KEYNOTE LECTURERS

Keynote Lecture 1: Tuesday, January 20, 2009, 11:00-12:00, Main Building MG-02

Tow-placed Unconventional Composite Laminates-- Design and Optimization

Prof. Zafer Gürdal
Aerospace Structures Chair
Faculty of Aerospace Engineering
Delft University of Technology, The Netherlands

Abstract

Developments in automated composite production technologies enable rapid penetration of composite materials into many different products. In addition to providing a fabrication process that reduces the production cycle time, automated processes are natural solution to address problems associated with variability in quality resulting from the manual processes, and provide a repeatable and improved quality component production. Automated fibre/tow-placement machines in particular provide a technology that enable production of complex shapes, but also result in high quality parts. The real advantage of such processes for designers, however, is the possibility to use unconventional layups for components, and therefore better exploit tailoring potential of composite materials. By unconventional laminates two different constructions of laminates are implied. One of these constructions is the layups with ply orientation angles other than traditional 0, ±45, and 90 degrees. These layups may include fiber orientation angles dispersed with small orientation increments, such as 5 to 10 degrees or smaller, that will be assumed constant over skin laminates. The second construction are laminates with steered fibers that will produce continuous variation of ply orientation angles over the planform of skin laminates of thin-walled structures. Such a construction is often referred to as a variable stiffness laminate, though besides the stiffness properties any other fiber orientation depended property of the laminate, such as coefficient of thermal expansion, also varies as a function of location in the laminate.
After providing a brief introduction of unconventional laminates and their advantages, the main emphasis of the talk will be a new design methodology that allows design optimization of large scale composite parts with large number of discrete orientation angles. The new methodology, which is referred to as a two step design framework, takes advantage of a guide-based design approach for blended composite laminates and is a computationally efficient algorithm. In the first step of this two step approach, the panel level optimization is done using mathematical optimization techniques with lamination parameters as continuous design variables using optimality criteria. The lamination parameters provide a simple way of dictating or representing the stiffness properties of a multi-layered laminate with only a few variables and also eliminate the need for discrete optimization. The continuous optimization is carried out using a successive convex approximation method based on a recently developed novel approximation scheme. After the optimal distribution of lamination parameters is obtained, a guide-based Genetic Algorithm approach is applied in the second step to generate the fully blended solution with discrete stacking sequence as design variables.

Biography

Prof. Gürdal received his B.S. in Mechanical Eng., Middle East Technical University 1979, M.S. in Mechanical & Aerospace Eng., Illinois Institute of Technology 1981, and his Ph.D. in Aerospace & Ocean Eng, Virginia Polytechnic Institute and State University 1985. He subsequently joined the Engineering Science & Mechanics Department as an Assistant Professor, and became a Full Professor in 1995. He held a jointly appointed position of the Professor of Aerospace & Ocean Eng. and Engineering Science & Mechanics at Virginia Tech until 2006. In 2004 Prof. Gürdal was appointed as the Chair Holder of Aerospace Structures at TU Delft in The Netherlands. He also holds an honorary Professor Emeritus appointment at Virginia Tech.
Prof. Gürdal’s research interests are in structural and multidisciplinary design and optimization, design and optimization of composite materials and structures, adaptive structures, buckling postbuckling of thin-walled structures, global/local design methodologies for optimization of large complex systems, and computational methods for design. His research has been largely funded by NASA Langley Research Center and Air Force Office of Scientific Research. Research funding from various companies include Sikorsky Aircraft, Ford Motor Co., Schneider Electric, Boeing Defense and Space, Mc. Donnell Douglas, Boeing Helicopter, Lockheed Martin, Newport News Shipbuilding, Material Science Corporation, and ALCOA. He also worked with a number of small companies, and is one of the founders of ADOPTECH, Inc., a small business in Blacksburg Virginia.
Prof. Gürdal’s research contributions resulted in more than 200 publications. He is a co-author of 3 books; “Design and Optimization of Laminated Composite Materials.” John Wiley & Sons, Inc., “Elements of Structural Optimization”, 3rd Ed., Kluwer Academic Publishers., “Optimal Design: Theory and Applications to Material and Structures.” Technomic Publishing Co., Inc. Prof. Gürdal taught several AIAA Professional Development Short courses. He served as the graduate thesis advisor for 32 masters and 17 doctoral students. He is a Lifetime member and Associate Fellow of the American Institute of Aeronautics and Astronautics (AIAA), and one of the founding members of the AIAA Multidisciplinary Technical Committee (MDO-TC).

Keynote Lecture 2: Wednesday, January 21, 2009, 9:00-10:00, Main Building MG-02

Design by Shopping using Multi-Solution Genetic Algorithms

Prof. Richard Balling
Department of Civil and Environmental Engineering
Brigham Young University, USA

Abstract

Many optimization algorithms produce a single optimum design. The human designer is then left with the choice of either accepting or rejecting that design. Since genetic algorithms operate on populations of designs, the potential exists for the genetic algorithm to produce a population of diverse optimum designs. The human designer can then "shop" for a design, and may even use non-quantifiable preferences in the decision-making process. There are many advantages to this paradigm, where the computer serves as preprocessor to decision-making.
The talk will describe two types of Multi-Solution Genetic algorithms. The first type includes Multi-Objective Genetic Algorithms (MOGA's). MOGA's will be illustrated with an application to city planning of a high-growth metropolitan region. The second type includes Multi-Concept Genetic Algorithms (MCGA's). MCGA's will be illustrated with an application to the conceptual design of structural frames and trusses.

Biography

Dr. Richard Balling is a professor of Civil and Environmental Engineering at Brigham Young University, USA. He received his doctorate at the University of California at Berkeley in 1982. He has published over 100 peer-reviewed articles on the application of optimization to structural engineering, city planning, aircraft design, mechanical design, fluid flow, geometric packaging, software design, and multidisciplinary collaboration. He has twice been the co-recipient of the American Society of Civil Engineering (ASCE) State-of-the-Art Award. He has served as associate editor of the ASCE Journal of Structural Engineering. He is co-author of the OPTDES.BYU and OPTDESX optimization software packages that have been licensed to hundreds of industrial companies, and he has been the instructor at several short courses on engineering optimization. He currently directs a study abroad program titled "China Megastructures" where he takes engineering students to China to study the remarkable structures there. He came to Dubai to see if he should offer a "Dubai Megastructures" program.

Keynote Lecture 3: Thursday, January 22, 2009, 9:00-10:00, Main Building MG-02

Distributed Parameter Systems: Hybrid Dynamic Inequalities and Applications

Dr. Gangaram Ladde
Department of Mathematics and Statistics
University of South Florida, USA

Abstract

In this work, a mathematical model for interconnected distributed parameter dynamic phenomenon evolving under different measure chains with state dependent discrete events is formulated. By introducing an arbitrary pair of functional of a pair of flows (measured dynamic flows), a system of systems of dynamic inequalities with corresponding comparison hybrid dynamic system is outlined. An arbitrary pair of functional of dynamic flows evolving in two different time scales satisfying a system of systems of hybrid dynamic inequalities is estimated by the corresponding comparison system of systems of impulsive hybrid dynamic equations. Moreover, employing vector Lyapunov/energy functions as functionals of hybrid dynamic flows, several comparison results are developed to estimate solution processes of nonlinear non stationary hybrid dynamic system in systematic and coherent manner. The obtained results extend and generalize the existing results in a systematic and unified way.

Biography

Dr. Gangaram Ladde received his B. Sc. (First Class) in Chemistry, Mathematics and Physics, from People's College, Nanded (India) in 1963, M. Sc. (First Class First) in Mathematics from Marathwada University, Aurangabad (India) in 1965, and Ph. D. in Mathematics from University of Rhode Island in 1972. Dr. Ladde joined the State University of New York at Potsdam in 1973 and was promoted to the Tenured Full Professorship in 1980; he relocated to the University of Texas at Arlington in 1980. Dr. Ladde's research interests are Stochastic Time Analysis and Estimation Theory: Stability Theory, Control Theory, Differential Games, Filtering Theory, Oscillation Theory, and Qualitative Analysis of Competitive-Cooperative Process in Biological, Physical and Social Sciences under random environmental perturbations. He has published more than 150 papers, has co-authored 4 monographs, and co-edited 6 proceedings of international conferences, notably, (i) Stochastic Versus Deterministic Systems of Differential Equations, (with M. Sambandham), Marcel Dekker, Inc, New York, 2004 and (ii) Random Differential Inequalities (with V. Lakshmikantham), Academic Press, New York, 1980. Dr. Ladde is the Founder and Joint Editor-in Chief (1983-present) of an International Journal of Stochastic Analysis and Applications. He is also a Member of Editorial Board of several journals in Mathematical Sciences. Dr. Ladde is recipient of several research awards and grants.