Advances on Links Between Mathematics and Industry

CTMI 2019

herausgegeben von: Prof. Peregrina Quintela Estévez, Prof. Bartomeu Coll, Assoc. Prof. Rosa M. Crujeiras, Prof. José Durany, Prof. Laureano Escudero

Verlag: Springer International Publishing

Buchreihe : Sxi — Springer per l’Innovazione / Sxi — Springer for Innovation

Enthalten in: Springer Professional "Wirtschaft+Technik" , Springer Professional "Technik" , Springer Professional "Wirtschaft"

Einloggen, um Zugang zu erhalten

Über dieses Buch

This book results from the talks presented at the First Conference on Transfer between Mathematics & Industry (CTMI 2019). Its goal is to promote and disseminate the mathematical tools for Statistics & Big Data, MSO (Modeling, Simulation and Optimization) and their industrial applications. In this volume, the reader will find innovative advances in the automotive, energy, railway, logistics, and materials sectors. In addition, Advances CTMI 2019 promotes the opening of new research lines aiming to provide suitable solutions for the industrial and societal challenges. Fostering effective interaction between Academia and Industry is our main purpose with this book. CTMI conferences are one of the main forums where significant advances in industrial mathematics are presented, bringing together outstanding leaders from business, science and Academia to promote the use of mathematics for an innovative industry.

Inhaltsverzeichnis

Frontmatter

Recent Advances in Computational Models for the Discrete and Continuous Optimization of Industrial Process Systems

Abstract

An overview of the mathematical formulations used for discrete and continuous optimization are presented. These include Linear Programming, Nonlinear Programming, Integer Programming, Mixed-Integer Linear Programming, Mixed-Integer Nonlinear Programming, Logic-based Optimization, Stochastic Programming, Robust Optimization, and Flexibility Analysis. Successful applications of optimization models in industry are presented in the following fields: upstream oil & gas, materials blending, natural gas, biofuels, water treatment, electricity market integration, plant reliability, and supply chain design. Ongoing projects applying computational models to optimize industrial process systems are also mentioned. Implementations of customized optimization techniques that improve computational performance and enable finding solutions to otherwise unsolvable optimization problems are highlighted. These include strengthening cuts, decomposition strategies, model reformulation, and linearization, among others.

Hector D. Perez, Ignacio E. Grossmann

Optimal Design of a Railway Bypass at Parga, Northwest of Spain

Abstract

In the past few years, different models have been proposed to obtain the optimal design of a linear infrastructure (road or railway) connecting two given points. This paper analyses the usefulness of one of these models to design a railway bypass, in a real case study, on the railway line A Coruña-Palencia (Spain), where it passes through the urban area of Parga. Firstly, to show the good performance of the model, it is applied to a situation whose solution is known “a priori”. Next, the problems arising in the real case are shown, where it is quite complicated to design a layout avoiding existing buildings and other restricted areas, and a two-stage method is proposed generating the layout of the requested bypass. Throughout the development of this method, it arises the need to connect a given circular curve with an also given tangent. To address this problem, an algorithm is proposed for computing the transition curve (clothoid) performing that connection. The work ends with some interesting conclusions and a brief description of future work.

Gerardo Casal, Alberte Castro, Duarte Santamarina, Miguel E. Vázquez-Méndez

Reduced Models for Liquid Food Packaging Systems

Abstract

Simulation tools are nowadays key elements for effective production, design and maintenance processes in various industrial applications. Thanks to the advances that have been achieved in the past three decades, accurate and efficient solvers for computational fluid dynamics and computational mechanics are routinely adopted for the design of many products and systems. However, the most accurate models accounting for the complete three-dimensional complex physics (of even multi-physics) are not always the best option to pursue, in particular in the preliminary design phase or whenever very fast evaluations are required. In this paper, we present a set of reduced numerical models that have been developed in the past few years to support the design of paperboard packaging systems.

Nicola Parolini, Chiara Riccobene, Elisa Schenone

Reduced-Order Modeling in the Manufacturing Process of Wire Rod: Applications for Fast Temperature Predictions and Optimal Selection of Process Parameters

Abstract

The number of operational variables that determines the cooling process of steel wire, given by the conveyor velocity and the different fan sections powers (controlled independently), lead to a dependency of the cooling on a high multidimensional parameter space whose potential combinations are impossible to be analyzed, either experimentally or by numerical simulation of a thermal–metallurgical model. To tackle this problem, an efficient strategy, based on the use of Higher Order Singular Value Decomposition (HOSVD), is presented. The approach presented provides a Reduced-Order Model (ROM) capable of predicting quite accurately the cooling curve for any combination of the process parameters. Fast online predictions of the cooling rates allow to incorporate accurate modeling results in many Engineering tools, such as model predictive control algorithms or plant simulation software. Also, the ROM in combination with an optimization tool finds the adequate operational parameters with significant reduction of energy consumption.

Elena B. Martín, Fernando Varas, Iván Viéitez

Modeling and Numerical Simulation of the Quenching Heat Treatment. Application to the Industrial Quenching of Automotive Spindles

Abstract

The quenching heat treatment consists in the immersion of a steel piece (previously heated up to the austenization temperature range) in fluid. The fast cooling undergone by the piece induces microstructure transformations (from austenite to usually martensitic microstructure) aimed to provide the piece with specific mechanical properties (high hardness). The numerical model needed to mimic the cooling process and, therefore, to predict the final crystallographic structure, involves the following strongly coupled problems: a two-phase turbulent thermo-fluid-dynamic model (due to the presence of liquid and vapor caused by the high solid temperatures), and a thermal-metallurgical model for the piece. In this work, the heat flow on the surface of the spindle is characterized using a compilation of correlations (based on the specialized literature and also adjusted from simplified experiments and/or simulations) aimed to describe the different heat transfer mechanisms, extensively described in Nukiyama’s experiments. This approach allows to describe (up to a certain degree of accuracy) the cooling process without solving a complex fluid-dynamic multiphase model, and hence in a computationally affordable way. The final model is eventually used to optimize the manufacturing parameters of the steel industrial quenching process of spindles in the automotive industry.

Carlos Coroas, Elena B. Martín

Single Particle Models for the Numerical Simulation of Lithium-Ion Cells

Abstract

In the design of Battery Management Systems (BMS) for a lithium-ion cell, it is crucial to accurately simulate the device in real time using mathematical models. Often, Equivalent Circuit Models (ECM) are used to this end, due to their simplicity and efficiency. However, they are purely phenomenological (their parameters are fitted to emulate empirical data) and their internal variables lack physical meaning. On the other hand, the most popular physics-based electrochemical model in the literature, the pseudo-two-dimensional (P2D) model, presents a high computational cost. In this paper, we review the single particle model (SPM), a physics-based model of reduced complexity that is suitable for real-time applications.

Alfredo Ríos-Alborés, Jerónimo Rodríguez

Fracture Propagation Using a Phase Field Approach

Abstract

Phase field models have received a lot of attention during the last 20 years and they have reached maturity, being in the last years ubiquitous, finding nice examples in a range of applications in physical sciences and engineering, from the classical spinodal decomposition in multiphase flows to qualitative studies of motility in metastatic tumor cells. In this paper, we give a brief introduction to the theory of the method and present a review of some striking applications in order to show the huge potential and versatility of the technique. In this work, we are interested in very specific problems related to applications in energy storage and fracture dynamics. This research began by using a simple model to study the propagation of fractures in elastic homogeneous materials and eventually evolved into a coupled model that includes fracture propagation and flow in elastic-porous media.

David Casasnovas, Ángel Rivero

Phase Space Learning with Neural Networks

Abstract

This work proposes an autoencoder neural network as a non-linear generalization of projection-based methods for solving Partial Differential Equations (PDEs). The proposed deep learning architecture presented is capable of generating the dynamics of PDEs by integrating them completely in a very reduced latent space without intermediate reconstructions, to then decode the latent solution back to the original space. The learned latent trajectories are represented and their physical plausibility is analysed. It is shown the reliability of properly regularized neural networks to learn the global characteristics of a dynamical system’s phase space from the sample data of a single path, as well as its ability to predict unseen bifurcations.

Jaime López García, Ángel Rivero

Backmatter

Titel: Advances on Links Between Mathematics and Industry
herausgegeben von: Prof. Peregrina Quintela Estévez
Prof. Bartomeu Coll
Assoc. Prof. Rosa M. Crujeiras
Prof. José Durany
Prof. Laureano Escudero
Verlag: Springer International Publishing
Electronic ISBN: 978-3-030-59223-3
Print ISBN: 978-3-030-59222-6
DOI: https://doi.org/10.1007/978-3-030-59223-3