New approaches to high speed civil transport multidisciplinary design and optimization

New Approaches to High Speed Civil Transport Multidisciplinary Design and Optimization
P. Scott Zink, Dan A. DeLaurentis, Mark A. Hale, Vitali V. Volovoi Daniel P. Schrage, James I. Craig, Robert E. Fulton, Farrokh Mistree, Dimitri N. Mavris Georgia Institute of Technology Atlanta, GA 30332-0150 Wei Chen University of Illinois Kemper E. Lewis University at Buffalo Carlos E.S. Cesnik Massachusetts Institute of Technology Peter J. R?hl General Electric Corporate Research and Development Patrick N. Koch Engineous Software, Incorporated Timothy W. Simpson The Pennsylvania State University
ABSTRACT
New approaches to multidisciplinary design optimization have been developed and demonstrated in a NASA Langley funded project to look at a particularly challenging aeronautics problem – High Speed Civil Transport Aeroelastic Wing Design. In particular, the problem is how to incorporate key multidisciplinary information and knowledge as it becomes available. This problem results from such information not being available early in the design process due to dependence on higher fidelity models, databases and analysis tools. A Robust Design Synthesis methodology involving the application of response surface equations was developed and successfully applied in the solution of this problem. The approach also involved the incorporation of stochastic models for the design parameters and the development of tools for their propagation through the approximations generated for the design space under consideration. The research addresses several key wing design domains including aerodynamics, structural analysis and controls as necessary to address the wing aeroelasticity and transonic flutter problems, including the potential application to active flexible wing technologies. AFW ASOP DOE DREAMS DSIDES DSP DSPT FIDO FLOPS GSRP HSCT HSR IDEF IDES IMAGE ISMD MDO RCEM RDS RSE RSM SST
ACRONYMS
Active Flexible Wing Affordable Systems Optimization Process Design of Experiments Developing Robust Engineering Analysis Models and Specifications Decision Support in the Design of Engineering Systems Decision Support Problem Decision Support Problem Technique Framework for Interdisciplinary Design Optimization Flight Optimization System Graduate Student Research Program High Speed Civil Transport High Speed Research Integration Definition Integrated Design Engineering Simulator Intelligent Multidisciplinary Aircraft Generation Environment Integrated Structure/Maneuver Design Multidisciplinary Design Optimization Robust Concept Exploration Method Robust Design Synthesis Response Surface Equation Response Surface Methodology Supersonic Transport
TABLE OF CONTENTS ABSTRACT 1. INTRODUCTION 2. MDO METHODOLOGY 3. FOCUS AREAS IN MDO SOLUTION 4. CLOSURE 6. REFERENCES 7. ACKNOWLEDGEMENTS 8. AUTHORS 1 2 2 3 8 9 12 12

1. INTRODUCTION
Multidisciplinary design optimization (MDO) addresses the considerable challenge of concurrently incorporating analysis models and design parameters from several different discipline areas into a design synthesis process that is implemented using powerful decision-support tools. Aerospace systems are inherently multidisciplinary in nature, and therefore MDO is a key part of the design process. However, it has been only in the last two decades that the problem complexities have risen well beyond the human abilities of the individual designer or the traditional design team. As a result, new emphasis has been placed on the development of powerful, flexible and robust MDO methods applicable to high-fidelity problems. The initial efforts towards MDO focused on specific disciplinary interactions that were growing troublesome to handle or that had immediate and profound impact on the design process, as illustrated in Figure 1. While there are still significant challenges in these areas, the problem addressed in the present research involves the development of MDO methods to handle aerospace vehicle synthesis and sizing at the conceptual/preliminary phases of design.
“Traditional” MDO
Aero
Year 1
Year 2
Year 3
Problem Description
? AFW MDO Structure ? DSPT Research ? Computing Requirements
Problem Development ? Design Framework ? DSP Palette ? Analysis Tools
Problem Solution ? Analysis & RSE Integration ? IMAGE Architecture ? Design Exploration
Figure 2. Three Year Research Schedule The Year 1 effort focused on identification of a specific (and academically “tractable”) portion of a broader scope of HSCT design. The initial attention was on the HSCT wing design with consideration of product and process development aspects [50], but the focus shifted towards the multidisciplinary effort to handle the aeroelastic design of the wing. Year 1 was also spent on adaptation and development of basic decision support methods to MDO and on the development of computing requirements for a practical system, one in which information can be obtained for decision-making from high-fidelity tools. The Year 2 effort involved the further development of the wing design framework, the development of specific classes of decision support, and the identification and development of specific analysis tools. The Year 3 effort centered on the incorporation of robust design simulation methods involving the use of response surface equations (RSE) to bring highfidelity, discipline specific analysis and modeling methods forward into conceptual design studies from their more traditional places in subsystem level preliminary design. The methods and tools were then tested and evaluated in sample MDO studies using a design computing framework called IMAGE (Intelligent Multidisciplinary Aircraft Generations Environment) [16].
Conceptual/Preliminary Vehicle Sizing & Synthesis
Struct. Aero
Discipline interactions optimizing discipline objectives (Ww, Cd, etc.)
Aero Prop.
Struct.
Control Control Prop.
? The REAL objectives $, range, weight, size ? Mission Requirements ? Evaluate Technologies ? Systems Approach
Barrier
Figure 1. Present Solution thrust Areas Contrasted to Traditional MDO The research presented here was a planned three year effort and was aimed first at the description of a High Speed Civil Transport (HSCT) MDO problem (described in the next section as one that has conflicting objectives of cost and performance and requires high-fidelity analyses to solve), next at the development of MDO methods for the solution of the problem, and finally at the implementation of a solution to a significant portion of the problem. These phases are illustrated in Figure 2.
2. MDO METHODOLOGY
The MDO methodology utilized in this project is illustrated in Figure 3 as a hierarchical system decomposition. The complex problem of finding good designs for a flexible HSCT wing is based on the combined (and generally conflicting) objectives of minimum cost and maximum performance. The solution of this problem requires the combined analysis capabilities from the aerodynamics, structures, and controls disciplines. In addition, the simulation is multi-leveled, with objectives, such as takeoff gross weight and cost, calculated at the system level through sizing and synthesis, and with subsystem objectives, such as wing weight and lift to drag ratio, distributed in the subsystem level disciplines. The contributing analyses introduced through RSEs allow a designer to perform tradeoffs in terms of the size of the design space searched and complexity of the tools used. The left half flow in Figure 3 illustrates the general problem solution process following a decision support process referred to here as the Decision Support Problem Technique (DSPT) [66]. It also illustrates how the subsystem

objectives are related to system goals through the use of RSEs. The right hand flow in Figure 3 illustrates how this process was implemented for an aeroelastic wing design problem. These MDO methodologies were coordinated and implemented in an infrastructure and integration project which was initially identified as the Integrated Design Engineering Simulator (IDES) but later came to be known as DREAMS (Developing Robust Engineering Analysis Models and Specifications) [25]. The methods were implemented in an open computing infrastructure deployed at Georgia Tech that facilitates the design of complex engineering systems called IMAGE.
Generic Problem Solution Process Design Satisficing Solution Aeroelastic Wing Design Implementation
Through the use of response surfaces, analysis-oriented methods are incorporated into DREAMS requiring only a one time investment for a given class of vehicles. Further, increased design freedom and knowledge, as well as reduced cycle time at the conceptual level, make the resulting analysis portion of DREAMS more amenable to implementation in IMAGE, the computer framework in which the whole process is implemented. 3.1 HSCT Approaches Wing aeroelasticity and the calculation of flutter during transonic flight has been the “long pole in the tent” for the development of supersonic aircraft. During the Supersonic Transport (SST) development of the 1960s and early 1970s aeroelastic calculation of flutter for closing the loop with design was left open – to be resolved during prototype development. The same relates to the recently terminated NASA/Industry High Speed Research (HSR) program. Because of the importance of flutter in viability of these programs, this particular aspect of the HSCT design problem was the subject for the methodology evaluation. Thus it can be seen that an appropriate grand challenge was being tackled. Wing design with consideration of both static and dynamic aeroelasticity presents the classical MDO challenge to designers working at the conceptual level: how to incorporate key multidisciplinary information and knowledge that is usually not available early in the design process due to its dependence on higher fidelity models, databases and analysis tools. For example, wing aeroelasticity studies and transonic flutter calculations require much more detailed (higher fidelity) structural and aerodynamic models than are typically available at the conceptual phase of design. Yet it is well understood that it is at this early stage when these complex problems can be most effectively and economically addressed. The new MDO approaches for HSCT wing design directly address the challenge of providing higher fidelity design models at the earliest stages of the design process. Rather than directly linking these higher fidelity discipline analyses to the system level, synthesis/sizing code (FLOPS in Figure 3), the general characteristics of each discipline were represented by much simpler RSEs (typically no greater than second order) that define the variation of key discipline responses with respect to the design parameters of interest. This accomplished three important goals: (a) adequate representation of the physics of the problem under study, (b) representation of the results in a form simple enough for use in conceptual phase studies and (c) modularization of contributing analyses so that alternate tools could be easily substituted if desired. RSEs are certainly not new, and others have tried with varying degrees of success to apply them in similar situations. However, the success in the present application was derived in no small part from the development of powerful and statistically accountable ways to generate the needed RSEs and to regenerate them, when necessary, to extend the design variable
DSIDES
System Variables
System Goals RSE Sub-System Objectives Discipline FLOPS Wing Weight RSE
Deviation Variables
Structural Weight
Flight Condition Planform
APAS
ASTROS
Sub-Discipline
Maneuver Loads
SIC
AIC Aeroelastic Wing Design Procedure
ISMD
Figure 3. Hierarchical System Decomposition
3. FOCUS AREAS IN MDO SOLUTION
A drawback of using complex models (CFD, FEM, etc.) and complex tools during the research of new approaches to design is that the resources and time are often not available to synthesize more than a handful of different design points. The findings in this paper will show that a more desirable avenue for exploring new approaches, the project employed DREAMS to execute the integrated aero-structures-control methodology for the multidisciplinary design and optimization of an HSCT wing, as illustrated in Figure 3. A synopsis of the DREAMS method flow is: ? Use of the DSPT for a HSCT wing using a satisficing approach which includes identification of goals, deviation variables, and design variables ? Identification of required analysis/simulation tools ? Identification of information exchange between aero-structure-control modules ? Set up of the Design of Experiments (DOE) for design parameters and required system responses ? Development of RSEs from simulation results for use in preliminary DREAMS demonstration ? Implementation of aero-structure-controls integrated tool in IMAGE, with RSEs as agents ? Robust Design studies using RSEs integrated with the synthesis/sizing tool, FLOPS (Flight OPtimization System), for a point design solution

Design Space Definition
Case 1 2 3 . .
DOE Table
X1 -1 -1 -1 . . Y1 1 1 -1 . . X5 1 -1 1 . . Sref 1 1 -1 . . t/c 1 -1 1 . .
Linear Aero Linear Aero (APAS) (APAS)
Parametric Parametric Geometry Generator Geometry Generator (in house developed) (in house developed)
Paneled Geometry
Grid definition
Grid Generator/Translator Grid Generator/Translator (in house developed) (in house developed)
Grid/Panel Interpolation
FEM FEM (ASTROS) (ASTROS)
? Paneled Geometry ? AICs Maneuver Requirements
Trimmed Maneuver Loads Trimmed Maneuver Loads Generation Generation (ISMD) (ISMD)
? SICS ? Opt. Mass Distribution Trimmed Net Loads Iteration (until weight converges)
Case 1 2 3 . .
X1 -1 -1 -1 . .
Y1 1 1 -1 . .
X5 1 -1 1 . .
Sref 1 1 -1 . .
t/c 1 -1 1 . .
Wing Wt. WW(1) WW(2) WW(3) . .
Aeroelastic Wing Weight Equation W=b0 + b1X1+….+b11X12+….+b12X1· Y1+….
Figure 4. RSM/DOE Application to Aeroelastic Wing Design space. The key insight was to base the generation of the RSEs on a classical DOE approach to vastly simplify the combinatorial explosion that results from trying to use more than a dozen or so design variables. The initial Response Surface Methodology (RSM), a method for finding and validating RSEs, applications done by the investigators were towards the analysis of HSCT economics, but the research has extended the applications to aerodynamics, structural analysis and controls as necessary to address the wing aeroelasticity and transonic flutter design, including the potential application to active flexible wing (AFW) technologies. Figure 3 above summarizes the overall MDO methodology including system level considerations, while Figure 4 below details the more “discipline-focused” aeroelastic wing design. The HSCT aeroelastic wing design method used in conjunction with the DOE/RSM is described in detail in Ref. [12] and [14]. The problem addresses a finite-element based structural optimization of a wing box under aerodynamic loads that is subjected to stress and flutter constraints. The wing is represented by a varying complexity spar and rib model and utilizes multiple shape functions for distribution of design parameters. A maneuver load program, called Integrated Structure/Maneuver Design (ISMD), provides for the computation of static external loads. The key objective of the wing design procedure is to balance the desire for a parametric procedure and a desire for increased analysis accuracy. Taken alone, the RSM only addresses the modeling of higher fidelity analyses at the conceptual level. Even with the availability of this new approach, one is still left with substantial uncertainty in the fundamental design data itself. For example, costs such as fuel or manufacturing, as well as projections of new technology availability are subject to considerable uncertainty, so that simply using purely deterministic values in an RSM to model a higher fidelity analysis may not be justifiable at the conceptual phase. To address this issue, a Robust Design Synthesis (RDS) approach was developed and applied to the HSCT conceptual design problem. The basic approach involves the incorporation of stochastic models for the design parameters and the development of tools for their propagation through the RSEs. The initial effort focused on basic Monte Carlo approaches to this problem but subsequent development has pursued the use of semianalytical methods and so-called fast probability integration methods. The fundamental developments in robust design methodologies are described in more detail in the following section.

3.2 Decision Support Methods The basic HSCT wing design methodology outlined in the previous section addressed the challenge of moving higher fidelity design models and analysis tools forward into the conceptual and preliminary design phases of the design timeline and of handling the fundamental uncertainty inherent in the system parameters and design variables. This section summarizes the research effort to develop decision support methods to allow design decisions to be made in this environment with due account given to the level of fidelity as well as to the stochastic nature of the models and information. The DREAMS methodology is based on the DSPT, shown on the left of Figure 3 in which the design process is organized according to the types of decisions which are being made and the domain-dependent information which is available to make those decisions. Within the DSPT, two principal types of decisions are available to a designer selection decisions [56] and compromise decisions [39]. Each of these types of decisions is accomplished within the framework of a Decision Support Problem (DSP). The details of a compromise DSP in the context of the HSCT wing design problem under study are illustrated in Figure 5. The compromise DSP and the DSIDES software which implements it form the foundation for design exploration within this environment.
SATISFY GIVEN
Mission Profile
M = 2.4 ft. 60000 50000 ft. CRUISE 35000 ft. CLIMB DESC ENT M = 0.85 CRUISE LOI TER M = 0.6 25000 ft. ABOR T 3000 ft. TAXI & T.O. F.L. = 1100 0 ft. 200 n.m. 5000 n.m. 500 n.m. DESC ENT M = 0.6 25000 ft. RESERVE LAND F.L. = 1100 0 ft. S.D. & S.L. 100 n.m. 200 n.m.
RCEM was then used to develop ranged sets of specifications which are common and good for a family of general aviation aircraft [64]. In addition to focusing on the inclusion of robustness in design, the research addressed methods for increasing efficiency, increasing design knowledge and maintaining design freedom during the early stages of design for open engineering systems [65]. It also dealt with methods for modeling design uncertainty in design formulations and on methods for prioritizing design objectives [64,65]. Subsystems may not be isolated, self-supporting entities. Constraints, goals and design variables may be shared between entities. However, full communication and cooperation often does not exist. The information may be incomplete, or one subsystem may dominate the design. Game theory was applied to the DSP involving the design of two coupled subsystems, one of which dominates the design process [37,44,45]. A conceptual framework for the application of game theory in complex systems design was developed and applied initially to the multidisciplinary design of a subsonic passenger transport aircraft, as illustrated in Figure 6. This research also led to development of an algorithm for solving mathematical models involving nonlinear functions of both discrete and continuous design variables [41,46].
FIND
Aero Constraints Aero Goals
missed approaches landing field length take-off field length aspect ratio Continuous Wing Area Discrete Fuselage Length Wing Span
Aerodynamic Configuration
HSCT Overall Design Requirements Flutter and Maneuver Constraints Control Surface Settings
GIVEN
Mission Profile
aspect ratio climb gradients landing field length take-off field length drag coefficients
SATISFY
Xaero , Saero
Compromise DSP Given
Planform Shape/Size
Xwing 0, 0
Wing Design Concepts control surface layouts Generic HSCT Baseline
Find Satisfy Minimize
FIND
(X1, Y1) naY2 naY1
Aerodynamic constants Weight constants Engine choices
Xweight , Sweight
System Configuration
SATISFY
Weight Goals Weight Constraints
useful load fuel available climb gradient landing field length take-off field length
FIND
Engine Requirements and Fuel Balance
(X2,1) Y-axis (X3,1) (X4, Y1) X5, 0
productivity index useful load fraction fuel balance missed approaches landing field length take-off field length
MINIMIZE
Deviation for achievable and desired goals for: 1) Ticket Price 2) Gross Weight 3) Flyover Noise
Continuous Take-off Weight Discrete Installed Thrust
Modeled as a 2 Player Game Modeled as a 2 Player Game
Figure 6. Alternative Decision Strategy: Game Theory Complex systems usually involve an extensive hierarchical structure and this can be used to advantage in the decision support problem formulation. A hierarchical robust preliminary design exploration method was developed to facilitate concurrent system and subsystem design exploitation, and specifically for the concurrent generation of robust system and subsystem specifications for the preliminary design of multilevel, multi-objective, largescale complex systems [34]. This method was developed through the integration and expansion of current design techniques including:
Figure 5. Compromise Decision Support Problem for HSCT Wing Design Problem In the initial DSP application, the compromise DSP was used as the foundation for the development of a Robust Concept Exploration Method (RCEM) to facilitate the quick evaluation of design alternatives and the generation of toplevel specifications with quality considerations in the early stages of design of a complex system [1,2]. The RCEM was implemented by integrating several metrics and tools including Taguchi’s robust design [1], Suh’s independence axiom [37], and DOE/RSM [11] into one mathematical construct - the Compromise Decision Support Problem. The RCEM was demonstrated in the context of the design of an integrated HSCT airframe/propulsion subsystem [2,8]. The

TIME AND SCHEDULE REQUIREMENTS REQUIREMENTS AND SPECIFICATIONS
AVAILABLE TEST DATA
REQUEST FOR PROPOSAL (RFP) EXISTING AIRCRAFT DESIGNS PREVIOUS RESEARCH
A0
DESIGN HSCT AIRCRAFT
HSCT DESIGN
COMPUTER RESOURCES DESIGN EXPERIENCE
WIND TUNNEL
SCIENTISTS AND ENGINEERS WATER TUNNEL
(a) IDEF0 - Level 0
REQUIREMENTS AND SPECIFICATIONS TIME AND SCHEDULE REQUEST FOR PROPOSAL (RFP) EXISTING DESIGNS PREVIOUS RESEARCH REQUIREMENTS AND TIME AND SCHEDULE SPECIFICATIONS AVAILABLE DATA REJECTED DESIGNS A2 PERFORM PROPOSED CANDIDATE 3-VIEW DESIGNS DESIGN CAD/CAM EXPERIENCE AIR SHOWS 1 ORDER SIZING PROGRAMS AIRCRAFT DESIGNERS/ CONFIGURATORS
st
A1 DEVELOP CONCEPTUAL 3-VIEW BASELINE DESIGNS
INITIAL ANALYSIS OF PROPOSED 3-VIEW DESIGNS
REQUIREMENTS AND SPECIFICATIONS TIME AND SCHEDULE
EXISTING AIRCRAFT
DETAILED ANALYSIS AND AIRCRAFT SIZING DESIGN PEFORMANCE ENGINEERS PROGRAMS CODES
SUITABLE DESIGNS FOR SURFACING
A3 DEVELOP SURFACED MODELS OF CANDIDATE DESIGNS
SUITABLE DESIGNS FOR REFINEMENT
DESIGN EXPERIENCE
CAD/CAM
AIRCRAFT DESIGNERS/ CONFIGURATORS
CANDIDATE CONFIGURATION GEOMETRIES REQUIREMENTS AND TIME AND SCHEDULE SPECIFICATIONS AVAILABLE DATA REQUIREMENTS AND SPECIFICATIONS TIME AND SCHEDULE
DEVELOP REFERENCE CONFIGURATION DATABASE
A5
A6 DEMONSTRATE PERFORMANCE GUARANTEES CONCEPTUAL HSCT DESIGN DEVELOP PRELIMINARY HSCT DESIGN PRELIMINARY HSCT DESIGN
AIRCRAFT WIND ANALYSIS TECHNOLOGY TUNNEL CODES ENGINEERS WATER AIRCRAFT TUNNEL DESIGNERS/ CONFIGURATORS
OPTIMIZED REFERENCE DESIGN
PERFORMANCE AIRCRAFT CODES TECHNOLOGY ENGINEERS AIRCRAFT DESIGNERS/ CONFIGURATORS
(b) IDEF0 - Level 1 Figure 8. IDEF0 Diagrams Hierarchical partitioning and modeling techniques for partitioning large-scale complex systems into more tractable parts and allowing integration of sub-problems for system synthesis [25, 34] ? Statistical experimentation and approximation techniques for increasing both the efficiency and the comprehensiveness of preliminary design exploration [35] ? Noise modeling techniques for implementing robust preliminary design when approximate models are employed [6,7] These techniques were developed for a case study performed with Rolls-Royce Allison and are based on an existing engine designed for midsize commercial, regional business jets. The solutions obtained are similar to the design model ? of the existing commercial engine, but are better with respect to many of the requirements [34]. Additional supporting research provided a method for coupling objectives related to technical and economic efficiency in an environment involving the simultaneous consideration of multiple objectives, multilevel decisions and uncertainty [48]. This work further compared a single objective model with methods founded on the Compromise Decision Support Problem [56, 77, 78]. 3.3 Information Management in MDO During the course of implementing this design scenario of DREAMS in IMAGE, the need to have a well defined data model became evident. In addition, research has shown that advances in the aircraft technologies have resulted in a commensurate increase in the amount of data required to

define a design during the conceptual stages [28]. Conceptual design requires a tight multidisciplinary effort requiring large amounts of data exchange. In order to effectively implement the new MDO design processes described above, it is crucial that a top-down data management design structure be in place in the early phases of the design. This structure must provide consistency in data format and allow ease of data exchange between the various disciplines involved in the design process. In the conceptual design phase, consideration must be given to the changing structure of the database as the product design evolves. Current database design approaches are typically limited to the detailed design phase where the data organization is fixed and unambiguous. The data modeling problem is encountered for both design product and process models. The use of IDEF0 diagrams for representing a design (Figure 8) was employed in the research, and the use of these diagrams was extended to the use of Design Palettes as illustrated for the aeroelastic wing design problem [28]. A data model is also required for the product information. A IDEF1X model for a generalized aircraft component has also been developed. In this example, an aircraft configuration is made up of the components: engine, fuselage, landing gear, inlet, nozzle, canard, horizontal tail, vertical tail, and wing. This data model is instantiated within the IMAGE architecture. An expansion of this data model using EXPRESS, a modeling language, to provide an object oriented perspective was also carried out in a separate environment [31] and showed some of the assets and liabilities of such models. It was concluded that a relational model was often sufficient for the early design area but that an object model had more flexibility for evolving with the design. The data modeling approach was also investigated [9,10] as a way of identifying and tracking the impact of local design changes on overall system design. The data
Design Activities
model provides a system level view and appears useful in tracking the propagation of change throughout a system. 3.4 MDO Computational Frameworks The new MDO methodologies developed in the research were implemented in an MDO framework and integration project. The methodology was initially identified as IDES, but it ultimately was identified by the name DREAMS. Much of the computational architecture development was pursued under GSRP funding with the project name IMAGE. However, the guidance for much of this was based on the requirements established by DREAMS to accommodate the new MDO approaches being developed under the grant. Additional specifications were derived from related projects such as FIDO (at NASA LaRC) and ASOP (at Rockwell). The computational and information environment was designed and constructed to provide a framework for consistently applying a general decision-based design methodology within an integrated computing environment across the design timeline for open engineering systems. It is based around an agent integration technology, and results have demonstrated the feasibility in situations of practical complexity level [17-20]. Side research was done on methods for coarse and fine grained parallelization to enhance the computational efficiency as documented in Reference [71]. A distributed, object-oriented database definition with dynamic schema editing was also demonstrated.
Available Assets
Database
Visualization
Optimization
Geometry Algorithmic Heuristic Simulation Simulation
Agent Collaboration
Model Wrap
Model Wrap
Model Wrap
Model Wrap
Model Wrap
Model Wrap
Model Wrap
Model Wrap
Model Wrap
Model Wrap
Tasking
Parallelism
Protocols
Scheduling
Network
Computing Architecture
Figure 9. IMAGE Infrastructure Which Supports DREAMS

The infrastructure was designed to support the DREAMS methodology by incorporating: ? A design partitioning process ? A mechanism for solving the resulting design subproblems ? A design information model In addition, the infrastructure supports: ? Information generation in context for informed decision-making ? Efficient and cost-effective application of design resources ? Geographically distributed design activities The basic approach in a schematic form is shown in Figure 9, and it refers to the following specific functional capabilities incorporated in the system: ? Design Activities in which a designer partitions a problem into activities for solution; this also provides for comprehensive information management ? Available Assets which include a variety of design resources (e.g., programs) that provide aid in the generation of design knowledge; resources may include performance simulation codes, objectoriented databases, CAD packages, etc. ? Agent Collaboration as implemented with a generic toolkit that allows resources to be incorporated into the design infrastructure with minimal effort by the engineering developers; the incorporation of a “model” (which describes precisely what an agent is capable of doing or providing and how it is accomplished) within the toolkit allows for knowledge to be generated in context allowing a designer to interrogate knowledge for the who, what, where, when, and how the information was created. ? Computing Architecture which includes components that are required for objects to operate in a distributed, homogeneous computing environment are included in an underlying infrastructure. The computational framework is considered open because it provides freedom for a designer to model both processes and information as required at a particular point in a design’s timeline. This was accomplished through an information model which incorporated schema evolution to capture time-dependent product and process characteristics at varying degrees of accuracy and fidelity. This model evolved from the IDEF models described earlier. As a result, product descriptions can be modified as fidelity increases. For example, in the case of the HSCT wing design application considered in this project, an initial product description was based on parametric components. During finite element analysis, a more detailed model was required that included node and member definitions. Both of these representations coexisted in the information model. Moreover, specific instances (e.g. values) were accumulated for decision-making and optimization.
The computational framework was demonstrated in several different HSCT wing design studies during the third and fourth years of the grant. Based on evaluations of the framework performance as well as from user feedback, a number of refinements and improvements were incorporated. Much of this focused on improving usability of the integration tools and on enhancements to the user interface. Continuing development is moving the framework from its heavy dependence on the Unix environment to a more open architecture incorporating emerging web technologies and a “lean server” technology. When used in a client-server approach, a lean server using the hypertext transport protocol (http) provides a much more simplified server-side interface to analysis codes and decision support tools. Ultimately, the architecture will migrate towards a web-based and largely platform-independent configuration with most of the user interface implemented using the hypertext markup language (html). A current evaluation version of the IMAGE framework is available for download from: https://www.wendangku.net/doc/d412685186.html,/image/.
4. CLOSURE
Research accomplishments under grant support are extensive and wide-ranging as documented in the grant bibliography listed here in the references. The findings include: ? Extension of traditional MDO methods to conceptual/preliminary design ? Ability to do multi-level, high fidelity system studies ? Wing aeroelastic analysis which permits design trades to be done on a series of computers ? Development of HSCT aeroelastic wing weight equation ? Integration of decision support in design process ? Development of IDEF HSCT data models Consistent with the multidisciplinary theme of the grant, the research team included faculty and students in both Aerospace Engineering and Mechanical Engineering at Georgia Tech along with industrial participation from Rockwell International and Lockheed Martin (see below). The objective as stated in the initial proposal and refined in subsequent work statements for each of the 3 years was to develop a new MDO approach to aerospace systems design and to apply it to the HSCT design problem. The research plan was outlined in Section 2 above in four primary areas and substantial achievements have been made in all the areas. A new approach to MDO focusing on the conceptual (advanced or pre-) design and early preliminary design phases was developed under the first two tasks described above. The approach involved a combination of consistent refinement in the fidelity of the analysis models with the incorporation of decision support problem (DSP) techniques to synthesize form to meet functional requirements. Both deterministic and stochastic approaches were developed.

The remaining two tasks focused on the implementation of the new MDO methodologies and their evaluation when applied to the HSCT design problem outlined in the initial proposal. One area of research considered the development of parallel computing strategies to support MDO. The second addressed the development of computational frameworks to support MDO across the enterprise. New approaches to MDO have been developed and demonstrated during this project on a particularly challenging aeronautics problem – HSCT Aeroelastic Wing Design. This problem represents a grand challenge to the aerospace community because of the complex interactions that occur in its solution. A new framework for systematic problem solution, called DREAMS, was developed. This framework included aspects of decision-making (DSPT, RSM, RCEM), information technology (IDEF, EXPRESS), simulation capability (Parallelization), and scalable software design (IMAGE). As this framework was developed, a parallel effort was placed on its application to detailed aeroelastic wing design. At this level of complexity, many intricacies were identified and addressed. Here it was found that an application was needed in order to verify, as well as drive requirement formulation, for the methodology. The resulting framework is available for application to other complex multidisciplinary challenges.
8.
9.
10.
11.
12.
13.
6. REFERENCES
The following publications were produced in whole or in part with funding from the NASA Grant. 1. Chen, W., “A Robust Concept Exploration Method for Configuring Complex Systems,” Ph.D. Dissertation, School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA., 1995. Chen, W., Allen, J.K., Mavris, D., Mistree, F, "A Concept Exploration Method for Determining Robust Top-Level Specifications," Engineering Optimization, vol. 26, pp. 137-158, 1996. Chen, W., Simpson, T.W., Allen, J.K, and Mistree, F., “Satisfying Ranged Sets of Design Requirements using Design Capability Indices as Metrics,” Engineering Optimization, vol. 31, no.4., pp. 615-639, 1999. Chen, W., Allen, J.K, Tsui, K-L and Mistree, F., "A Procedure for Robust Design: Minimizing Variations Caused by Noise Factors and Control Factors," ASME Journal of Mechanical Design, vol. 118, no. 4, pp. 478485, 1996. Chen, W., Allen, J.K. and Mistree, F., “System Configuration: Concurrent Subsystem Embodiment and System Synthesis,” ASME Journal of Mechanical Design, vol. 118, no. 2, pp. 165-170, 1996. Chen, W., Allen, J.K, Tsui, K-L and Mistree, F., "A Procedure for Robust Design: Minimizing Variations Caused by Noise Factors and Control Factors," ASME Journal of Mechanical Design,vol. 118, no. 4, pp.478485, 1996. Chen, W., Allen, J.K., Schrage, D.P. and Mistree, F., 14.
2.
15.
3.
16.
4.
17.
5.
18.
6.
19.
“Statistical Experimentation For Affordable Concurrent Design,” AIAA Journal, vol. 35, no. 5, pp. 893-900, 1997. Chen, W., Allen, J. K., and Mistree, F., "The Robust Concept Exploration Method for Enhancing Concurrent Systems Design", Concurrent Engineering: Research and Applications, vol. 5, no. 3, pp. 203-217, 1997. Cohen, T., “A Data Approach to Tracking and Evaluating Engineering Changes”, Ph.D. Thesis, Georgia Institute of Technology, 1997. Cohen, T., Fulton, R.E. (1998) “A Data Approach to Tracking and Evaluating Engineering Changes”, 1998 ASME Design Engineering Technical Conferences, Proceedings of DETC’98, Atlanta. DeLaurentis, D., “A Probabilistic Approach to Aircraft Design Emphasizing Stability and Control Uncertainties,” Ph.D. Dissertation, School of Aerospace Engineering, Georgia Institute of Technology, November 1998. DeLaurentis, D., Cesnik, C., Lee, J., Mavris, D., Schrage, D., "A New Approach to Integrated Wing Design in Conceptual Synthesis and Optimization," 6th AIAA/USAF/NASA/ISSMO Symposium on Multidisciplinary Analysis and Optimization, Bellevue, WA, September 4-6, 1996. AIAA-96-4000. DeLaurentis, D.A., Mavris, D.N., “An IPPD Approach to the Preliminary Design Optimization of an HSCT using Design of Experiments”, 20th ICAS Congress, Sorrento, Italy, September 1996. DeLaurentis, D.A., Zink, P.S., Mavris, D.N., Cesnik, C.E.S., Schrage, D.P., "New Approaches to Multidisciplinary Synthesis: An Aero-StructuresControl Application Using Stastical Techniques," 1st AIAA/SAE World Aviation Congress, Los Angeles, CA, October 21-24, 1996. AIAA-96-5501. Hale, M.A., "A Computing Infrastructure that Facilitates Integrated Product and Process Development from a Decision-Based Perspective," Ph.D. Thesis Proposal, School of Aerospace Engineering, Georgia Institute of Technology, Atlanta, GA, January, 1995. Hale, M.A., "An Open Computing Infrastructure that Facilitates Integrated Product and Process Development from a Decision-Based Perspective," Doctoral Thesis, School of Aerospace Engineering, Georgia Institute of Technology, Atlanta, GA. July 1996. Hale, M. A. and Craig, J. I., "Preliminary Development of Agent Technologies for a Design Integration Framework," 5th AIAA/NASA/USAF/ISSMO Symposium on Multidisciplinary Analysis and Optimization - Panama City, September 7-9, 1994. AIAA-94-4297. Hale, M. A. and Craig, J. I., "Use of Agents to Implement an Integrated Computing Environment," Computing in Aerospace 10, AIAA, San Antonio, TX, March 28-30, 1995. AIAA-95-1001. Hale, M. A. and J. I. Craig, "Techniques for Integrating Computer Programs into Design Architectures," Sixth
7.

20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
AIAA / NASA / USAF / ISSMO Symposium on Multidisciplinary Analysis and Optimization, Bellevue, WA, September 4-6, 1996. AIAA-96-4166. Hale, M. A., "Dynamic CATIA CATGEO Access: Powerful Interpretive Technologies Based on Tk/tcl," presented at the CATIA Operators Exchange Meeting, Long Beach, CA, October 8-12, 1995. Hale, M. A., "IMAGE: A Design Integration Framework Applied to the High Speed Civil Transport," HM301: First University/Industry Symposium on High Speed Civil Transport Vehicles, North Carolina State University, December 4-6,1994. Hale, M. A., "Preliminary Agent Technologies with CATIA," presented at the CATIA Operators Exchange Meeting, Dallas, October 9-13, 1994. Hale, M. A., Craig, J. I., Mistree, F. and Schrage, D.P., "On the Development of a Computing Infrastructure that Facilitates IPPD from a Decision-Based Design Perspective," 1st AIAA Aircraft Engineering, Technology, and Operations Congress, Anaheim, CA, September 19-21, 1995. AIAA-95-3880. Hale, M. A., Craig, J. I., Mistree, F., Schrage, D. P., "Implementing an IPPD Environment from a DecisionBased Design Perspective," ICASE/LaRC Workshop on Multidisciplinary Design Optimization, Hampton, VA, March 13-16, 1995. Hale, M. A., J. I. Craig, F. Mistree and D. P. Schrage, "DREAMS & IMAGE: A Model and Computer Implementation for Concurrent, Life-Cycle Design of Complex Systems," Concurrent Engineering: Research and Applications, vol. 4, no. 2, pp. 171-186, June 1996. Hall, N., and Fulton, R.E., "A Relational Database Application to Multidisciplinary Conceptual Design for HSCT," (Submitted for the publication). Hall, N., and R. Fulton, "A Relational Database Approach to a Multidisciplinary Conceptual Design for the HSCT," Georgia Institute of Technology, September, 1994. Hall, N. S. and Fulton, R. E., “An Investigation of a Relational Database Approach to a Multidisciplinary Conceptual Design for the HSCT”, 1996 ASME Design Engineering Technical Conferences and Computers in Engineering Conference, Irvine, California, August 1822, 1996, Paper Number 96-DETC/EIM-1425. Hall, N. S. and Fulton, R. E., “Impact of Data Modeling and Database Implementation Methods on the Optimization of Conceptual Aircraft Design”, Research Paper, School of Mechanical Engineering, Georgia Institute of Technology, 1996. Hall, N.S.; Fulton, R.E. (1997), “Impact of Data Modeling and Database Implementation Methods on the Optimization of Conceptual Aircraft Design”, ASME Engineering Information Management Symposium (EIMS’97) at DETC’97, Sacramento. Hall, N.S., “Impact of Data Modeling and Database Implementation Methods on the Optimization of
32.
33. 34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
Conceptual Aircraft Design”, M.S. Thesis, Georgia Institute of Technology, 1996. Har, J. and Fulton, R. E., "A Technical Note: Parallel Finite Element Implementation for Wing Models," Parallel Computing Research Lab, School of Mechanical Engineering, Georgia Institute of Technology, May 5, 1995. Har, J. (1998) “A New Scalable Parallel Finite Element Approach for Contact-Impact Problems”. Koch, P., "Hierarchical Modeling and Robust Synthesis for the Preliminary Design of Large Scale Complex Systems," Ph.D. Dissertation, School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA., 1997. Koch, P.N., Allen, J.K. and Mistree, F., “Problem of Size in Robust Design,” Advances in Design Automation, (Parkinson, A. Ed.), New York: ASME 1997. ASME97-DETC/CAD-3983. Koch, P.N., Simpson, T.W., Allen, J.K., and Mistree, F., "Statistical Approximations for Multidisciplinary Optimization: The Problem of Size," Journal of Aircraft, Special MDO Issue, vol. 36, no.1, pp. 275286, 1999. Lewis, K, "An Algorithm for Concurrent Subsystem Embodiment and System Synthesis," Ph.D. Dissertation, School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA., 1996. Lewis, K. and Mistree, F, "On Developing a Taxonomy for Multidisciplinary Design Optimization: A DecisionBased Perspective," First World Congress of Structural and Multidisciplinary Optimization, Goslar, Germany. Paper number 118., 1995. Lewis, K., Smith W.F. and Mistree, F., “Ranged Set of Specifications for Complex Engineering Systems” in Simultaneous Engineering: Methodologies and Applications, pages 279-304, (U. Roy, J.M. Usher and H.R. Parsaei, Eds.), New York: Chapman-Hall, (1999). 40. Lewis, K. and Mistree, F., “Collaborative, Sequential and Isolated Decisions in Design,” ASME Design Theory and Methodology ’97, (Shah, J. Ed.), New York: ASME 1997. ASME97-DETC97/DTM3883. Lewis, K. and Mistree, F., “Foraging-Directed Adaptive Linear Programming: An Algorithm for Solving Nonlinear Mixed Discrete/Continuous Design Problems,” Advances in Design Automation, (Dutta, D. Ed.), New York: ASME 1996. ASME96-DETC/CAD1601. Lewis, K., Lucas, T. and Mistree, F., "A DecisionBased Approach for Developing A Ranged Top-Level Aircraft Specification: A Conceptual Exposition," 5th AIAA/NASA/ USAF/ISSMO Symposium on Multidisciplinary Analysis and Optimization, Panama City, FL, September 7-9, 1994. AIAA-94-4304. Lewis, K. and Mistree, F., "The Other Side of Multidisciplinary Design Optimization: Accommodating a Multiobjective, Uncertain and Non-

44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
Deterministic World," Engineering Optimization, vol. 31, no. 2, 1998. Lewis, K. and Mistree, F., “Modeling Interactions in Multidisciplinary Design: A Game Theoretic Approach,” AIAA Journal, vol. 35, no. 8, pp. 13871392, 1997. Lewis, K. and Mistree, F., "Collaborative, Sequential and Isolated Decisions in Design," ASME Journal of Mechanical Design, vol. 120, no.4, pp. 643-652, 1998. Lewis, K. and Mistree, F., "FALP: Foraging-Directed Adaptive Linear Programming - A Hybrid Algorithm for Discrete/Continuous Problems," Engineering Optimization, (in press). Lucas, T., “Formulation and Solution of Hierarchical Decision Support Problems,” M.S. Thesis, School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA, 1995. Lucas, T., Vadde, S., Chen, W., Allen, J.K. and Mistree, F., "Utilization of Fuzzy Compromise DSPs for Hierarchical Design Problems," AIAA/ASME/ASCE/AHS/ACS 35th Structures, Structural Dynamics and Materials Conference, Hilton Head, South Carolina, April 18-20, 1994. AIAA-941543. Marx, W. J., Mavris, D. N., and Schrage, D. P., "A Hierarchical Aircraft Life Cycle Cost Analysis Model," 1st AIAA Aircraft Engineering, Technology, and Operations Congress, Los Angeles, CA, September 1921, 1995. AIAA-95-3861. Marx, W. J., Mavris, D. N., and Schrage, D. P., "Integrated Design and Manufacturing for the High Speed Civil Transport," 19th ICAS Congress / AIAA Aircraft Systems Conference, Anaheim, CA, September 18-23, 1994. ICAS-94-10.8.4. Marx, W. J., Mavris, D. N., and Schrage, D. P., "Integrated Product Development for the Wing Structural Design of the High Speed Civil Transport," 5th AIAA/NASA/USAF/ISSMO Symposium on Multidisciplinary Analysis and Optimization, Panama City, FL, September 7-9, 1994. AIAA-94-4253. Marx, W. J., Mavris, D. N., and Schrage, D. P., "Knowledge-based Manufacturing and Structural Design for a High Speed Civil Transport," AS205: 1st Industry/Academy Symposium on Research for Future Supersonic and Hypersonic Vehicles, Greensboro, NC, December 4-6, 1994. Marx, W. J., Schrage, D. P., and Mavris, D. N., "An Application of Artificial Intelligence for ComputerAided Design and Manufacturing," International Conference on Computational Engineering Science; Supercomputing in Multidisciplinary Analysis and Design, Mauna Lani, HI, July 30 - August 3, 1995. ICES-95-B6-3. Mavris, D.N., Bandte, O., and Schrage, D.P., "Economic Uncertainty Assessment of an HSCT Using a Combined Design of Experiments/Monte Carlo Simulation Approach", 17th Annual Conference of the
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
International Society of Parametric Analysts, San Diego, CA, May 1995. Miller, G.D., “An Active Flexible Wing Multidisciplinary Design Optimization Method”, AIAA/USAF/NASA/ISSMO Symposium on Multidisciplinary Analysis and Optimization, Panama City, FL, 7-9 September 1994. Mistree, F., Lewis, K. and Stonis, L., "Selection in the Conceptual Design of Aircraft," 5th AIAA/NASA/USAF/ISSMO Symposium on Multidisciplinary Analysis and Optimization, Panama City, FL, September 7-9, 1994. AIAA-94-4382. Mistree, F., Patel, B. and Vadde, S., "On Modeling Multiple Objectives and Multi-Level Decisions in Concurrent Design," Advances in Design Automation, pp. 151-161, (Gilmore, B.J., Hoeltzel, D., Dutta, D. and Eschenauer, H., Eds.), New York: ASME, (1994). ASME DE-Vol. 69-2. Nottingham, C.R., Ortega, R.A., Rangarajan, B., Koch, P.N., and Mistree, F., “Designing for Maintainability in the Preliminary Stages of Aircraft Engine Design,” Advances in Design Automation, (Parkinson, A. Ed.) New York: ASME 1997. ASME97-DETC97/DAC3963. R?hl, P., Mavris, D.N., and Schrage, D.P., "A Multilevel Wing Design Procedure Centered on the ASTROS Structural Optimization System," 5th AIAA/NASA/USAF/ISSMO Symposium on Multidisciplinary Analysis and Optimization - Panama City, September 7-9, 1994. AIAA-94-4411. R?hl, P.J., , "A Multilevel Decomposition Procedure for the Preliminary Wing Design of a High Speed Civil Transport Aircraft," Ph.D. Thesis, School of Aerospace Engineering, Georgia Institute of Technology, June 1995. R?hl, P.J., Mavris, D.N., and Schrage, D.P., "A Multilevel Decomposition Procedure for the Preliminary Wing Design of High-Speed Civil Transport Aircraft," First Industry/Academy Symposium on Research for Future Supersonic and Hypersonic Vehicles, Greensboro, NC, December 1994. R?hl, P.J., Mavris, D.N., and Schrage, D.P., "Preliminary HSCT Wing Design Through Multilevel Decomposition," 1st AIAA Aircraft Engineering, Technology, and Operations Congress, Los Angeles, CA, September 19-21, 1995, AIAA 95-3944. R?hl, P.J., Schrage, D.P. and Mavris, D.N., "Combined Aerodynamic and Structural Optimization of a HighSpeed Civil Transport Wing," 36th AIAA Structures, Dynamics, and Materials Conference, New Orleans, LA, April 1995, AIAA 95-1222. Simpson, T.W., Chen, W., Allen, J.K. and Mistree, F., "Conceptual Design of a Family of Products Through the Use of the Robust Concept Exploration Method," AIAA/NASA/USAF/ISSMO Symposium on Multidisciplinary Analysis and Optimization, Bellevue, Washington, September 4-6, 1996, pp. 1535-1545.

AIAA Paper Number 96-4161. 65. Simpson, T.W., “Development of a Design Process for Realizing Open Engineering Systems,” M.S. Thesis, School of Mechanical Engineering, Georgia Institute of Technology, August 1995. 66. Mistree, F., Smith, W. F., Bras, B., Allen, J. K. and Muster, D., "Decision-Based Design: A Contemporary Paradigm for Ship Design", Transactions, Society of Naval Architects and Marine Engineers, vol. 98, pp. 565-597, 1990. 67. Synn, S.Y. and Fulton, R.E., "Practical Strategy for Concurrent Substructure Analysis," Journal of Computers Structures, vol.54, no.5, 1995. 68. Synn, S.Y. and Fulton, R.E., "The Concurrent Element Level Processing for Nonlinear Dynamic Analysis on a Massively Parallel Computer," Third National Symposium on Large-Scale Structural Analysis for High-Performance Computers and Workstations, Norfolk, VA, November 8-11, 1994. 69. Synn, S.Y. and Fulton, R.E., "The Prediction of Parallel Skyline Solver and It's Implementation For Large Scale Structural Analysis," Third National Symposium on Large-Scale Structural Analysis for High-Performance Computers and Workstations, Norfolk, VA, November 8-11, 1994 (also in Journal of Computer Systems in Engineering). 70. Synn, S.Y. and Fulton, R.E., "The Prediction of Parallel Skyline Solver and its Implementation for Large Scale Structural Analysis," Third National Symposium on Large-Scale Structural Analysis for High-Performance Computers and Workstations, Norfolk, VA, November 8-11, 1994, (Also, in Journal of Computer Systems in Engineering). 71. Synn, S.Y., "Practical Domain Decomposition Approaches for Parallel Finite Element Analysis", Ph.D. Thesis, School of Mechanical Engineering, Georgia Institute of Technology, January 1995. 72. Synn, S.Y., K. Schwan and Fulton, R.E., "Analysis of Large Scale Heterogeneous Structures on Massively Parallel Computers," Journal of Concurrency, Practice/Experience, 1995. 73. Synn, S.Y.; Fulton, R.E. (1995), “Performance Prediction of a Parallel Skyline Solver and its Implementation for Large Scale Structure Analysis”, Computing Systems in Engineering: An International Journal, Vol., 6, No. 3, Pergamon Press Inc., Tarrytown, NY, USA, 275-284. 74. Synn, S.Y.; Fulton, R.E. (1995), “Concurrent Element Level Processing for Nonlinear Dynamic Analysis on a Massively Parallel Computer,” Computing Systems in Engineering: An International Journal, Vol. 6, No. 3, Pergamon Press Inc., Tarrytown, NY, USA, 285-293. 75. Synn, S.Y., “Practical Domain Decomposition Approaches for Parellel Finite Element Analysis”, Ph.D. Thesis, Georgia Institute of Technology, 1995. 76. Synn, Sang Y. and Fulton, Robert E., “Prediction of Parallel Computing Performance”, Research Paper,
Parallel Processing Lab, School of Mechanical Engineering, Georgia Institute of Technology, 1996. 77. Vadde, S., “Modeling Multiple Objectives and Multilevel Decisions in Concurrent Design of Engineering Systems,” M.S. Thesis, School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA, 1995. 78. Vadde, S., Allen, J.K., Lucas, T. and Mistree, F., "On Modeling Design Evolution Along a Design TimeLine," 5th AIAA/NASA/USAF/ISSMO Symposium on Multidisciplinary Analysis and Optimization - Panama City, September 7-9, 1994. AIAA-94-4313. 79. Vadde, S. and Mistree, F., "Design of a Shaft-Disk System: Modeling Interactions Between Design and Manufacture," AIAA/NASA/USAF/ISSMO Symposium on Multidisciplinary Analysis and Optimization, Bellevue, Washington, September 4-6, 1996, pp. 1546-1557. AIAA Paper Number 96-4162.
7. ACKNOWLEDGEMENTS
The authors wish to acknowledge support from the NASA Langley Multidisciplinary Optimization Brach for this work under grant NGT 51102L. Advisors included Dr. David Rudy, Dr. Thomas Zang, and Dr. Robert Weston. Research results from students funded by NASA Multidisciplinary Analysis (MDA) and Graduate Student Research Program (GSRP) is acknowledged. Grant industrial partners included Rockwell International – North American Aircraft (now Boeing), Seal Beach, CA and Lockheed Aeronautical Systems Company (now Lockheed Martin), Marietta, GA. Their assistance throughout this effort is appreciated. All research done by the authors was conducted while in studies at Georgia Tech funded under the projects given above. Current affiliations are listed in the introduction.
8. AUTHORS
P. Scott Zink is a Ph.D. Student in the Georgia Tech Aerospace Systems Design Laboratory. He is researching integrated aeroelastic optimization techniques for active aeroelastic wing technology and developing a method by which structural designs can be acquired that are robust to maneuver load uncertainty. He received a Bachelor of Aerospace Engineering and M.S. in Aerospace Engineering from Georgia Tech in 1995 and 1997, respectively. Mr. Zink can be reached at 404/894-3343 and scott.zink@https://www.wendangku.net/doc/d412685186.html,.

Daniel A. DeLaurentis, Ph.D., is a Research Engineer II in the School of Aerospace Engineering at Georgia Tech and co-leads the advanced design methodology thrust area for the ASDL and Center for Aerospace Systems Analysis (CASA), which includes elements of system engineering and system analysis evaluation. He is the author of several papers on the topics of uncertainty modeling for complex systems, robust design methods, and multidisciplinary design (MDO). Dr. DeLaurentis’ education includes a B.S. in Aerospace Engineering at the Florida Institute of Technology and a M.S. and Ph.D. in Aerospace Engineering at the Georgia Institute of Technology. Dr. DeLaurentis can be reached at 404/8948280 and dan.delaurentis@https://www.wendangku.net/doc/d412685186.html,. Mark A. Hale, Ph.D., conducts research on computational enterprise management for integrated product and process development from a decisionbased perspective at the Georgia Tech Center for Aerospace Systems Analysis. He developed a computer architecture name IMAGE (Intelligent Multidisciplinary Aircraft Generation Environment) for this research. Dr. Hale has over 15 refereed publications and conference publications in the field. He is a member of IEEE, AIAA, SAE, and ASME. Dr. Hale can be reached at 404/894-9810 and mark.hale@https://www.wendangku.net/doc/d412685186.html,. Vitali V. Volovoi, Ph.D. is a Post Doctoral Fellow in the School of Aerospace Engineering at the Georgia Institute of Technology. Dr. Volovoi conducts research in structural mechanics with emphasis on the analysis and design of advanced composite structures including elastic tailoring of composites, integration of high fidelity structural tools into the general framework for conceptual and preliminary design of modern aircraft, structural optimization techniques, failure characterization of composites; finite element methods for structural and multi-body dynamics, rotorcraft and fixed wing aeroelasticity. Dr. Volovoi has over 10 publications in refereed journals and conference proceedings. His education includes a University Diploma Cum Laude in Applied Mathematics from the Moscow State University and a Ph.D. in Aerospace Engineering from the Georgia Institute of Technology.
Daniel P. Schrage, Ph.D., is director of the Center for Excellence in Rotorcraft Technology, and a Professor at the School of Aerospace Engineering at the Georgia Institute of Technology. His education includes a B.S., General Engr., USMA,. West Point, NY; M.S., Aerospace Engr., Georgia Tech; D.Sc., Mechanical Engineering, Washington U., St. Louis, MO; and M.A., Bus. Admin., Webster U. Government experience includes engineering and management experience at AVSCOM as Director for Advanced Systems; Chief, Structures & Aeromechanics Div. and Aeromechanics Branch Chief. Former member of the Army Science Board, current member of the Air Force Studies Board and NASA's ARTS Committee. Military experience: Reserves: COL, MOBDES Faculty USMA Dept. Mech. & Civil Eng.; Active Duty: 9 years an an Army Aviator with combat experience and 2 years as a Field Artillery Officer. Publications include over 50 technical papers/reports with 18 refereed and contributed chapters to three books. James I. Craig, Ph.D., is codirector of the Georgia Tech Center for Aerospace Systems Analysis. His current research interests and sponsored projects are focused on problems involving complex engineering systems design, structural dynamics and active structural control. Dr. Craig may be reached at 404/8943042 and james.craig@https://www.wendangku.net/doc/d412685186.html,. Farrokh Mistree’s, Ph.D., design experience spans mechanical, aeronautical, structural, industrial and software engineering. His research focus is on developing the means to enhance the ability of human designers to make decisions in the early stages of any product realization process. He is particularly interested in learning how to manage design freedom associated with the design, deployment, operation and support of open and sustainable engineering systems. Since 1992 he has been Professor of Mechanical Engineering at Georgia Tech. He received his Bachelor of Technology with Honours from I.I.T. Kharagpur, India and his M.S. and Ph.D. from the University of California, Berkeley in 1970 and 1974, respectively. Dr. Mistree can be reached at 404/894-8412 and farrokh.mistree@https://www.wendangku.net/doc/d412685186.html,.

Dimitris N. Mavris, Ph.D. is an Assistant Professor and Manager of the Aerospace Systems Design Laboratory Engineering at the Georgia Institute of Technology. Dr. Mavris is responsible for the research of 25 graduate students working in a variety of research areas which are sponsored by the U.S. Army, Air Force, Navy, NASA, and industry. He has authored 35 publications, and his working experience includes involvement in the study of rotor/fuselage interaction aerodynamics, helicopter and aircraft design, Navy Ship airwake turbulence modeling, and the High Speed Civil Transport metrics program as part of the NASA team. Dr. Mavris is presently involved with the study of advanced propulsion concepts for the Joint Advanced Strike Technology (JAST) program, and the design for affordability of future supersonic, very large subsonic transports, general aviation, and civil tiltrotors. Wei Chen, Ph.D., is an Assistant Professor in the Department of Mechanical Engineering at the University of Illinois at Chicago, where she is director of the Design & Integration Engineering Laboratory. Her main research area has been engineering design and manufacturing. Her research goal is to advance the theory, methods, and tools for those design and manufacturing problems with a magnitude of complexity, in the context of globally competitive markets and the need to quickly respond to the change. Her current research involves issues such as robust design, multicriteria decision making under risk and uncertainty, concept exploration of complex systems, distributed collaborative systems design, integrated product and process development, and computer aided design (CAD) and optimization. Dr. Chen received her Ph.D. from the Georgia Institute of Technology in 1995. She is the recipient of the 1996 NSF Faculty Early Career Award and 1998 ASME Pi Tau Sigma Gold Medal achievement award. She is a member of ASME, AIAA, and ASQC. Dr. Chen can be reached at 312/996-6072 and weichen1@https://www.wendangku.net/doc/d412685186.html,. Peter R?hl, Ph.D., received his Master Degree in Mechanical Engineering from Technical University of Braunschweig, Germany, in 1990, and his Ph.D. in Aerospace Engineering from the Georgia Institute of Technology in 1995, focusing on structural optimization and multidisciplinary optimization applied
transport aircraft design. Since then he has been working at the General Electric Corporate Research and Development Center in Schenectady, NY. His areas of research include the integration of design and manufacturing process simulation tools into an optimization and robust design framework, parametric CAD modeling, integration of CAD and CAE software and overall design tools development. Dr. R?hl can be reached at 518/3874004 and rohl@https://www.wendangku.net/doc/d412685186.html,. Kemper Lewis, Ph.D., is an Assistant Professor at the University at Buffalo in the Department of Mechanical and Aerospace Engineering. His research and education interests lie in developing decision support tools for multidisciplinary design, characterized by multiple objectives, uncertainty, and distributed design environments. Dr. Lewis is funded by the National Science Foundation, NASA-Langley Research Center, Praxair Inc., Rodgard Co., Allison-Rolls Royce Engine Co., and the Australian Defence Force Academy. Dr. Lewis can be reached at 716/645-2593, ext.2232 and kelewis@https://www.wendangku.net/doc/d412685186.html,. Patrick N. Koch, Ph.D., is currently employed by Engineous Software, Inc., where he is focusing on the development, implementation, and application of methods for addressing uncertainty, including robust optimization and six sigma techniques, reliability analysis and probabilistic optimization techniques, and Monte Carlo simulation techniques. Dr. Koch completed his doctoral research at Georgia Tech in December of 1997. He received his M.S. degree in Mechanical Engineering from Georgia Tech in 1994, and his B.S. in Mechanical Engineering with distinction from the University of Minnesota, Institute of Technology in 1992. Dr. Koch can be reached at 919/3197666 ext. 262 and koch@https://www.wendangku.net/doc/d412685186.html,.
to
supersonic

Carlos E. S. Cesnik, Ph.D., received his Engineering Degree in Aeronautics and Masters in Aeronautical Engineering from the Instituto Tecnológico de Aeronáutica (ITA, Brazil), and Ph.D. in Aerospace Engineering from the Georgia Institute of Technology. Previously to his current appointment as the Boeing Assistant Professor of Aeronautics and Astronautics at MIT, he worked as a research engineer at Embraer and as a research fellow at Georgia Tech. Prof. Cesnik’s educational and research interests include nonlinear aeroelastic modeling and analysis for multidisciplinary design environments, with emphasis on highly flexible structures and distributed embedded actuation. He teaches structural dynamics, aeroelasticity, and structural design courses. Dr. Cesnik can be reached at (617) 252-1518 and ccesnik@https://www.wendangku.net/doc/d412685186.html,. Tim Simpson, Ph.D., is an assistant professor at Penn State University with a joint appointment in the Departments of Mechanical & Nuclear Engineering and Industrial & Manufacturing Engineering. He obtained his Ph.D. and M.S. degrees in Mechanical Engineering in 1998 and 1995, respectively, from the Georgia Institute of Technology. His research interests include multidisciplinary design and optimization, surrogate metamodels for engineering design, and product family and product platform design for mass customization. He is a member of ASME, AIAA, and ASEE. Dr. Simpson can be reached at 814/863-7136 and tws8@https://www.wendangku.net/doc/d412685186.html,.