Chuck A. Baker

Response surface models combining linear and Euler aerodynamics for supersonic transport design

A method has been developed to efficiently implement supersonic aerodynamic predictions from Euler solutions into a highly constrained, muItidisciplinary design optimization of a High-Speed Civil Transport. The method alleviates the large computational burden associated with performing computational fluid dynamics analyses through the use of variable-complexity modeling techniques, response surface (RS) methodologies, and coarse-grained parallel computing. Using information gained from lower-fidelity aerodynamic models, reduced-term RS models representing a correction to the linear theory RS model predictions are constructed using Euler solutions. Studies into 5-, 10-, 15-, and 20-variable design problems show that accurate results can be obtained with the reduced-term models at a fraction of the cost of creating the full-term quadratic RS models. Specifically, a savings of 255 CPU hours out of 392 CPU hours required to create the full-term RS model is obtained for the 20-variable problem on a single 75MHz IP21 processor of a Silicon Graphics, Inc. Power Challenge.

A fully Distributed Parallel Global Search Algorithm

The n‐dimensional direct search algorithm, DIRECT, developed by Jones, Perttunen, and Stuckman has attracted recent attention from the multidisciplinary design optimization community. Since DIRECT only requires function values (or ranking) and balances global exploration with local refinement better than n‐dimensional bisection, it is well suited to the noisy function values typical of realistic simulations. While not efficient for high accuracy optimization, DIRECT is appropriate for the sort of global design space exploration done in large scale engineering design. Direct and pattern search schemes have the potential to exploit massive parallelism, but efficient use of massively parallel machines is non‐trivial to achieve. A fully‐distributed control version of DIRECT that is designed for massively parallel (distributed memory) architectures is presented. Parallel results are presented for a multidisciplinary design optimization problem – configuration design of a high speed civil transport.

VizCraft: a problem-solving environment for aircraft configuration design

The VizCraft problem-solving environment aids aircraft designers during conceptual design of a high-speed civil transport (HSCT). It integrates simulation codes that evaluate a design with visualizations for analyzing a design individually or in contrast to other designs. VizCraft provides a graphical user interface to a widely used suite of simulation and analysis codes for HSCT design, and it provides tools for visualizing the outputs of these codes. So, VizCraft provides an environment that combines visualization and computation, encouraging the designer to think in terms of the overall problem-solving task, not simply using the visualization to view the computations results.

HSCT configuration design using response surface approximations of supersonic Euler aerodynamics

A method has been developed to efficiently implement supersonic aerodynamic predictions from Euler solutions into a highly constrained, multidisciplinary design optimization of a High-Speed Civil Transport (HSCT) configuration. The method alleviates the large computational burden associated with performing CFD analyses and eliminates the numerical noise present in the analyses through the use of response surface (RS) methodologies, a variation of the variable-complexity modeling (VCM) technique, and coarse grained parallel computing. Variablecomplexity modeling techniques allow one to take advantage of information gained from inexpensive lower fidelity models while maintaining the accuracy of the more expensive high fidelity methods. In this research, simple conceptual level aerodynamic models provide the functional form of the drag polar. Response surface models are therefore created for the intervening functions (drag polar shape parameters) revealed by the simple models instead of for the drag itself. Optimization results using linear theory RS models are used to select the allowable ranges of the design variables. Stepwise regression analysis, performed using data from linear theory aerodynamic results, provides information on the relative importance of each term in the polynomial RS models. With this information, reduced term RS models representing a correction to the linear theory RS model predictions are constructed using fewer Euler evaluations. Studies into five, ten, fifteen, and ∗Graduate Research Assistant, Dept. of Aerospace and Ocean Engineering. Current Position: Postdoctoral Research Associate, Dept. of Aeronautics and Astronautics, University of Washington, Seattle, WA, Member AIAA. †Postdoctoral Research Associate, National Research Council/NASA Langley Research Center, Hampton, VA, Member AIAA. ‡Graduate Research Assistant, Dept. of Aerospace and Ocean Engineering, Student Member AIAA. §Professor and Dept. Head of Aerospace and Ocean Engineering, Associate Fellow AIAA. ¶Professor of Aerospace and Ocean Engineering, Associate Fellow AIAA. ‖Professor of Aerospace Engineering, Mechanics and Engineering Science, University of Florida, Gainesville, FL, Fellow AIAA ∗∗Professor of Computer Science and Mathematics twenty variable HSCT design problems show that accurate results can be obtained with the reduced term models at a fraction of the cost of creating the full term quadratic RS models. Specifically, 11 hour, 47 hour, 115 hour, and 255 hour savings in CPU time on a single 75 MHz IP21 processor of a SGI Power Challenge are obtained for the five, ten, fifteen, and twenty variable design problems, respectively. Errors in the RS model cruise drag predictions, based on actual Euler calculations, for the optimal designs range from 0.1 counts to 0.8 counts for the twenty variable optimum.

Distributed Control Parallelism in Multidisciplinary Aircraft Design

Multidisciplinary design optimization (MDO) for large-scale engineering problems poses many challenges (e.g., the design of an efficient concurrent paradigm for global optimization based on disciplinary analyses, expensive computations over vast data sets, etc.) This work focuses on the application of distributed schemes for massively parallel architectures to MDO problems, as a tool for reducing computation time and solving larger problems. The specific problem considered here is configuraton optimization of a high speed civil transport (HSCT), and the efficient parallelization of the embedded paradigm for reasonable design space identification. Two distributed dynamic load balancing techniques (random polling and global round robin with message combining) and two necessary termination detection schemes (global task count and token passing) were implemented and evaluated in terms of effectiveness and scalability to large problem sizes and a thousand processors. The effect of certain parameters on execution time was also inspected. Empirical results demonstrated stable performance and effectiveness for all schemes, and the parametric study showed that the selected algorithmic parameters have a negligible effect on performance.

High-Speed Civil Transport Design Space Exploration Using Aerodynamic Response Surface Approximations

A method has been developed to generate and use polynomial approximations to the range and cruise drag components in a highly constrained, multidisciplinary design optimization of a high-speed civil transport (HSCT) cone guration. The method improves optimization performance by eliminating the numerical noise present in the analyses through the use of response surface methodology. In this implementation quadratic polynomials are e t within variable bounds to data gathered from a series of numerical analyses of different aircraft designs. Because the HSCT optimization process contains noise and suffers from a nonconvex design space even when noise is e ltered out, multiple optimization runs are performed from different starting points with and without the response surface models in order to evaluate both their effectiveness as surrogate functions and as a design exploration tool. The alternative method used is variable complexity modeling (VCM). It is shown that response surface methodology facilitates design space exploration, allowing improvements in terms of both convergence performance and computational effort when multiple starting points are required, although using VCM usually produces better e nal designs.

VizCraft: a multidimensional visualization tool for aircraft configuration design

We describe a visualization tool to aid aircraft designers during the conceptual design stage. The conceptual design for an aircraft is defined by a vector of 10-30 parameters. The goal is to find a vector that minimizes an objective function while meeting a series of constraints. VizCraft integrates the simulation code that evaluates the design with visualizations for analyzing the design individually or in contrast to other designs. VizCraft allows the designer to easily switch between the view of a design in the form of a parameter set, and a visualization of the corresponding aircraft. The user can easily see which, if any, constraints are violated. VizCraft also allows the user to view a database of designs using parallel coordinates.

STUDY OF A GLOBAL DESIGN SPACE EXPLORATION METHOD FOR AEROSPACE VEHICLES

The preliminary design space exploration for large, interdisciplinary engineering problems is often a difficult and time-consuming task. General techniques are needed that efficiently and methodically search the design space. This work focuses on the use of parallel load balancing techniques integrated with a global optimizer to reduce the computational time of the design space exploration. The method is applied to the multidisciplinary design of a High Speed Civil Transport (HSCT). A modified Lipschitzian optimization algorithm generates large sets of design points that are evaluated concurrently using a variety of load balancing schemes. The load balancing schemes implemented in this study are: static load balancing, dynamic load balancing with a master-slave organization, fully distributed dynamic load balancing, and fully distributed dynamic load balancing via threads. All of the parallel computing schemes have high parallel efficiencies. When the variation in the design evaluation times is small, the computational overhead needed for fully distributed dynamic load balancing is substantial enough so that it is more efficient to use a master-slave paradigm. However, when the variation in evaluation times is increased, fully distributed load balancing is the most efficient. * Graduate Research Assistant, Dept. of Aerospace

Distributed control parallelism for multidisciplinary design of a high speed civil transport

Large scale multidisciplinary design optimization (MDO) problems often involve massive computation over vast data sets; Regardless of the MDO problem solving methodology, advanced computing technologies and architectures are indispensable. The data parallelism inherent in some engineering problems makes massively parallel architectures a natural choice, but efficiently harnessing the power of massive parallelism requires sophisticated algorithms and techniques. This paper presents an effort to apply massively scalable distributed control and dynamic load balancing techniques to the reasonable design space identification phase of a variable complexity approach to the multidisciplinary design optimization of a high speed civil transport (HSCT). The scalability and performance of two dynamic load balancing techniques, random polling and global round robin with message combining, and two termination detection schemes, token passing and global task count, are studied. The extent to which such techniques are applicable to other MDO paradigms, and to the potential for parallel multidisciplinary design with current large-scale disciplinary codes, is of particular interest.