performance and improved fuel efficiency. Additionally, increasingly stringent international regulations regarding exhaust emissions require manufacturers to seek to lighten vehicles as much as possible. To aid engineers in optimizing lightweight designs, this comprehensive book covers topics by a variety of industry experts, including compensation by annealing, low-power welding, punch profile radius and tool-integrated springback measuring systems. It ends by looking at the future trends within the industry and the potential for further innovation within the field. This work will benefit car manufacturers and stamping plants that face springback issues within their production, particularly in the implementation of TWB production into existing facilities. It will also be of interest to students and researchers in automotive and aerospace engineering.

Predicting polymeric material behavior during processing and predicting final part properties continues to be a strong research focus within the scientific community as it involves taking into consideration a wide range of time-dependent variables. By use of data-driven modeling, the materials development process can be accelerated, and the highly predictive modeling techniques can facilitate the development of smart manufacturing systems. This dissertation worked on solving polymer engineering problems by use of data-driven modeling techniques. The first strategy was using data-driven modeling to provide a predictive model with statistical insights of the injection molding process to ensure part quality is maximized for a highly viscoelastic material blend. By injection molding highly viscoelastic materials, the probability of part defects is increased, therefore, it was crucial to use advanced computational techniques to understand the nuances of this highly non-linear process and to predict the outcome before creating material waste from faulty trials. The second strategy was in the use of data-driven modeling for reverse engineering purposes, specifically within materials development. By combining experimental characterization and data-driven modeling, algorithms were developed and compared to prove how highly predictive models can be used as reverse engineering toolboxes. This ultimately informed users of the optimal formulation which would reach the specified target material properties. The final strategy explored using data-driven modeling to validate the high influence of viscous heating within the pressure melt removal process, therefore, work was done in implementing a viscous heating system within a fused filament fabrication (FFF) 3D printer to accelerate the 3D printing process. The instrumented FFF 3D printer proved capable of accelerating print speeds and improving mechanical performance of 3D printed parts, working towards solving two of the largest bottlenecks within additive manufacturing: lead times and part quality. Given the unique capabilities of the data-driven modeling, the novel 3D printer was tested and evaluated via data-driven modeling to provide statistical information regarding which processing parameters were the most influential for improving overall performance of the 3D printing system. The results of this work provide a basis for future research endeavors related to combining data-driven modeling and polymer science, such as in optimizing the newly developed viscous heating 3D printer.

We exemplify and propose extending the use of genetic programs (GPs) – a genetic algorithm (GA) that evolves computer programs via mechanisms similar to genetics and natural selection – to symbolically regress key functional relationships between materials data, especially from electronic structure. GPs can extract structure–property relations or enable simulations across multiple scales of time and/or length. Uniquely, GP-based regression permits “data discovery” – finding relevant data and/or extracting correlations (data reduction/data mining) – in contrast to searching for what you know, or you think you know (intuition). First, catalysis-related materials correlations are discussed, where simple electronic-structure-based rules are revealed using well-developed intuition, and then, after introducing the concepts, GP regression is used to obtain (i) a constitutive relation between flow stress and strain rate in aluminum, and (ii) multi-time-scale kinetics for surface alloys. We close with some outlook for a range of applications (materials discovery, excited-state chemistry, and multiscaling) that could rely primarily on density functional theory results.

Engineering design problems can be discussed under the framework of decision making, namely, engineering design decisions. Inherently, accounting for uncertainty factors is an indispensable part of these decision processes. The goal of design decisions is to control or reduce the variational effect in decision consequences induced by many uncertainty factors. If we look at current technological trends, specifically the industry 4.0 movement, we can quickly appreciate the big push in science and technology for the digitalization of design, manufacturing, and management processes to reduce the amount of uncertainty present during innovation attempts. This work explores the value of data-driven integrated design and digital fabrication and how it allowed us to drive innovation in more than one domain. From examples in biomimicry discoveries to prostheses and unmanned aerial vehicle designs to the use of drones for emergency response, the key ideas of the proposed data-driven design paradigm are demonstrated. Earlier works on data-driven design and digital manufacturing have demonstrated its potential to disrupt the way we think about engineering design processes. However, constant modernization in these fields keeps pushing the boundaries on what is possible, and these territories remain relatively uncharted. This research aims to explore how a combination of spatial data sets can serve as a point of entry for data-driven innovative designs. The process starts with a different range of data acquisition tools and processing techniques, followed by computational analysis and optimization designs, all the way to digital manufacturing by means of 3D printing and validation via mechanical and functional testing. These data sets enabled the synthesis of digital twin models, which allowed us to begin a reverse engineering process for a series of multiple purposes. To begin our study, we focused on new methods of additive manufacturing with a special focus on composite 3D printing. We explored the current state of knowledge in the field of composite additive manufacturing. We investigated all different methods of 3D printing and the current broad range of materials available. We also gained a deep understanding of the different optimization opportunities that can be gained by incorporating fibers, chopped or continuous, into polymer filament additive manufacturing. Now that we know we can design and manufacture almost anything we can imagine we asked ourselves what would that be? At this point we explored new trends in the field of digital modeling, simulation, and optimization techniques. Starting in the cyber context we can create a digital twin that satisfies the objective functions of an engineering system such as lightweight, strong, controllable, manufacturable, and then use these objective functions as an opportunity to optimize over the design space. To prove this concept, we selected an engineering system design challenge: The design and optimization of a novel box wing vertical takeoff and landing aircraft intended to serve in long endurance environmental and archeological recognition missions as well as serving as the starting point for the development of the next generation of urban air mobility platforms. During the design of the Prandtl Box wing aircraft system we found that if we wanted to design better aeronautical systems, we needed to find a way to design lighter and stronger structures. This is the point when we decided to look into nature's library. We dived deep into biomimicry and proved how data and visualization driven research together with traditional mechanical testing, allows us to grasp a better insight on evolutionary optimization and its applications on structural design and material science. The ability to optimize and build stronger performing structures that follow form to function allowed us to add an extra level of complexity to our engineering system design. We added a human in the loop. This presented us with a unique set of functional requirements. We were faced with the challenge on how to translate these functional requirements into new objective functions. Using this new set of objective functions and the engineering system design methodology developed in previous studies we designed and tested a bioinspired transtibial prosthesis, which can be entirely 3D printed in a single piece allowing us to solve a global accessibility challenge. Additional work was done where we focused on the use of drones for emergency response after natural events and its applications on data-driven structural damage assessment during and after earthquakes. This work is not presented in this dissertation, but published material can be found online and the vita section of my thesis.