AIAA JOURNAL

Vol. 51, No. 9, September 2013

Multidisciplinary Design Optimization: A Survey of Architectures

Joaquim R. R. A. Martins
University of Michigan, Ann Arbor, Michigan 48109
and
Andrew B. Lambe
University of Toronto, Toronto, Ontario M3H 5T6, Canada
DOI: 10.2514/1.J051895
Multidisciplinary design optimization is a field of research that studies the application of numerical optimization
techniques to the design of engineering systems involving multiple disciplines or components. Since the inception of
multidisciplinary design optimization, various methods (architectures) have been developed and applied to solve
multidisciplinary design-optimization problems. This paper provides a survey of all the architectures that have been
presented in the literature so far. All architectures are explained in detail using a unified description that includes
optimization problem statements, diagrams, and detailed algorithms. The diagrams show both data and process flow
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

through the multidisciplinary system and computational elements, which facilitate the understanding of the various
architectures, and how they relate to each other. A classification of the multidisciplinary design-optimization
architectures based on their problem formulations and decomposition strategies is also provided, and the benefits and
drawbacks of the architectures are discussed from both theoretical and experimental perspectives. For each
architecture, several applications to the solution of engineering-design problems are cited. The result is a
comprehensive but straightforward introduction to multidisciplinary design optimization for nonspecialists and a
reference detailing all current multidisciplinary design-optimization architectures for specialists.

Nomenclature n = length of given variable vector

c = vector of design constraints R = governing equations of a discipline analysis in residual
cc = vector of consistency constraints form (discipline analysis constraints)
f = objective function x = vector of design variables
m = length of given constraint vector y = vector of coupling variables (outputs from a discipline
N = number of disciplines analysis)

Joaquim R. R. A. Martins is an Associate Professor at the University of Michigan, where he heads the
Multidisciplinary Design Optimization (MDO) Laboratory in the Department of Aerospace Engineering. His
research involves the development and application of MDO methodologies to the design of aircraft configurations,
with a focus on high-fidelity simulations that take advantage of high-performance parallel computing. Before joining
the University of Michigan faculty in September 2009, he was an Associate Professor at the University of Toronto
Institute for Aerospace Studies, where from 2002 he held a Tier II Canada Research Chair in Multidisciplinary
Optimization. He received his undergraduate degree in Aeronautical Engineering from Imperial College, London,
with a British Aerospace Award. He obtained both his M.Sc. and Ph.D. degrees from Stanford University, where he
was awarded the Ballhaus prize for best thesis in the Department of Aeronautics and Astronautics. He was a keynote
speaker at the International Forum on Aeroelasticity and Structural Dynamics in 2007 and the Aircraft Structural
Design Conference in 2010. He has received the Best Paper Award in the AIAA Multidisciplinary Analysis and
Optimization Conference twice (2002 and 2006). He is a member of the AIAA MDO Technical Committee and was the
technical cochair for the 2008 AIAA Multidisciplinary Analysis and Optimization Conference. He is an Associate
Editor for Optimization and Engineering and the AIAA Journal. He is also an Associate Fellow of the AIAA.

Andrew B. Lambe is a Ph.D. candidate at the University of Toronto Institute for Aerospace Studies. He received his
B.A.Sc. in Engineering Science from the University of Toronto, specializing in aerospace engineering. Before his
graduate studies, he worked at Environment Canada, applying optimization techniques to a remote sensing
application. He is currently engaged in research on gradient-based optimization algorithms, algorithms for large-
scale structural and multidisciplinary design problems, architectures for multidisciplinary design optimization, and
visualization methods for multidisciplinary analysis and optimization processes. He is a Student Member of the AIAA.

Received 8 March 2012; revision received 14 September 2012; accepted for publication 13 November 2012; published online 16 July 2013. Copyright © 2013 by
the authors. Published by the American Institute of Aeronautics and Astronautics, Inc., with permission. Copies of this paper may be made for personal or internal
use, on condition that the copier pay the $10.00 per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923; include the code
1533-385X/13 and $10.00 in correspondence with the CCC.
2049
 2050 MARTINS AND LAMBE
y
= vector of state variables (variables used inside only one
discipline analysis)
Subscripts
i = functions or variables that apply only to discipline i
0 = functions or variables that are shared by more than one
discipline
Superscripts/Oversets

= functions or variables at their optimal value
∼ = approximations of a given function or vector of functions
^ = independent copies of variables distributed to other
disciplines
I. Introduction
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
M ULTIDISCIPLINARY design optimization (MDO) is a
field of engineering that focuses on the use of numerical
optimization for the design of systems that involve a number of
disciplines or subsystems. The main motivation for using MDO is
that the performance of a multidisciplinary system is driven not only
by the performance of the individual disciplines but also by their
interactions. Considering these interactions in an optimization
problem generally requires a sound mathematical formulation. By
solving the MDO problem early in the design process and taking
advantage of advanced computational analysis tools, designers can
simultaneously improve the design and reduce the time and cost of
the design cycle.
The origins of MDO can be traced back to Schmit, Haftka, and
their collaborators [1–8], who extended their experience in structural
optimization to include other disciplines. One of the first applications
of MDO was aircraft wing design, where aerodynamics, structures,
and controls are three strongly coupled disciplines [9–16]. Since
then, the application of MDO has been extended to complete aircraft
[17–21] and a wide range of other engineering systems, such as
bridges [22], buildings [23,24], railway cars [25,26], microscopes
[27], automobiles [28,29], ships [30,31], propellers [32,33],
rotorcraft [34,35], wind turbines [36–38], and spacecraft [39–41].
One of the most important considerations when implementing
MDO is how to organize the discipline-analysis models, approxima-
tion models (if any), and optimization software in concert with the
problem formulation so that an optimal design is achieved. (In this
work, we mean “optimal” in the local sense since it is in general
difficult to assess global optimality.) Such a combination of problem
formulation and organizational strategy is referred to as an MDO
architecture. The MDO architecture defines both how the different
models are coupled and how the overall optimization problem is solved.
The architecture can be either monolithic or distributed. In a monolithic
approach, a single optimization problem is solved. In a distributed
approach, the same problem is partitioned into multiple subproblems
containing smaller subsets of the variables and constraints.
While many different architectures can be used to solve a given
optimal design problem, and just as many algorithms may be used to
solve a given optimization problem, the choice of the architecture has
a significant influence on both the solution time and the final design.
For example, using a global optimization algorithm rather than a
gradient-based algorithm may lead to a better final design because the
gradient-based optimizer may converge to a local minimum early
in the design process. However, if gradients can be computed
efficiently, the computational cost of the gradient-based optimization
may be far less than that of the global optimization because the
discipline analyses do not need to be run as many times [42]. If the
calculations required by a given architecture are easy to run in parallel
and if parallel computing facilities are available, a distributed
architecture may be preferred over a monolithic architecture despite
the increased computational expense. Human and computational
organization can also play a role. In practice, careful consideration of
the human and computing environments, the available algorithms,
and the design problem at hand is necessary to decide on the most
appropriate MDO architecture.
In this survey, we are primarily focused on methods for solving
MDO problems with a single objective function and continuous
design variables. We assume that the optimality of a given design
corresponds to the satisfaction of the Karush–Kuhn–Tucker (KKT)
optimality conditions (see, for example, Nocedal and Wright [43]
for a comprehensive mathematical definition). These conditions
are necessary for local optimality, so it is possible for different
architectures to obtain different (yet equally valid) local optima. The
KKT conditions require the availability of function gradients, so we
assume that the objective and constraint functions are differentiable.
However, various architectures have been developed specifically for
multi-objective problems [28,44–46] or problems with discrete
variables [47,48]. Other architectures determine optimality without
using gradients [49,50] or use optimality concepts from game theory
[51–53]. Whenever possible, we comment on the connections with
the work that go beyond the assumptions of this survey.
In the MDO literature, several terms are used to describe what
we call “architecture”: method [17,54–57], methodology [58–60],
problem formulation [61–64], strategy [65,66], procedure [67,68],
and algorithm [69–72] have all been used. Some authors use a variety
of terms and occasionally use them interchangeably in the same
paper. Our preference for the term architecture [73–76] comes from
the fact that the relationship between the problem formulation and the
solution algorithm is not one-to-one. For example, replacing a
particular discipline analysis with a surrogate model or reordering the
discipline analyses does not affect the problem formulation but
strongly affects the solution algorithm.
There have been a number of surveys of MDO over the last two
decades. Haftka et al. [77] were among the first to review the MDO
architectures known at the time. Cramer et al. [61] formalized
the monolithic architectures and detailed the required gradient
computation methods. Balling and Sobieszczanski-Sobieski [78]
identified a number of possible monolithic approaches and esti-
mated their computational cost. In a collection of articles entitled
Multidisciplinary Design Optimization: State of the Art, edited by
Alexandrov and Hussaini [79], Kroo [74] provided a comprehensive
overview of MDO, including a description of both monolithic and
distributed architectures. In the same volume, Alexandrov [80]
discussed the convergence properties of certain partitioning strat-
egies for distributed architectures and Balling [81] focused on
partitioning as a way to provide disciplinary autonomy. In the
same year, Sobieszczanski-Sobieski and Haftka [67] published an
exhaustive survey of the MDO literature up to that time.
Since this last series of surveys, MDO has continued to be an active
field of research. New architectures have been developed, and various
successful applications of MDO have taken place in industry [57,82–
84]. A recent paper by Tosserams et al. [85] identified numerous
architectures developed in the last decade that were not covered by
the previous surveys. However, there is currently no comprehensive
description of all the existing architectures that compares the features,
merits, and performance of each.
The purpose of this paper is to survey the available MDO
architectures and present them in a unified notation to facilitate
understanding and comparison. Furthermore, we propose the use of a
new standard diagram to visualize the algorithm of a given MDO
architecture, including how its components are organized and its data
flow. We pay particular attention to the newer MDO architectures that
have yet to gain widespread use. For each architecture, we discuss
its features and expected performance. We also present a new
classification of MDO architectures and show how they relate
mathematically. This classification is especially novel because it
is able to highlight similarities between architectures that were
developed independently.
For readers who want a quick overview, it is possible to skip most
of the text and still learn the essence of the various MDO architectures
by: 1) consulting the mathematical notation in the Nomenclature,
2) studying the problem formulation and pseudocode for each
architecture, and 3) consulting the MDO architecture classification
diagram in Fig. 1.
 Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

Fig. 1
Classification and summary of the MDO architectures.

2051 MARTINS AND LAMBE

 2052 MARTINS AND LAMBE
We have organized this paper as follows. In Sec. II, we present the
unified notation and diagrams for describing MDO architectures. In
Sec. III, we define the general MDO problem and describe the
three basic monolithic architectures and their derivation from a
common problem statement. In Sec. IV, we focus on the distributed
architectures; we discuss the motivation for using these methods and
their mathematical derivation. We also describe a new classification
of distributed architectures by drawing parallels with the monolithic
architectures, and we then explain the distributed MDO architectures
in detail, discussing their capabilities. In Sec. V, we survey some of
the benchmarking studies that have been performed to help decide
which architectures are most efficient for certain classes of design
problems. Finally, we summarize our conclusions in Sec. VI and
comment on possible future research in the field.
II. Unified Description of MDO Architectures
A. Terminology and Mathematical Notation
Before introducing the mathematical background, we introduce
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
the notation that we use throughout this paper. This notation allows us
to compare the various problem formulations within the architectures
and to identify how similar features of the general MDO problem are
handled in each case. The notation is listed in the Nomenclature. This
is not a comprehensive list; additional notation specific to particular
architectures will be introduced as necessary. We also take this
opportunity to clarify many of the terms we use that are specific to the
field of MDO.
A design variable is a quantity in the MDO problem that is always
under the explicit control of an optimizer. In traditional engineering
design, the values of these variables are selected explicitly by the
designer or design team. Design variables may be local (i.e., pertain to
a single discipline) or they may be shared by multiple disciplines. We
denote the vector of design variables local to discipline i by xi and the
shared variables by x0. The full vector of design variables is given by
x  xT0 ; xT1 ; : : : ; xTN T . The subscripts for local and shared data are
also used when describing objectives and constraints.
A discipline analysis is a simulation that models the behavior of
one aspect of a multidisciplinary system. Running a discipline
analysis consists of solving a system of equations (such as the
Navier–Stokes equations in fluid mechanics, the static equilibrium
equations in structural mechanics, or the equations of motion in a
control simulation) to compute a set of discipline responses, known
as state variables. State variables may or may not be controlled by the
optimization, depending on the formulation employed. We denote
the vector of state variables computed within discipline i by y
i. We
denote the associated set of disciplinary equations in residual form by
Ri , so that the expression Ri  0 represents the solution of these
equations with respect to y
i .
In a multidisciplinary system, most disciplines are required to
exchange coupling variables to model the interactions of the whole
system. Often, the number of variables exchanged is much smaller
than the total number of state variables computed in a particular
discipline. For example, in aircraft design, the structural analysis
does not require the states for the entire flowfield. Instead, only the
surface aerodynamic loads are required. The coupling variables
supplied by a given discipline i are denoted by y i. Another common
term for y i is response variables, since they describe the response of
the analysis to a design decision. In general, a transformation is
required to compute y i from y
i for each discipline [61]. Similarly, a
transformation may be needed to convert input coupling variables
into a usable format within each discipline [61]. In this work, the
mappings between y i and y
i are lumped into the analysis equations
Ri . This simplifies our notation with no loss of generality.
In many formulations, copies of the coupling variables must be
made to allow discipline analyses to run independently and in
parallel. These copies, which function as design variables in the
problem formulation, are sometimes called target variables. We
denote the coupling variable copies by y^. For example, the copy of the
coupling variables produced by discipline i is denoted y^ i . These
variables are independent of the corresponding original variables and
are used as part of the input to disciplines that are coupled to
discipline i through y i . To preserve consistency between the coupling
variable inputs and outputs at the optimal solution, we define a set
of consistency constraints, cci  y^ i − y i , which we add to the
optimization problem formulation.
B. Architecture Diagrams: The Extended Design Structure Matrix
While rewriting problem formulations for each architecture using a
common notation is a straightforward task, describing the sequence
of operations in the implementation in a convenient way presents a
significant challenge. Some authors merely present the problem
formulation and leave the readers to work out the implementation.
This is acceptable for monolithic architectures. For some of the
distributed architectures, however, the implementation is not
obvious. Other authors use an algorithm or flowchart as an aid, but
these are often inadequate for describing the data flow between
the different software components. Furthermore, more complex
strategies with multiple loops and parallel processes are difficult to
describe compactly using this technique. The lack of a standard
convenient graphical representation to describe the solution strategy
in MDO architectures is another impediment to understanding and
comparing their relative merits.
To enhance our exposition, each of the architectures is presented
with a new diagram that we call the extended design structure matrix,
or XDSM [86]. As the name suggests, the XDSM was based on the
design structure matrix (DSM) [87,88], a common diagram in
systems engineering that is used to visualize the interconnections
among components of a complex system. The traditional DSM
shows components and connections between components, but the
meaning of the connections is left ambiguous. To represent MDO
architectures, we need two types of connections: data dependency
and process flow. This need motivated the development of XDSM,
which simultaneously communicates data dependency and process
flow between computational components of MDO architectures on a
single diagram. We present a brief overview of the XDSM; for further
details, see Lambe and Martins [86].
We explain the basics of the XDSM using two simple examples.
The first example is shown in Fig. 2, and it represents a Gauss–Seidel
multidisciplinary analysis (MDA) procedure for three disciplines.
The pseudocode for this procedure is listed in Algorithm 1. As with
the traditional DSM, the components are laid out along the diagonal.
The components in this case consist of the discipline analyses and a
special component, known as a driver, that controls the iteration
and is represented by a rounded rectangle. The function of the
components is to process data. The data flow is shown as thick gray
lines. The components take data inputs from the vertical direction and
output data in the horizontal direction. Thus, the connections above
the diagonal flow from left to right and top to bottom, and the
connections below the diagonal flow from right to left and bottom to
top. The off-diagonal nodes in the shape of parallelograms are used to
label the data. Using this convention, it is easy to identify the inputs of
a given component by scanning the columns above and below the
component, while the outputs can be identified by scanning the row.
External inputs and outputs are placed on the outer edges of the
diagram, in the top row and leftmost column, respectively. In Fig. 2,
the external inputs are the design variables and an initial guess of the
system coupling variables. Each discipline analysis computes its own
set of coupling variables that is passed to other discipline analyses or
back to the driver. At the end of the MDA process, each discipline
returns the final set of coupling variables computed.
The thin black lines show the process flow. The direction of these
lines follows the convention for the data-flow lines. In addition, a
numbering system is used to show the order in which the components
are executed. These numbers are presented inside each component
in the diagram followed by a colon and the component name. As
the algorithm executes, every time a number is reached, the
corresponding component performs a relevant computation. Multiple
numbers indicate that the component is called several times within the
algorithm. The algorithm starts at component 0 and proceeds in
numerical order, following the process lines. Loops are denoted by
j → k for k < j, indicating that the algorithm must return to step k until
 MARTINS AND LAMBE
Fig. 2 Block Gauss–Seidel MDA process to solve a three-discipline
coupled system.
some condition required by the driver is satisfied. The data nodes are
also labeled with numbers to indicate when the input data are retrieved.
The second XDSM example, illustrated in Fig. 3, is the solution
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
process for an analytic optimization problem using gradient-based
optimization. The problem has a single objective and a vector of
constraints. Figure 3 shows separate components to compute the
objective, the constraints, and their gradients, as well as a driver to
control the iteration. We assume that the gradients can be computed
without knowledge of the objective and constraint function values.
Notice that multiple components are evaluated at step 1 of the
algorithm; this numbering denotes parallel execution. In some cases,
it may be advisable to combine components, e.g., the objective and
the constraints, to reflect underlying problem structures. In the
following sections, we have done so to simplify the presentation; we
will mention other simplifications as we proceed.
III. Monolithic Architectures
A. Introduction
If we ignore the discipline boundaries, an MDO problem is nothing
more than a standard constrained nonlinear programming problem:
we must find the values of the design variables that maximize or
minimize a particular objective function, subject to the constraints.
The choice of the objective, the constraints, and even what variables
to change in a given system is strictly up to the designer. The behavior
of each component, or discipline, within the system is modeled using
a discipline analysis. Each analysis is usually available in the form of
a computer program and can range in complexity from empirical
curve-fit data to a highly detailed physics-based simulation.
One of the major challenges of MDO is how to manage the
coupling of the system under consideration. Like the disciplines they
model, the discipline analyses are mutually interdependent: one
analysis requires the outputs of other analyses as input. Furthermore,
the objective and constraint functions, in general, depend on both the
design variables and the analysis outputs from multiple disciplines.
While this interdependence is sometimes ignored in practice through
the use of single-discipline optimizations occurring in parallel or in
sequence, taking the interdependence into account generally leads to
a more accurate representation of the behavior of the system. MDO
architectures provide a consistent, formal setting for managing this
interdependence in the design process [74].
The architectures presented in this section are referred to as
monolithic architectures. Each architecture solves the MDO problem
Algorithm 1 Block Gauss–Seidel
multidisciplinary analysis for three disciplines
Input: Design variables x
Output: Coupling variables y
0: Initiate MDA iteration loop
repeat
1: Evaluate analysis 1 and update y 1 y 2 ; y 3 
2: Evaluate analysis 2 and update y 2 y 1 ; y 3 
3: Evaluate analysis 3 and update y 3 y 1 ; y 2 
until 4 → 1: MDA has converged
2053
Fig. 3 Gradient-based optimization procedure.
by casting it as a single optimization problem; different strategies
are used to achieve multidisciplinary feasibility. Architectures that
decompose the optimization problem into smaller problems, i.e.,
distributed architectures, are presented in Sec. IV.
B. All-at-Once Problem Statement
Before discussing specific architectures, we show the fundamental
optimization problem from which all other problem statements
are derived. We can describe the MDO problem in its most general
form as
X
N
minimize f0 x; y  fi x0 ; xi ; y i 
i1
with respect to x; y^ ; y; y

subject to c0 x; y ≥ 0
ci x0 ; xi ; y i  ≥ 0 for i  1; : : : ; N
cci  y^ i − y i  0 for i  1; : : : ; N
Ri x0 ; xi ; y^ j≠i ; y
i ; y i   0 for i  1; : : : ; N (1)
which is known as the all-at-once (AAO) problem. This form
of the design-optimization problem includes all coupling variables,
coupling variable copies, state variables, consistency constraints, and
residuals of the governing equations directly in the problem statement.
For the design variables, we concatenate the discipline variable
groups as follows: x  xT0 ; xT1 ; : : : ; xTN T . The same is done for
the coupling variables, y  y T1 ; : : : ; y TN T , as well as the
state variables. The objective functions fi represent local objective
functions, i.e., objective functions that depend only on shared design
variables and other variables that are local with respect to only one
discipline (discipline i). We omit the sum of local objective functions
except when necessary to highlight situations where an architecture
exploits this term.
For the design constraints, c  cT0 ; cT1 ; : : : ; cTN T , we adopt the
convention of using only greater-than-or-equal-to inequalities. There
is no loss of generality with this convention, since the sign can be
switched for less-than-or-equal-to inequalities, and equalities can be
stated as pairs of inequalities with opposing signs.
Figure 4 shows the XDSM for the AAO problem. To keep the
diagrams compact, we adopt two conventions. First, unlike Fig. 3, we
do not explicitly show the gradient computation steps unless they are
intrinsic to the architecture. Second, any block referring to discipline
i represents a repeated pattern for every discipline. Thus, in Fig. 4, a
residual block exists for every discipline, and each block can be
executed in parallel. As an added visual cue, the “Residual i”
component is displayed as a stack of similar components.
There is a conflict with the literature when it comes to the labeling
of problem (1). What most authors refer to as AAO, following the
lead of Cramer et al. [61], others label as the simultaneous analysis
and design (SAND) [78,89] problem. Our AAO problem is most like
what Cramer et al. refer to as simply “the most general formulation”
[61]. Herein, we classify formulation (1) as the AAO problem,
because it includes all the design, state, and input and output coupling
variables in the problem, so the optimizer is responsible for all
 2054
Fig. 4 Diagram for the AAO problem.
variables at once. The SAND architecture has a different problem
statement and is presented in Sec. III.C.
The AAO problem is never solved in this form in practice because
the consistency constraints, which are linear in this formulation,
can be easily eliminated. Eliminating these constraints reduces
the problem size without compromising the performance of the
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
optimization algorithm. As we will see, eliminating the consistency
constraints from problem (1) results in the problem solved by the
SAND architecture. However, we have presented the AAO problem
first because it functions as a common starting point for deriving both
the SAND problem and the individual discipline feasible (IDF)
problem and, subsequently, all other equivalent MDO problems.
Depending on which equality constraint groups are eliminated
from problem (1), we can derive the other three monolithic
architectures: multidisciplinary feasible (MDF), individual discipline
feasible, and simultaneous analysis and design. All three have been
known in the literature for a long time [61,78,89,90]. In the next three
subsections, we describe how each architecture is derived and the
relative advantages and disadvantages of each. We emphasize that, in
spite of the elements added or removed by each architecture, we are
always solving the same MDO problem. Furthermore, each problem
has the same set of optimal solutions.
C. Simultaneous Analysis and Design
The most obvious simplification of problem (1) is to eliminate the
consistency constraints, cci  y^ i − y i  0, by introducing a single
group of coupling variables to replace the separate target and
response groups. This simplification yields the SAND architecture
[89], which solves the following optimization problem:
minimize f0 x; y
with respect to x; y; y

subject to c0 x; y ≥ 0
ci x0 ; xi ; y i  ≥ 0 for i  1; : : : ; N
Ri x0 ; xi ; y; y
i   0 for i  1; : : : ; N
The XDSM for SAND is shown in Fig. 5. Cramer et al. [61] refer to
this architecture as “all-at-once”. However, we use the name SAND
to reflect the consistent set of analysis and design variables chosen by
the optimizer. The optimizer, therefore, can simultaneously analyze
and design the system.
Fig. 5 Diagram for the SAND architecture.
MARTINS AND LAMBE
Several features of the SAND architecture are noteworthy.
Because we do not need to solve any discipline analysis explicitly or
exactly at each iteration, the optimization problem can potentially be
solved quickly by letting the optimizer explore regions that are
infeasible with respect to the analysis constraints Ri . The SAND
methodology is not restricted to multidisciplinary systems and can
be used in single-discipline optimization as well. In that case, we
need to define only a single group of design constraints c. If the
discipline residual equations are simply discretized partial differ-
ential equations (PDEs), the SAND problem is just a PDE-
constrained optimization problem like many others in the literature.
(See Biegler et al. [91] for an overview of this field.)
Two major issues are present in the SAND architecture. First, the
problem formulation still requires all state variables and discipline-
analysis equations, so problem size and potential premature
termination of the optimizer at an infeasible design can be issues in
practice. Second, and more important, is the fact that the discipline-
analysis equations are treated explicitly as constraints, which means
that the residual values (and possibly their derivatives) need to be
available to the optimizer. In other words, rather than computing
coupling variables y i and state variables y
i , each discipline i accepts
predetermined values of y i and y
i and returns analysis equation
residuals Ri . (In our notation, Ri includes the transformation
between y and y
for disciplines coupled to discipline i.) In
engineering design, software for discipline analysis often operates in
a black-box fashion, directly computing the coupling variables while
hiding the discipline-analysis residuals and state variables. Even if
the software can be modified to return the residuals, the cost and effort
required may be excessive. Therefore, most practical MDO problems
require an architecture that can take advantage of existing discipline-
analysis software. The following two monolithic architectures
address this concern.
D. Individual Discipline Feasible
By eliminating the discipline-analysis constraints Ri x0 ; xi ;
y i ; y^ j≠i ; y
i   0 from problem (1), we obtain the IDF architecture
[61]. As commonly noted in the literature, this type of elimination is
achieved by applying the implicit function theorem to the Ri
constraints so that y
i and y i become functions of design variables and
coupling variable copies. The IDF architecture is also known as
distributed analysis optimization [62] and optimizer-based decom-
position [74]. The optimization problem for the IDF architecture is
minimize f0 x; yx; y^ 
with respect to x; y^
subject to c0 x; yx; y^  ≥ 0
ci x0 ; xi ; y i x0 ; xi ; y^ j≠i  ≥ 0 for i  1; : : : ; N
(2) cci  y^ i − y i x0 ; xi ; y^ j≠i   0 for i  1; : : : ; N (3)
The most important consequence of this reformulation is the removal
of all the state variables and discipline-analysis equations from the
problem statement. All the coupling variables are now implicit
functions of design variables and coupling variable copies as a result
of the discipline-analysis equations being solved exactly at each
iteration.
The XDSM for IDF is shown in Fig. 6. This architecture enables
the discipline analyses to be performed in parallel, since the coupling
between the disciplines is resolved by the coupling variable copies y^
and consistency constraints cc . Within the optimization iteration,
specialized software for the discipline analyses can now be used to
return the coupling variable values to the objective and constraint
function calculations. The net effect is that the IDF problem is
substantially smaller than the SAND problem and requires minimal
modification to existing discipline analyses. In the field of PDE-
constrained optimization, the IDF architecture is exactly analogous to
a reduced-space method [91].
However, the size of the IDF problem can still be an issue. If the
number of coupling variables is large, the resulting optimization
 MARTINS AND LAMBE
Fig. 6 Diagram of the IDF architecture.
problem might be too large to solve efficiently. This can be mitigated
to some extent by careful selection of the discipline variable partitions
or aggregation of the coupling variables to reduce the information
transfer between disciplines.
If gradient-based optimization software is used to solve the IDF
problem, the gradient computation can become an issue. When the
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
discipline analyses are expensive, evaluating the objective and
constraint function gradients becomes costly. This is because the
gradients themselves must be discipline-feasible, i.e., the changes in
the design variables cannot cause the output coupling variables to
violate the discipline-analysis equations to the first order.
In practice, gradients are often calculated using some type of finite-
differencing procedure, where the discipline analysis is evaluated for
each design variable. While this approach preserves discipline
feasibility, it is costly and unreliable. If the discipline-analysis code
allows for the use of complex numbers, the complex-step method
[92,93] is an alternative approach that gives machine-precision
derivative estimates. If the analysis codes require a particularly long
time to evaluate, automatic differentiation or analytic derivative
calculations (direct or adjoint methods) can be used to avoid multiple
discipline-analysis evaluations [94–98]. While the development
time for these methods can be long, the reward is accurate deriv-
ative estimates and massive reductions in the computational cost,
especially for design optimization based on high-fidelity models [99–
101]. We refer the reader to the work of Martins and Hwang [102]
for a detailed discussion of the options available for computing
derivatives in MDO problems.
E. Multidisciplinary Feasible
If both the analysis and consistency constraints are removed from
problem (1), we obtain the MDF architecture [61]. This architecture
has also been referred to in the literature as fully integrated
optimization [62] and nested analysis and design [78]. The resulting
optimization problem is
Fig. 7 Diagram for the MDF architecture with a Gauss–Seidel multidisciplinary analysis.
2055
minimize f0 x; yx; y
with respect to x
subject to c0 x; yx; y ≥ 0
ci x0 ; xi ; y i x0 ; xi ; y j≠i  ≥ 0 for i  1; : : : ; N (4)
The MDF architecture XDSM for three disciplines is shown in Fig. 7.
Typically, a fixed-point iteration, such as the block Gauss–Seidel
iteration shown in Fig. 7, is used to converge the MDA, where each
discipline is solved in turn. This approach usually exhibits slow
convergence rates. Reordering the sequence of disciplines can
improve the convergence rate of Gauss–Seidel [103], but even better
convergence rates can be achieved through the use of Newton-based
methods [104]. Because of the sequential nature of the Gauss–Seidel
iteration, we cannot evaluate the disciplines in parallel and cannot
apply our convention for compacting the XDSM. Using a different
MDA method results in a different XDSM.
An obvious advantage of MDF over the other monolithic
architectures is that the optimization problem is as small as it can be
for a monolithic architecture, since only the design variables,
objective function, and design constraints are under the direct control
of the optimizer. Another benefit is that MDF returns a system design
that always satisfies the consistency constraints, even if the
optimization process is terminated early. This is advantageous in
an engineering-design context if time is limited and our concern
is to find an improved design that need not be optimal in the
strict mathematical sense. Note, however, that design constraint
satisfaction is not guaranteed if the optimization is terminated early;
this depends on whether or not the optimization algorithm maintains
a feasible design point. In particular, methods of feasible directions
[105] require and maintain a feasible design point while many robust
sequential quadratic programming [106] and interior point methods
[107] do not.
The main disadvantage of MDF is that a consistent set of coupling
variables must be computed and returned to the optimizer every time
the objective and constraint functions are reevaluated. In other words,
the architecture requires a full MDA to be performed for every
optimization iteration. Instead of simply running each individual
discipline analysis once per iteration, as we do in IDF, we need to run
every analysis multiple times until a consistent set of coupling
variables is found. This task requires its own specialized iterative
procedure outside the optimization. Developing an MDA procedure
can be time-consuming.
Gradient calculations are also much more difficult for MDF than
for IDF. Just as the gradient information in IDF must be discipline-
feasible, the gradient information in MDF must be feasible with
respect to all disciplines. Fortunately, research in the sensitivity of
 2056
coupled systems is fairly mature, and there are semi-analytic methods
that drastically reduce the cost of this step by eliminating finite
differencing over the full MDA [99,100,102,108–110]. There has
also been some work on automating the implementation of these
coupled sensitivity methods [41,111,112]. The required partial
derivatives can be obtained using any of the methods described in
Sec. III.D for the individual disciplines in IDF.
IV. Distributed Architectures
A. Motivation
Thus far, we have focused our discussion on monolithic
MDO architectures: those that form and solve a single optimiza-
tion problem. Many more architectures have been developed
that decompose this optimization problem into a set of smaller
optimization problems, or subproblems, that have the same solution
when reassembled. These are the distributed MDO architectures.
Before reviewing and classifying the distributed architectures, we
discuss the motivation for the development of this new class of MDO
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
architectures.
Early in the history of optimization, the motivation for decom-
position methods was to exploit the structure of the problem to reduce
solution time. Many large optimization problems, such as network
flow problems and resource allocation problems, exhibit such special
structure [113].
To better understand decomposition, consider the following
problem:
X
N defined in terms of knowledge areas, e.g., aircraft design disciplines
minimize fi xi 
i1
with respect to x1 ; : : : ; xN
subject to c0 x1 ; : : : ; xN  ≥ 0
c1 x1  ≥ 0; : : : ; cN xN  ≥ 0
In this problem, there are no shared design variables x0 , and the
objective function is separable, i.e., it can be expressed as a sum
of functions, each of which depends only on the corresponding
local design variables xi . On the other hand, the constraints include a
set of constraints c0 that depends on more than one set of design
variables. This is referred to as a problem with complicating
constraints [114]; if c0 did not exist, we could simply decompose this
optimization problem into N independent problems and solve them in
parallel.
Another possibility is that a problem may include shared design
variables and a separable objective function, with no complicating
constraints, i.e.,
X
N
minimize fi x0 ; xi 
i1
with respect to x0 ; x1 ; : : : ; xN
subject to c1 x0 ; x1  ≥ 0; : : : ; cN x0 ; xN  ≥ 0
This is referred to as a problem with complicating variables [114]. In
this case, the decomposition would be straightforward if there were
no shared design variables x0 and we could solve N optimization
problems independently and in parallel.
Specialized decomposition methods were developed to reintro-
duce the complicating variables or constraints into these problems
with only small increases in time and cost relative to the N
independent problems. Examples of these methods include the
Dantzig–Wolfe decomposition [115] and Benders decomposi-
tion [116] for problems (5) and (6), respectively. However, these
decomposition methods were designed to work with the simplex
algorithm on linear programming problems. In the simplex algo-
rithm, the active set changes by only one constraint at a time, so
decomposition is the only way to exploit the special structure.
Algorithms for nonlinear optimization that are based on Newton’s
MARTINS AND LAMBE
method, such as sequential quadratic programming and interior point
methods, may also use specialized matrix factorization techniques to
exploit sparsity in the problem. Nonlinear decomposition algorithms
have also been developed but, to the best of our knowledge, no
performance comparisons have been made between decomposition
algorithms and Newton-like algorithms that employ sparse matrix
factorization. Intuition suggests that the latter should be faster since
Newton methods can exploit second-order problem information.
Thus, while decomposition methods do exist for nonlinear problems,
the problem structure is not the primary motivation for their
development.
The primary motivation for decomposing the MDO problem
comes from the structure of the engineering-design environment.
Typical industrial practice involves breaking up the design of a large
system and distributing it among specific engineering groups. These
groups may be geographically distributed and may communicate
infrequently. Furthermore, these groups typically like to retain
control of their own design procedures and make use of in-house
expertise, rather than simply passing on the discipline-analysis
results to a central design authority [74]. Decomposition through
distributed architectures allows individual design groups to work in
isolation, controlling their own sets of design variables, while
periodically receiving updated information from other groups to
improve their aspect of the overall design. This approach conforms
more closely to current industrial design practice than does the
approach of the monolithic architectures.
In an industrial setting, the notion of a discipline in an MDO
problem can take many forms. Traditionally, disciplines have been
include aerodynamics, structures, propulsion, and control. This
definition conforms well with many existing analysis codes. In
principle, however, a discipline can constitute any suitable partition
of an MDO problem. For example, if the structural design is broken
down by component (e.g., wings, tail, and fuselage), the design of
(5) each component could also be considered a discipline. Therefore, an
industrial-scale MDO problem could contain hundreds of disciplines,
depending on the company architecture. How the disciplines are
arranged within a distributed MDO architecture is up to the company,
but there is some literature on choosing discipline partitions to reduce
the coupling between distributed groups [117–119].
The structure of discipline design groups working in isolation has a
profound effect on the timing of each discipline-analysis evaluation.
In a monolithic architecture, all discipline-analysis programs are run
exactly the same number of times, based on requests from the
optimizer or MDA program. In the context of parallel computing, this
approach can be thought of as a synchronous algorithm [120]. In
instances where some analyses or optimizations are much more
expensive than others, such as multifidelity optimization [121,122],
the performance suffers because the processors performing the
inexpensive analyses and optimizations experience long periods of
inactivity while waiting for updates from other processors. In the
language of parallel computing, the computation is said to exhibit
poor load balancing. Another example is aerostructural optimization,
in which a nonlinear aerodynamics solver may require an order of
(6)
magnitude more time to run than a linear structural solver [123].
By decomposing the optimization problem, we can balance the
processor workloads by allowing discipline analyses with lower
computational costs to perform more optimization on their own.
Those disciplines with less demanding optimizations may also
be allowed to make more progress before updating nonlocal
information. In other words, the design process occurs not only in
parallel but also asynchronously. Overall, this may result in more
computational effort, but the intrinsically parallel nature of the
architecture allows much of the work to proceed concurrently,
reducing the wall-clock time.
B. Classification
We now introduce a new approach to classifying MDO
architectures. Some of the previous classifications were based on
which constraints could be controlled by the optimizer [78,124].
 MARTINS AND LAMBE
Alexandrov and Lewis [124] used the terms “closed” to indicate that a
set of constraints cannot be satisfied via explicit action of the
optimizer and “open” to indicate that it can. For example, the
MDF architecture is closed with respect to both the analysis and
consistency constraints, because their satisfaction is determined
through the process of converging the MDA. Similarly, IDF is closed
with respect to the former but not the latter, since the consistency
constraints can be satisfied by the optimizer adjusting the coupling
variable copies and design variables. Tosserams et al. [85] expanded
on this classification scheme by discussing whether distributed
architectures used open or closed local design constraints in the
system subproblem. Closure of the constraints is an important
consideration when selecting an architecture, because most robust
optimization software will permit the exploration of infeasible
regions of the design space. Such exploration can result in faster
solution via fewer optimization iterations, but this must be weighed
against the increased problem size and the risk of terminating the
optimization at an infeasible point.
The central idea in our classification is that distributed MDO
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
architectures can be classified based on their monolithic analogs:
MDF, IDF, or SAND. This stems from the different approaches to the
state and coupling variables in the monolithic architectures. Those
distributed architectures that follow MDF and use an MDA (or an
approximation of an MDA) to enforce coupling variable consistency
at the final solution are classified as distributed MDF architectures.
Similarly, those distributed architectures that follow IDF and use
coupling variable copies and consistency constraints to enforce
consistency at the final solution are classified as distributed IDF
architectures.
Our classification is similar to the previous classifications in
that an equality constraint must be removed from the optimization
problem (i.e., closed) for every variable removed from the problem
statement. However, our classification makes it much easier to see the
connections between distributed architectures, even when these
architectures are developed in isolation. In many cases, the problem
formulation in the distributed architecture can be derived directly
from that of the monolithic architecture by adding certain elements to
the problem, by making certain assumptions, or by applying a
specific decomposition scheme. Our classification can also be viewed
as a framework in which researchers can develop new distributed
architectures, since the starting point for a distributed architecture is
always a monolithic architecture.
This architecture classification is shown in Fig. 1. The relationships
between the architectures are shown by arrows. Some distributed
architectures have many variants, such as versions incorporating
surrogate models (also known as metamodels or response surfaces)
and multi-objective versions; we have included only the core
architectures in this diagram. We discuss the different versions of each
distributed architecture in the corresponding sections.
No distributed architecture developed to date is an analog of
SAND. As discussed in Sec. III, the desire to use independent black-
box computer software for the discipline analyses necessarily
excluded consideration of the SAND formulation as a starting point.
Nevertheless, the techniques used to derive distributed architectures
from IDF and MDF may also be useful when using SAND as a
foundation.
Our classification scheme does not distinguish between the dif-
ferent solution techniques for the distributed optimization problems.
For example, we have not focused on the order in which the
distributed problems are solved. Coordination schemes are partially
addressed in the distributed IDF group, where we have classified
the architectures as either penalty or multilevel, based on whether
the coordination uses penalty functions or a problem hierarchy.
This grouping follows from the work of De Wit and Van
Keulen [125].
One area that is not well explored in MDO is the use of hybrid
architectures. A hybrid architecture incorporates elements of two or
more architectures in such a way that different discipline analyses
or optimizations are treated differently. For example, a hybrid
monolithic architecture could be created from MDF and IDF by
resolving the coupling of some disciplines within an MDA, while the
2057
remaining coupling variables are resolved through constraints. Some
ideas for hybrid architectures have been proposed by Marriage and
Martins [41,111,112] and Geethaikrishnan et al. [126]. Such
architectures could be especially useful in applications where
the coupling characteristics vary widely among the disciplines.
However, general rules need to be developed to specify under what
conditions the use of certain architectures is advantageous. As we will
note in Sec. V, much work remains to be done in this area.
The relative performance of the MDO architectures is discussed
later, in Sec. V. However, we should make an important remark on
benchmarking before describing the various distributed MDO
architectures. Because timing results vary based on the hardware and
software implementations, the most reliable performance metric
when using a gradient-based optimizer is to count the number of
times the discipline analyses (and, by extension, the objective
and constraint functions) are evaluated. Furthermore, because the
discipline analyses often constitute the bulk of the computational
work, this metric can serve as a proxy for the total computational
time. However, this does not take into account the parallel nature of
many computations. Therefore, we should keep in mind that the
comparison is less reliable for measuring the wall-clock time in a
parallel computing environment.
In the following sections, in chronological order of their initial
development, we introduce the distributed architectures for MDO.
We prefer to use the term “distributed” as opposed to hierarchical or
multilevel because these architectures do not necessarily create a
hierarchy of problems to solve. In some cases, it is better to think of all
the optimization problems as being on the same level. Furthermore,
neither the systems being designed nor the design team organization
need to be hierarchical in nature for these architectures to be
applicable. Our focus here is to provide a unified description of these
architectures and to explain the advantages and disadvantages of
each. Along the way, we will point out variations and applications
that can be found in the literature. We also aim to review the state of
the art, since the most recent detailed architecture survey in the
literature is now more than 15 years old [67]. More recent surveys,
such as that of Agte et al. [57], discuss MDO more generally without
discussing the architectures in detail.
C. Concurrent Subspace Optimization
Concurrent subspace optimization (CSSO) is one of the oldest
distributed architectures for large-scale MDO problems. The original
formulation [127,128] decomposes the system problem into inde-
pendent subproblems with disjoint sets of variables. Coupled
sensitivity information is calculated at each iteration to give each
subproblem a linear approximation to an MDA, improving the
convergence behavior. At the system level, a coordination problem is
solved to recompute the “responsibility,” “tradeoff,” and “switch”
coefficients assigned to each discipline to provide information on the
design variable preferences for nonlocal constraint satisfaction.
Using these coefficients gives each discipline a certain degree of
autonomy within the system as a whole. Shankar et al. [129] proposed
several improvements to the original architecture, including methods
for updating the coefficients, and tested them on two- and three-
variable quadratic optimization problems. Unfortunately, they note
that the architecture performance is sensitive to parameter selection,
and extensive tuning may be required to run CSSO efficiently on
larger nonlinear problems.
Several variations of this architecture have been developed to
incorporate surrogate models [65,69,130] and higher-order informa-
tion sharing among the disciplines [131]. More recently, the
architecture has been adapted to solve multi-objective problems
[132–134]. Parashar and Bloebaum [135] extended a multi-objective
CSSO formulation to robust design-optimization problems. An
application to the design of high-temperature aircraft engine
components was presented by Tappeta et al. [136].
The version we consider here, due to Sellar et al. [69], uses
surrogate models of each discipline analysis to efficiently model
multidisciplinary interactions. Using our unified notation, the CSSO
system subproblem is given by
 2058 MARTINS AND LAMBE
minimize f0 x; yx;
~ y~
with respect to x
subject to c0 x; yx;
~ y~ ≥0
ci x0 ; xi ; y~ i x0 ; xi ; y~ j≠i  ≥ 0 for i  1; : : : ; N (7)
and the discipline i subproblem is given by
minimize f0 x; y i xi ; y~ j≠i ; y~ j≠i 
with respect to x0 ; xi
subject to c0 x; yx;
~ y~ ≥0
ci x0 ; xi ; y i x0 ; xi ; y~ j≠i  ≥ 0
cj x0 ; y~ j x0 ; y
~ ≥0
for j  1; : : : ; i − 1; i  1; : : : ; N (8)
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
The CSSO architecture is depicted in Fig. 8, and the corresponding
steps are listed in Algorithm 2. Note that the architecture uses a
surrogate model for every discipline and MDA methods that may,
depending on the step in the algorithm, call the surrogate models
directly instead of the discipline analyses. A potential pitfall of this
architecture is the need to include all the design variables in the
system subproblem. For industrial-scale design problems, this may
not always be possible or practical.
There have been some benchmarks that compare CSSO with other
MDO architectures. Perez et al. [137], Yi et al. [138], and Tedford and
Martins [139] all compared CSSO to other architectures on low-
dimensional test problems with gradient-based optimization. Their
results show that CSSO required many more analysis calls and
function evaluations to converge to an optimal design. The results of
de Wit and van Keulen [140] show that CSSO was unable to reach the
optimal solution of even a simple minimum-weight two-bar truss
problem. Thus, CSSO seems to be largely ineffective when compared
with newer MDO architectures.
D. Collaborative Optimization
In collaborative optimization (CO), the discipline optimization
subproblems are made independent of each other by using copies of
the coupling and shared design variables [141,142]. These copies are
then shared with all the disciplines during every iteration of the
solution procedure. (In many of the references cited, these variable
copies are also known as targets.) The complete independence of the
discipline subproblems combined with the simplicity of the data-
sharing protocol makes this architecture attractive for problems with
a small amount of shared design information.
Braun [141] formulated two versions of the CO architecture: CO1
and CO2 . CO2 is more frequently used, so it will be the focus of our
discussion. The CO2 system subproblem is given by
minimize f0 x0 ; x^ 1 ; : : : ; x^ N ; y^ 
with respect to x0 ; x^ 1 ; : : : ; x^ N ; y^
subject to c0 x0 ; x^ 1 ; : : : ; x^ N ; y^  ≥ 0
J
i  kx^ 0i − x0 k22  kx^ i − xi k22 
k^y i − y i x^ 0i ; xi ; y^ j≠i k22  0 for i  1; : : : ; N (9)
where the x^ 0i are copies of the shared design variables passed to (and
manipulated by) discipline i, and the x^ i are copies of the local design
variables passed to the system subproblem. These copies are
independent variables for which the values are chosen by a different
subproblem. Equality constraints are used to ensure that both copies
agree on a single value at an optimal design. Copies of the local
design variables are made only if those variables directly influence
the objective. In CO1 , the quadratic equality constraints are replaced
with linear equality constraints for each shared variable and its copy.
In either case, if derivatives are required to solve the system
subproblem, they must be derivatives of the optimized function J
i
with respect to the system variables. Although the architecture has yet
to converge on an optimized system design, this step is referred to as a
postoptimality analysis because the subsystems have been optimized
with respect to their local information.
The discipline i subproblem in both CO1 and CO2 is
minimize Ji x^ 0i ; xi ; y i x^ 0i ; xi ; y^ j≠i 
with respect to x^ 0i ; xi
subject to ci x^ 0i ; xi ; y i x^ 0i ; xi ; y^ j≠i  ≥ 0 (10)
Thus, the system-level problem is responsible for minimizing the
design objective, while the discipline-level problems minimize
system inconsistency. Braun [141] showed that the CO problem
statement is mathematically equivalent to the IDF problem statement
(3), and therefore equivalent to the original MDO problem (1) as well.
In particular, if the CO architecture converges to a point that locally
minimizes f0 while satisfying the design constraints c0 and ci and
consistency constraints J i  0, the resulting point must also be a local
minimum of the IDF problem. This can be inferred from the special
structure of problems (9) and (10). CO is depicted by the XDSM in
Fig. 9. The corresponding procedure is detailed in Algorithm 3.
In spite of the organizational advantage of fully separate discipline
subproblems, CO has major weaknesses in the mathematical
formulation that lead to poor performance in practice [62,143]. In
particular, the system problem in CO1 has more equality constraints
than variables, so the system subproblem is often infeasible. This can
also happen in CO2 , but it is not the most problematic issue. The most
significant difficulty with CO2 is that the constraint gradients of the
system problem at an optimal solution are all zero vectors. This
represents a breakdown in the constraint qualification of the Karush–
Kuhn–Tucker optimality conditions, which slows down convergence
for most gradient-based optimization software [62]. In the worst case,
the CO2 formulation may not converge at all. These difficulties with
the original formulations of CO have inspired several researchers to
improve the behavior of the architecture.
In a few cases, problems have been solved with CO and a gradient-
free optimizer, such as a genetic algorithm [49] or a gradient-based
optimizer that handles the troublesome constraints by avoiding the
use of Lagrange multipliers in the termination condition [144]. While
such approaches do avoid the obvious problems with CO, they
introduce other issues. Gradient-free optimizers that do not employ
some kind of surrogate modeling tend to require many more func-
tion evaluations than do gradient-based optimizers. These addi-
tional function evaluations and discipline analyses can become
a bottleneck. Gradient-based optimizers that avoid Lagrange
multipliers, such as feasible direction methods, often fail in nonconvex
feasible regions. As pointed out by DeMiguel and Murray [143], the
CO system subproblem is set-constrained (i.e., nonconvex) because of
the need to satisfy optimality in the discipline subproblems.
Several researchers successfully use CO together with robust
gradient-based optimization methods. DeMiguel and Murray [143]
relax the troublesome constraints using an L1 exact penalty function
with a fixed penalty parameter value, and they add elastic variables to
preserve the smoothness of the problem. This revised approach is
called modified collaborative optimization (MCO). This approach
satisfies the requirement of mathematical rigor, since algorithms
using the penalty function formulation are known to converge to an
optimal solution under mild assumptions [43,145]. However, the
results of Brown and Olds [146] show strange behavior in a practical
design problem. In particular, when the penalty parameter was above a
threshold value, the algorithm could not improve on the initial design
point. Below a lower threshold value, the architecture showed poor
convergence. Finally, the authors could not find a penalty parameter
that produced a final design close to those computed by other
architectures. In light of these findings, they did not test MCO further.
Another idea, proposed by Sobieski and Kroo [75], uses surrogate
models to approximate the postoptimality behavior of the discipline
subproblems in the system subproblem. This both eliminates the
direct calculation of postoptimality derivatives and improves the
 Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

Fig. 8
Diagram for the CSSO architecture.

2059 MARTINS AND LAMBE

 2060 MARTINS AND LAMBE
Algorithm 2 Concurrent subspace optimization
Input: Initial design variables x0
Output: Optimal variables x
, objective function f
0 , and constraint values c

0: Initiate main CSSO iteration
repeat
1: Initiate design of experiments (DOEs) to generate design points
for each DOE point do
2: Initiate MDA that uses exact discipline information
repeat
3: Evaluate discipline analyses
4: Update coupling variables y
until 4 → 3: MDA has converged
5: Update discipline surrogate models with latest design
end for 6 → 2
7: Initiate independent discipline optimizations (in parallel)
for each discipline i do
repeat
8: Initiate MDA with exact coupling variables for discipline i and
approximate coupling variables for other disciplines
repeat
9: Evaluate discipline i outputs y i , and surrogate models for the
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
other disciplines, y~ j≠i
10: Update coupling variables y and y~
until 10 → 9: MDA has converged
11: Compute objective f0 and constraint functions c using current data
12: Compute new discipline design point
until 12 → 8: Discipline optimization i has converged
end for
13: Initiate DOE that uses subproblem solutions as sample points
for each subproblem solution i do
14: Initiate MDA that uses exact discipline information
repeat
15: Evaluate discipline analyses
16: Update coupling variables y
until 16 → 15 MDA has converged
17: Update discipline surrogate models with newest design
end for 18 → 14
19: Initiate system-level optimization
repeat
20: Initiate MDA that uses only surrogate model information
repeat
21: Evaluate discipline surrogate models
22: Update surrogate coupling variables y~
until 22 → 21: MDA has converged
23: Compute objective f0 and constraint function values c
24: Update system design point
until 24 → 20: System level problem has converged
until 25 → 1: CSSO has converged
treatment of the consistency constraints. While the approach does
seem to be effective for the problems they solve, to our knowledge, it
has not been adopted by any other researchers to date.
The simplest and most effective CO adjustment involves relaxing
the system subproblem equality constraints to inequalities with a
relaxation tolerance; this was originally proposed by Braun et al.
[142]. This approach was also successful in other test problems
[147,148], where the tolerance is a small fixed number, usually 10−6.
The effectiveness of this approach stems from the fact that a positive
inconsistency value causes the gradient of the constraint to be
nonzero if the constraint is active, eliminating the constraint-
qualification issue. Nonzero inconsistency is not an issue in a
practical design setting provided the inconsistency is small enough
such that other errors in the computational model dominate at the
final solution. Li et al. [149] built on this approach by adaptively
choosing the tolerance during the solution procedure so that the
system subproblem remains feasible at each iteration. This approach
appears to work for the test problems in [62] but has yet to be verified
on larger test problems.
Despite the numerical issues, CO has been widely implemented on
a number of MDO problems, mostly in the design of aerospace
systems. Examples include the design of launch vehicles [39],
rocket engines [40], satellite constellations [150], flight trajectories
[73,151], and flight control systems [152], as well as the preliminary
design of complete aircraft [17,18] and aircraft family design [153].
Beyond aerospace engineering, CO has been applied to problems
involving automobile engines [28], bridge design [22], railway cars
[26], and even the design of a scanning optical microscope [27].
Adaptations of the CO architecture have also been developed for
multi-objective, robust, and multifidelity MDO problems. Multi-
objective formulations were first described by Tappeta and Renaud
[44]. McAllister et al. [154] presented a multi-objective approach
using linear physical programming. Available robust design
formulations incorporate the decision-based models of Gu et al.
[155] and McAllister and Simpson [28], the implicit uncertainty
propagation method of Gu et al. [156], and the fuzzy computing
models of Huang et al. [157]. Zadeh and Toropov [121] integrated
multiple model fidelities into CO for an aircraft design problem.
The most recent version of CO, enhanced collaborative optimiza-
tion (ECO), was developed by Roth [158] and Roth and Kroo [56].
Figure 10 shows the XDSM corresponding to this architecture. ECO,
while still derived from the same basic problem as the original CO
architecture, is radically different, and therefore deserves attention. In
a sense, the roles of the system and discipline optimization have
been reversed in ECO: the system subproblem minimizes system
infeasibility, while the discipline subproblems minimize the system
objective. The system subproblem is
X
N
minimize J0  kx^ 0i − x0 k22  k^y i − y i x0 ; xi ; y^ j≠i k22
i1
with respect to x0 ; y^ (11)
Note that this subproblem is unconstrained. Also, unlike CO, post-
optimality derivatives are not required by the system subproblem,
because the discipline responses are treated as parameters. The
system subproblem chooses the shared design variables by averaging
all the discipline preferences.
The ith discipline subproblem is
minimize Ji  f~0 x^ 0i ; y i x^ 0i ; xi ; y^ j≠i 
wCi kx^ 0i − x0 k22  k^y i − y i x^ 0i ; xi ; y^ j≠i k22 
X
N X
ns
wFi sjk
j1;j≠i k1
with respect to x^ 0i ; xi ; sj≠i
subject to ci x^ 0i ; xi ; y i x^ 0i ; xi ; y^ j≠i  ≥ 0
c~ j x^ 0i   sj ≥ 0 j  1; : : : ; i − 1; i  1; : : : ; N
sj ≥ 0 j  1; : : : ; i − 1; i  1; : : : ; N (12)
where wCi and wFi are penalty weights for the consistency and
nonlocal design constraints, and s is a local set of elastic variables for
the constraint models. The wFi penalty weights are chosen to be
larger than the largest Lagrange multiplier, while the wCi weights
are chosen to guide the optimization toward a consistent solution.
Theoretically, each wCi must be driven to infinity to enforce
consistency exactly. However, smaller finite values are used in
practice to both provide an acceptable level of consistency and
explore infeasible regions of the design space [56].
The main new idea introduced in ECO is to include linear models
of nonlocal constraints, represented by c~ j≠i, and a quadratic model of
the system objective function, represented by f~0, in each discipline
subproblem. This is meant to increase each discipline’s “awareness”
of its influence on other disciplines and the global objective as a
whole. The construction of the constraint models deserves special
attention, because it strongly affects the structure of Fig. 10. The
constraint models for each discipline are constructed by first solving
the optimization problem that minimizes the constraint violation with
respect to the local elastic and design variables, i.e.,
 MARTINS AND LAMBE 2061

Fig. 9 Diagram for the CO architecture.

Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

X
ns
minimize sik with those from other distributed architectures [137,139,148] but still
k1
compare poorly with those from monolithic architectures.
with respect to xi ; si
E. Bilevel Integrated System Synthesis
subject to ci x0 ; xi ; y i x0 ; xi ; y^ j≠i   si ≥ 0 The bilevel integrated system synthesis (BLISS) architecture [70],
si ≥ 0 (13) like CSSO, is a method for decomposing the MDF problem along
discipline lines. Unlike CSSO, however, BLISS assigns local design
variables to discipline subproblems and shared design variables to the
where the shared design variables and the coupling variable copies system subproblem. The basic approach of the architecture is to form
are treated as fixed parameters. Postoptimality derivatives are then a path in the design space using a series of linear approximations to
computed to determine the change in the optimized local design the original design problem, with user-defined bounds on the design
variables with respect to the change in the shared design variables. variable steps, to prevent the design point from moving so far away
Combining these postoptimality derivatives with the appropriate that the approximations are too inaccurate. This is an idea similar
partial derivatives yields the linear constraint models. The optimized to that of trust-region methods [161]. These approximations are
local design variables and elastic variables from problem (13) are constructed at each iteration using coupled sensitivity information.
then used as part of the initial data for problem (12). The full The system-level subproblem is
algorithm for ECO is listed in Algorithm 4.


df0
Based on Roth and Kroo’s [56] and Roth’s results [158], ECO is minimize f
0 0  Δx0
effective in reducing the number of discipline analyses compared to dx0
CO. The tradeoff is in the additional time required to build and update with respect to Δx0
the models for each discipline weighed against the simplified solution

dc
0
to the decomposed optimization problems. The results also show subject to c
0 0  Δx0 ≥ 0
that ECO compares favorably with the analytical target cascading dx0



architecture, which we describe in Sec. IV.F. dci
While ECO seems to be effective, CO tends to be an inefficient c
i 0  Δx0 ≥ 0 for i  1; : : : ; N
dx0
architecture for MDO problems. Without the modifications discussed
in this section, the architecture requires a disproportionately large Δx0L ≤ Δx0 ≤ Δx0U (14)
number of function and discipline evaluations [138,140,159,160], The discipline i subproblem is given by
assuming it converges at all. When the system-level equality

df0
constraints are relaxed, the results from CO are more competitive minimize f0 0  Δxi
dxi
with respect to Δxi


Algorithm 3 Collaborative optimization dc0
subject to c0 0  Δxi ≥ 0
dxi
Input: Initial design variables x0

Output: Optimal variables x
, objective function f
0 , and constraint values c
dci
0: Initiate system optimization iteration ci 0  Δxi ≥ 0
dxi
repeat
1: Compute system subproblem objectives and constraints ΔxiL ≤ Δxi ≤ ΔxiU (15)
for each discipline i (in parallel), do
1.0: Initiate discipline optimization Note the extra set of constraints in both system and discipline
repeat subproblems denoting the design variable bounds.
1.1: Evaluate discipline analysis To prevent violation of the discipline constraints by changes in the
1.2: Compute discipline subproblem objective and constraints shared design variables, postoptimality derivative information (the
1.3: Compute new discipline subproblem design point and Ji change in the optimized discipline constraints with respect to a
until 1.3 → 1.1: Optimization i has converged change in the system design variables) is required to solve the system
end for
2: Compute new system subproblem design point
subproblem. For this step, Sobieszczanski–Sobieski [70] presents
until 2 → 1: System optimization has converged two methods: one based on a generalized version of the global
sensitivity equations [108], and another based on the “pricing”
 2062 MARTINS AND LAMBE
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
Fig. 10 Diagram for the ECO architecture.
interpretation of local Lagrange multipliers. The resulting variants of
BLISS are BLISS/A and BLISS/B, respectively. Other variations use
surrogate models to compute postoptimality derivatives [68,162].
Figure 11 shows the XDSM for BLISS, and the procedure is listed
in Algorithm 5. Because of the linear nature of the optimization
problems under consideration, repeated evaluation of the objective
and constraint functions is not necessary once gradient information is
available. However, this reliance on linear approximations is not
without difficulties. If the underlying problem is highly nonlinear, the
algorithm may converge slowly. The presence of user-defined
variable bounds may help the convergence if these bounds are
properly chosen, e.g., through a trust-region framework. Detailed
knowledge of the design space can also help, but this increases the
implementation overhead.
There are two adaptations of the original BLISS architecture. The
first is Ahn and Kwon’s proBLISS [163], an architecture for
reliability-based MDO. Their results show that the architecture is
competitive with reliability-based adaptations of MDF and IDF. The
second is LeGresley and Alonso’s BLISS/POD [164], an architecture
that integrates a reduced-order modeling technique called proper
Algorithm 4 Enhanced collaborative optimization (ECO)
Input: Initial design variables x0
Output: Optimal variables x
, objective function f
0 , and constraint values c

0: Initiate ECO iteration
repeat
for each discipline i, do
1: Create linear constraint models
2: Initiate discipline optimization
repeat
3: Evaluate nonlocal constraint models with local copies of shared
variables
3.0: Evaluate discipline analysis
3.1: Compute discipline subproblem objective and constraints
4: Compute new discipline subproblem design point and Ji
until 4 → 3: Discipline subproblem has converged
end for
5: Initiate system optimization
repeat
6: Compute J0
7: Compute updated values of x0 and y^
until 7 → 6: System optimization has converged
until 8 → 1: J0 is below specified tolerance
orthogonal decomposition (POD) [165] to reduce the cost of the
MDA and gradient computation steps. Their results show a signif-
icant improvement in the performance of BLISS, to the point where it
is almost competitive with MDF.
A radically different formulation called BLISS-2000 was developed
by Sobieszczanski-Sobieski et al. [54]. Because BLISS-2000 does
not require an MDA to restore the feasibility of the design, we have
separated it from the other BLISS variants in the classification shown
in Fig. 1. In fact, like other IDF-derived architectures, BLISS-2000
uses coupling variable copies to enforce consistency at the optimum.
Information exchange between the system and discipline sub-
problems occurs through surrogate models of the discipline optima.
The BLISS-2000 system subproblem is given by
minimize f0 x; yx;
~ y^ 
with respect to x0 ; y^ ; w
subject to c0 x; yx;
~ y^ ; w ≥ 0
y^ i − y~ i x0 ; xi ; y^ j≠i ; wi   0 for i  1; : : : ; N (16)
The BLISS-2000 discipline i subproblem is
minimize fi wTi y i x0 ; xi ; y^ j≠i 
with respect to xi
subject to ci x0 ; xi ; y i x0 ; xi ; y^ j≠i  ≥ 0 (17)
A unique aspect of this architecture is the use of a vector of weighting
coefficients wi attached to the discipline states. These weighting
coefficients give the user a measure of control over the state variable
preferences. Generally speaking, the coefficients should be chosen
based on the structure of the global objective to allow the discipline
subproblems to find an optimum more quickly. The impact of these
coefficients on convergence has yet to be determined. The XDSM for
the architecture is shown in Fig. 12, and the procedure is listed in
Algorithm 6.
BLISS-2000 has several advantages over the original BLISS
architecture. First, the solution procedure is much easier to
understand, and this simplicity greatly reduces the obstacles to
implementation. Second, the decomposed problem formulation is
equivalent to problem (1) provided the weighting coefficients
 MARTINS AND LAMBE
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
Fig. 11 Diagram for the BLISS architecture.
accurately capture the influence of the coupling variables on the
objective [54]. Third, by using surrogate models for each discipline,
rather than for the whole system, the calculations for BLISS-2000 can
be run in parallel with minimal communication between disciplines.
Finally, BLISS-2000 seems to be more flexible than its predecessor.
Recently, Sobieszczanski-Sobieski detailed an extension of BLISS-
2000 to handle multilevel system-of-systems problems [166]. In spite
of these advantages, it appears that BLISS-2000 has not been used
nearly as frequently as the original BLISS formulation.
Most of the benchmarking of BLISS used its original formulation.
Many of the results available [137,138,140] suggest that BLISS is not
competitive with other architectures in terms of computational cost.
Concerns include both the number of discipline analyses required
and the high cost of the postoptimality derivative computations.
These problems can be mitigated by the use of surrogate models
or reduced-order modeling, as described previously. The one
benchmarking result available for BLISS-2000 is the launch-vehicle
Algorithm 5 Bilevel system integrated system synthesis (BLISS)
Input: Initial design variables x0
Output: Optimal variables x
, objective function f
0 , and constraint values c

0: Initiate system optimization
repeat
1: Initiate MDA
repeat
2: Evaluate discipline analyses
3: Update coupling variables
until 3 → 2: MDA has converged
4: Initiate parallel discipline optimizations
for each discipline i, do
5: Evaluate discipline analysis
6: Compute objective and constraint function values and derivatives
with respect to local design variables
7: Compute optimal solution for discipline subproblem
end for
8: Initiate system optimization
9: Compute objective and constraint function values and derivatives
with respect to shared design variables using postoptimality analysis
10: Compute optimal solution to system subproblem
until 11 → 1: System optimization has converged
2063
design problem of Brown and Olds [146]. In this case, BLISS-2000
tended to outperform other distributed architectures, and it should be
cost-competitive with other monolithic architectures when coarse-
grained parallel processing is fully exploited. While this is a
promising development, more work is needed to confirm the result.
F. Analytical Target Cascading
The analytical target cascading (ATC) architecture was not initially
developed as an MDO architecture but as a method to propagate
system targets (i.e., requirements or desirable properties) through a
hierarchical system to achieve a feasible system design satisfying
these targets [58,167]. If the targets are unattainable, the ATC
architecture returns a design point that minimizes the unattainability.
Essentially, then, the ATC architecture is no different from an MDO
architecture with a system objective of minimizing the squared
difference between a set of system targets and the model responses.
By simply changing the objective function, we can solve general
MDO problems using ATC.
The ATC problem formulation that we present here is due to
Tosserams et al. [168]. This formulation conforms to our definition of
an MDO problem by explicitly including systemwide objective and
constraint functions. Like all the other ATC problem formulations, a
solution generated by this architecture will solve problem (1)
provided the penalty terms in the optimization problems all approach
zero. The ATC system subproblem is given by
X
N
minimize f0 x; y^   Φi x^ 0i − x0 ; y^ i − y i x0 ; xi ; y^ 
i1
Φ0 c0 x; y^ 
with respect to x0 ; y^ (18)
where Φ0 is a penalty relaxation of the global design constraints, and
Φi is a penalty relaxation of the discipline i consistency constraints.
The variable copies x^ 0i have the same function as in the CO
architecture, with the penalty function used to ensure all the copies
converge on the same value in an optimal design. The ith discipline
subproblem is
 2064 MARTINS AND LAMBE
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

Fig. 12 Diagram for the BLISS-2000 architecture.

minimize f0 x^ 0i ; xi ; y i x^ 0i ; xi ; y^ j≠i ; y^ j≠i  the penalty functions are applied to all constraints that combine local
information with information from the levels immediately above or
 fi x^ 0i ; xi ; y i x^ 0i ; xi ; y^ j≠i  below the current one. Postoptimality derivatives are not needed in
Φi ^y i − y i x^ 0i ; xi ; y^ j≠i ; x^ 0i − x0  any of the subproblems since nonlocal data are always treated as fixed
values in the current subproblem. Furthermore, as with penalty
Φ0 c0 x^ 0i ; xi ; y i x^ 0i ; xi ; y^ j≠i ; y^ j≠i  methods for general optimization problems [see, e.g., Nocedal and
with respect to x^ 0i ; xi Wright ([43] Chap. 17)], the solution of the MDO problem must be
computed to reasonable accuracy before the penalty weights are
subject to ci x^ 0i ; xi ; y i x^ 0i ; xi ; y^ j≠i  ≥ 0 (19) updated. However, because we are now dealing with a distributed set
of subproblems, the whole hierarchy of subproblems must be solved
for a given set of weights. This is because of the nonseparable nature
Figure 13 shows the ATC architecture XDSM, where w denotes the of the quadratic penalty function.
penalty function weights used in the determination of Φ0 and Φi . The The most common penalty functions in ATC are quadratic penalty
details of ATC are described in Algorithm 7. functions. In this case, the proper selection of the penalty weights is
ATC can be applied to a multilevel hierarchy of systems as well as important for both final consistency in the discipline models and
to a discipline-based nonhierarchical system. In the multilevel case, convergence of the algorithm. Michalek and Papalambros [60]
present an effective weight-update method that is especially useful
when unattainable targets have been set in a traditional ATC process.
Algorithm 6 BLISS-2000 Michelena et al. [59] present several coordination algorithms
using ATC with quadratic penalty functions and demonstrate the
Input: Initial design variables x0 convergence to an optimum for all of them. However, the analysis of
Output: Optimal variables x
, objective function f
0 , and constraint values c

0: Initiate main BLISS iteration
Michelena et al. is based on a top-level target satisfaction objective
repeat function that is convex. The behavior of these versions of ATC in the
for each discipline i, do presence of nonconvex objectives is unknown.
1: Initiate DOE Several other penalty function choices and associated coordination
for each DOE point, do approaches have been devised for ATC. Kim et al. [169] outline a
2: Initiate discipline optimization version of ATC that uses Lagrangian relaxation and a subgradient
repeat method to update the multiplier estimates. Tosserams et al. [170]
3: Evaluate discipline analysis use augmented Lagrangian relaxation with Bertsekas’ method of
4: Compute discipline objective fi and constraint functions ci multipliers [171] and the alternating-direction method of multipliers
5: Update local design variables xi
until 5 → 3 : Discipline optimization subproblem has converged
[120] to update the penalty weights. They also group this variant
6: Update surrogate model of optimized discipline subproblem with of ATC into a larger class of coordination algorithms known as
new solution augmented Lagrangian coordination [168]. Li et al. [172] apply the
end for 7 → 2 diagonal quadratic approximation approach of Ruszczynski [173]
end for to the augmented Lagrangian to eliminate subproblem coupling
8: Initiate system optimization through the quadratic terms and further parallelize the architecture.
repeat Finally, Han and Papalambros [174] propose a version of ATC based
9: Evaluate surrogate models with current values of system variables on sequential linear programming [175,176], where inconsistency is
10: Compute system objective and constraint function values penalized using infinity norms. They later present a convergence
11: Compute new system design point
until 11 → 9 System subproblem has converged
proof of this approach in a short note [177]. For each of the preceding
until 12 → 1: System optimization has converged penalty function choices, ATC was able to produce the same design
solutions as the monolithic architectures.
 MARTINS AND LAMBE
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
Fig. 13 Diagram for the ATC architecture.
Despite having been developed relatively recently, the ATC
architecture has been widely used. It has been most frequently
applied to design problems in the field for which it was developed: the
automotive industry [29,178–184]. However, it has also proved
useful in aircraft design [55,153,185] and building design [23].
Applications beyond engineering design include manufacturing
decisions [186], supply chain management [187], and marketing
decisions in product design [188]. Huang et al. [189] have developed
an ATC-specific web portal to solve optimization problems via
the ATC architecture. Etman et al. [190] discuss the automatic
implementation of coordination procedures, using ATC as an
example architecture. Formulations exist that can handle integer
variables [48] and probabilistic design problems [191,192]. The
method has also been adapted for problems with block-separable
linking constraints [193]. In this class of problems, c0 consists of
constraints that are sums of functions depending on only shared
variables and the local variables of one discipline.
The performance of ATC compared with other architectures is not
well known; only one result is available. In de Wit and van Keulen’s
architecture comparison [140], ATC is competitive with all the other
benchmarked distributed architectures, including standard versions
Algorithm 7 Analytical target cascading
Input: Initial design variables x0
Output: Optimal variables x
, objective function f
0 , and constraint values c

0: Initiate main ATC iteration
repeat
for each discipline i, do
1: Initiate discipline optimizer
repeat
2: Evaluate discipline analysis
3: Compute discipline objective and constraint functions and
penalty function values
4: Update discipline design variables
until 4 → 2: Discipline optimization has converged
end for
5: Initiate system optimizer
repeat
6: Compute system objective, constraints, and all penalty functions
7: Update system design variables and coupling variable copies
until 7 → 6: System optimization has converged
8: Update penalty weights
until 8 → 1: System optimization has converged
2065
of CSSO, CO, and BLISS. However, with a gradient-based optimizer,
ATC and the other distributed architectures are not competitive
with a monolithic architecture in terms of the number of function
and discipline evaluations. Different versions of ATC have been
benchmarked against each other. Tosserams et al. [170] compared the
augmented Lagrangian penalty approach with the quadratic penalty
approach and found much improved results with the alternating-
direction method of multipliers. Surprisingly, de Wit and van Keulen
[140] found that the augmented Lagrangian version performed worse
than the quadratic penalty method for their test problem. Han and
Papalambros [174] compared their sequential linear programming
version of ATC to several other approaches and found a significant
reduction in the number of function evaluations. However, they noted
that the coordination overhead is large compared to other ATC
versions, and this needs to be addressed.
G. Exact and Inexact Penalty Decompositions
If there are no systemwide constraints or objectives, i.e., if neither
f0 and c0 exist, the exact or inexact penalty decompositions (EPDs or
IPDs, respectively) [72,194] may be employed. Both formulations
rely on solving the discipline subproblem
minimize fi x^ 0i ; xi ; y i x^ 0i ; xi ; y^ j≠i 
 Φi x^ 0i − x0 ; y^ i − y i x^ 0i ; xi ; y^ j≠i 
with respect to x^ 0i ; xi
subject to ci x^ 0i ; xi ; y i x0i ; xi ; y^ j≠i  ≥ 0 (20)
Here, Φi denotes the penalty function associated with the incon-
sistency between the ith discipline’s information and the system
information. In EPD, Φi is an L1 penalty function with additional
variables and constraints added to ensure smoothness. In IPD, Φi is a
quadratic penalty function with appropriate penalty weights. The
notation x^ 0i denotes a local copy of the shared design variables in
discipline i, while x0 denotes the original variable. As in the ATC
architecture, the penalty function is used to ensure that all variable
copies converge to a single value in an optimal design.
At the system level, the subproblem is an unconstrained
minimization with respect to the shared variables and coupling
variable copies. The objective function is the sum of the optimized
discipline penalty terms, denoted as Φ
i :
 2066
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
Fig. 14 Diagram for the EPD and IPD architectures.
X
N   XN
minimize Φ
i x0 ; y^  Φi x^ 0i − x0 ; y^ i − y i x^ 0i ;xi ; y^ j≠i 
i1 i1
withrespectto x0 ; y^
The penalty weights are updated upon solution of the system
problem. Figure 14 shows the XDSM for this architecture, where w
represents the penalty weights. The sequence of operations in this
architecture is detailed in Algorithm 8.
Both EPD and IPD have mathematically provable convergence to a
KKT point under the linear independence constraint qualification,
and with mild assumptions on the update strategy for the penalty
weights [72]. In particular, similar to other quadratic penalty
methods, the penalty weight in IPD must monotonically increase
until the inconsistency is sufficiently small [43]. For EPD, the penalty
weight must be larger than the largest Lagrange multiplier, following
the established theory for the L1 penalty function [43], while the
barrier parameter must monotonically decrease as in an interior point
method [107]. If other penalty functions are employed, the parameter
values are selected and updated according to the corresponding
mathematical theory. Furthermore, under these conditions, the
solution obtained under EPD and IPD will also be a solution to
problem (1).
Only once in the literature has either architecture been tested
against others. The results of Tosserams et al. [195] suggest that
the performance depends on the penalty function employed. A
comparison of IPD with a quadratic penalty function and IPD with an
augmented Lagrangian penalty function showed that the latter
significantly outperformed the former on several test problems in
terms of both time and number of function evaluations.
H. MDO of Independent Subspaces
If the problem contains no systemwide constraints or objectives
(i.e., if f0 and c0 do not exist) and the problem does not include
shared design variables (i.e., if x0 does not exist), then the MDO of
independent subspaces (MDOISs) architecture [71] applies. In this
case, the discipline subproblems are fully separable (aside from the
coupled state variables) and given by
minimize fi xi ; y i xi ; y^ j≠i 
with respect to xi
subject to ci xi ; y i xi ; y^ j≠i  ≥ 0
The coupling variable copies are just parameters representing
the system state information. Upon solution of the discipline
MARTINS AND LAMBE
subproblems, which can access the output of individual discipline-
analysis codes, a full MDA is completed to update all parameter
values. Thus, rather than the system subproblem used by other
(21) architectures, the MDA is used to guide the discipline subproblems to
a design solution. Shin and Park [71] show that, under the given
problem assumptions, an optimal design of the original (MDF)
problem is found using this architecture. Figure 15 depicts this
process in an XDSM. MDOIS is detailed in Algorithm 9.
Benchmarking results are available that compare MDOIS to
some of the older architectures. These results are given by Yi et al.
[138]. In many cases, MDOIS requires fewer analysis calls than
MDF while still being able to reach the optimal solution. However,
MDOIS does not converge as fast as IDF, and the restrictive defini-
tion of the problem means that the architecture is not nearly as
flexible as MDF. A practical problem that can be solved using
MDOIS is the belt-integrated seat problem of Shin et al. [196].
However, the results have not been compared to those obtained using
other architectures.
I. Quasi-Separable Decomposition
Haftka and Watson [66] developed the quasi-separable decom-
position (QSD) architecture to solve quasi-separable optimization
problems. In a quasi-separable problem, the system objective and
Algorithm 8 Exact and inexact penalty decompositions
Input: Initial design variables x0
Output: Optimal variables x
, objective functions f
i , and constraint values c

0: Initiate main iteration
repeat
for each discipline i, do
repeat
1: Initiate discipline optimizer
2: Evaluate discipline analysis
3: Compute discipline objective and constraint functions, and
penalty function values
4: Update discipline design variables
until 4 → 2: Discipline optimization has converged
end for
5: Initiate system optimizer
repeat
6: Compute all penalty functions
(22) 7: Update system design variables and coupling variable copies
until 7 → 6: System optimization has converged
8: Update penalty weights
until 8 → 1: Penalty weights are large enough
 MARTINS AND LAMBE 2067

Fig. 15 Diagram for the MDOIS architecture.

Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

constraint functions are assumed to be dependent only on the shared minimize − si

design variables and coupling variables. This type of problem is
equivalent to the complicating-variable problems discussed in with respect to xi ; si
Sec. IV.A. We have not come across any practical design problems subject to c
ik  cik x0 ; xi ; y i x0 ; xi ; y^ j≠i  − si ≥ 0
that do not satisfy this property. However, if necessary, we can easily
transform the general MDO problem [Eq. (1)] into a quasi-separable for k  1; : : : ; mi
problem. This is accomplished by copying the relevant local f
i  bi − fi x0 ; xi ; y i x0 ; xi ; y^ j≠i  − si ≥ 0
variables and forcing the global objective to depend on the copies.
(The process is analogous to adapting the general MDO problem to cc  y^ i − y i x0 ; xi ; y^ j≠i   0 (24)
the original CO architecture.) By using a local equality constraint to
ensure that the copied variables converge to the same value, we where k is an element of the constraint vector ci. Because of the use of
ensure that a solution to the quasi-separable problem is also a solution coupling variable copies, we classify this architecture as distributed
to the original (IDF) problem. The system subproblem of QSD is IDF. Similar to CO, this is a bilevel architecture where the solutions of
given by the discipline subproblems are constraints in the system subproblem.
Therefore, postoptimality derivatives or surrogate model approx-
X
N
imations of the optimized discipline subproblems are required to
minimize f0 x0 ; y^   bi solve the system subproblem. The standard architecture is shown as
i1
an XDSM in Fig. 16. The sequence of operations in QSD is given by
with respect to x0 ; y^ ; b Algorithm 10. Haftka and Watson have extended the theory behind
subject to c0 x0 ; y^  ≥ 0 QSD to problems with a combination of discrete and continuous
variables [47]. Liu et al. [197] successfully applied QSD with
s
i x0 ; xi ; y i x0 ; xi ; y^ j≠i ; bi  ≥ 0 surrogate models to a structural optimization problem. However, they
for i  1; : : : ; N (23) made no comparisons to other architectures, not even QSD without
surrogates. A version of QSD without surrogate models was
benchmarked by de Wit and van Keulen [140]. Unfortunately, this
where si is the constraint margin for discipline i, and bi is the budget
architecture was the worst of all the architectures tested in terms of
assigned to each discipline objective. The discipline i subproblem
becomes discipline evaluations. Aversion using surrogate models should yield
improved performance, because of the smoothness introduced by
the model, but this version has not, to our knowledge, been
benchmarked.
Algorithm 9 Multidisciplinary design optimization of independent
subspaces J. Asymmetric Subspace Optimization
Input: Initial design variables x0 The asymmetric subspace optimization (ASO) architecture [123]
Output: Optimal variables x
, objective functions f
i , and constraint values c
is a new distributed-MDF architecture. It was motivated by high-
0: Initiate main iteration fidelity aerostructural optimization, where the aerodynamic analysis
repeat typically requires an order of magnitude more time to complete than
1: Initiate MDA the structural analysis [198]. To reduce the number of expensive
repeat aerodynamic analyses, the structural analysis is coupled with a
2: Evaluate discipline analyses structural optimization inside the MDA. This idea can be readily
3: Update coupling variables
until 3 → 2: MDA has converged generalized to any problem where there is a wide discrepancy
for each discipline i, do between the discipline-analysis times. Figure 17 shows the
4: Initiate discipline optimization optimization of the third discipline of a generic problem within the
repeat ASO architecture. The sequence of operations in ASO is listed in
5: Evaluate discipline analysis Algorithm 11.
6: Compute discipline objectives and constraints The optimality of the final solution is preserved by using the
7: Compute new discipline design point coupled postoptimality sensitivity (CPOS) equations, developed by
until 7 → 5: Discipline optimization has converged
end for
Chittick and Martins [123], to calculate gradients at the system level.
until 8 → 1 Main iteration has converged CPOS represents an extension of the coupled sensitivity equations
[99,108] to include the optimality conditions. ASO was later
 2068
Fig. 16 Diagram for the QSD architecture.
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
implemented using a coupled-adjoint approach as well [76].
Kennedy et al. [199] presented alternative strategies for computing
the discipline subproblem optima and the postoptimality sensitivity
analysis. As with other bilevel MDO architectures, the postoptimality
analysis is necessary to ensure convergence to an optimal design of
the original monolithic problem. We also note the work of Kaufman
et al. [200] and Hosder et al. [201], who used structural optimization
as an inner loop in aircraft design optimization with a surrogate model
replacing the postoptimality sensitivity equations.
The system subproblem is
minimize f0 x; yx; y
with respect to x0 ; xk
subject to c0 x; yx; y ≥ 0
ck x0 ; xk ; y k x0 ; xk ; y j≠k  ≥ 0 for all k
where subscript k denotes discipline information that is not optimized
within the MDA. The discipline subproblem for discipline i, which is
resolved inside the MDA, is
minimize f0 x; yx; y
with respect to xi
subject to ci x0 ; xi ; y i x0 ; xi ; y j≠i  ≥ 0
The results show a substantial reduction in the number of calls to the
aerodynamics analysis, and even a slight reduction in the number
of calls to the structural analysis [123]. However, the total time
for the optimization routine is competitive with MDF only if the
Algorithm 10 Quasi-separable decomposition
Input Initial design variables x 0
Output: Optimal variables x
, objective function f
0 , and constraint values c

0: Initiate system optimization
repeat
1: Compute system objectives and constraints
for each discipline i, do
1.0: Initiate discipline optimization
repeat
1.1: Evaluate discipline analysis
1.2: Compute discipline objective and constraints
1.3: Update discipline design point
until 1.3 → 1.1: Discipline optimization has converged
end for
2: Compute new system design point
until 2 → 1: System problem has converged
MARTINS AND LAMBE
aerodynamic analysis is substantially more costly than the structural
analysis. If the two analyses take roughly the same time, MDF is still
much faster. Furthermore, using CPOS increases the complexity of
the sensitivity analysis compared to using a normal coupled adjoint,
which adds to the overall solution time. As a result, this architecture
may appeal only to practitioners solving MDO problems with widely
varying computational costs in the discipline analysis.
V. Architecture Benchmarking Issues
One of the challenges facing the MDO community is determining
which architecture is most efficient for a given MDO problem or class
of problems. One way is to benchmark the architectures using simple
test problems. Many authors developing new architectures include a
few results in their work. However, it is especially important to test
multiple architectures on a given problem, since the most appropriate
(25) architecture often depends on the nature of the problem. There
are a number of benchmarking studies in the literature [137–
139,146,159,160,202,203]. However, they have limitations that need
to be considered when reading these studies and considering
future work.
The first is the fact that no two studies are likely to produce
identical results. This is because no two programmers will program
the same architecture in the same way. In addition, experts in a
particular architecture may be able to more fully optimize the
(26) performance of that architecture, unintentionally biasing the results
[138]. This problem can be overcome by performing the bench-
marking in a specialized MDO framework such as iSight,
ModelCenter, pyMDO [204], or OpenMDAO [205]. These frame-
works allow single definitions of the MDO problem and the
architecture to be frequently reused, eliminating most of the “human
factor” from the optimization. OpenMDAO and pyMDO can even
implement the MDO architectures automatically [204]. A framework
has also been proposed to go a step further and select architectures
automatically based on the structure of the problem and nest different
architectures in the solution of large MDO problems [206]. Padula
and Gillian [207] present a brief history of frameworks for MDO.
Kodiyalam and Sobieszczanski–Sobieski [208] and Kodiyalam et al.
[209] discuss requirements for an MDO framework and high-
performance computing considerations, respectively.
The second limitation is the choice of architectures for the
benchmarking studies. Most studies tend to focus on the most mature
architectures, especially MDF, IDF, CO, and CSSO. This should
become less of a limitation as the newer architectures become better
known. However, we must emphasize that implementing a new
architecture in an established MDO framework should allow more
rapid benchmarking and give a much earlier indication of the
architecture’s promise.
The third limitation is in the test problems themselves. Almost all
the test problems are of low dimensionality, and many have discipline
 MARTINS AND LAMBE
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
Fig. 17 Diagram for the ASO architecture.
analyses consisting of analytic functions. In high-fidelity design, the
MDO problem could have thousands of variables and constraints
with discipline analyses that take minutes or hours to evaluate, even
with high-performance parallel computers. While the test problems
used in benchmarking may never reach this level of complexity, they
should be large enough to establish a clear trend in terms of time and
number of function calls. Ideally, these problems should be scalable
to examine the effects of problem dimensionality on architecture
efficiency. An early attempt at such a scalable problem was presented
by Tedford and Martins [139] and allowed the user to define an
arbitrary number of disciplines, design variables, and coupling
variables. Another benchmarking problem that possesses some
scalability is the microaccelerometer problem of Tosserams et al.
[210]. Test problems with lower dimensionality but with multiple
local minima are presented by Tosserams et al. [211]. We also feel
that a test suite of MDO problems similar to the now-defunct NASA
MDO suite [212] would be a useful resource for researchers and
practitioners.
Algorithm 11 Asymmetric subspace optimization
Input: Initial design variables x0
Output: Optimal variables x
, objective function f
0 , and constraint values c

0: Initiate system optimization
repeat
1: Initiate MDA
repeat
2: Evaluate analysis 1
3: Evaluate analysis 2
4: Initiate optimization of discipline 3
repeat
5: Evaluate analysis 3
6: Compute discipline 3 objectives and constraints
7: Update local design variables
until 7 → 5: Discipline 3 optimization has converged
8: Update coupling variables
until 8 → 2: MDA has converged
9: Compute remaining objectives and constraints
10: Update remaining design variables
until 10 → 1: System optimization has converged
2069
VI. Conclusions
In this paper, all known multidisciplinary design optimization
(MDO) architectures were summarized and classified. The archi-
tectures were presented in a unified notation to enable comparisons
between them, thus contributing to a deeper and more encompassing
understanding of the theory. A classification of the architectures was
also introduced that built on previous work to show how some
architectures may be derived from others. The hope is that this work
will help point the way toward new MDO architectures and expand
the scope of MDO theory and applications.
From a more practical perspective, much work remains to be done
in the area of benchmarking the existing architectures. The per-
formance of the new architectures needs to be compared with that of
the older architectures on a predefined set of test problems, and the
results should be independently corroborated. The development of
MDO frameworks and test problem suites would be extremely useful.
Finally, new architectures are needed. A distributed architecture
that exhibits fast convergence in most medium- and large-scale
problems may be thought of as the “holy grail” of MDO. The newest
architectures in this survey represent significant progress toward this
objective. However, the distributed architectures still seem to be more
expensive than the monolithic architectures they are meant to
supersede. Until a distributed architecture can demonstrate fast
convergence for a wide range of problems, architecture development
will remain an active area of research.
Acknowledgments
This work was partially funded by a scholarship from the Natural
Science and Engineering Research Council of Canada. We would
like to thank the following colleagues for providing feedback on a
preprint of this paper: Evin Cramer, Pascal Etman, Raphael Haftka,
John Hwang, Harrison Kim, Brian Roth, and Jaroslaw Sobieski.
Additional helpful suggestions were provided by James Allison,
Sylvain Arreckx, Cristina Bloebaum, Robert Canfield, Justin Gray,
Jason Hicken, Graeme Kennedy, Alicia Kim, Michael Kokkolaras,
Zhoujie (Peter) Lyu, and Kenneth Moore. We are also grateful to the
anonymous reviewers for their diligence and insights.
 2070 MARTINS AND LAMBE
References
[1] Schmit, L. A., “Structural Design by Systematic Synthesis,” 2nd
Conference on Electronic Computation, American Society of Civil
Engineers, New York, 1960, pp. 105–132.
[2] Schmit, L. A., and Thornton, W. A., “Synthesis of an Airfoil at
Supersonic Mach Number,” NASA CR-144, Jan. 1965.
[3] Schmit, L. A., “Structural Synthesis: Its Genesis and Development,”
AIAA Journal, Vol. 19, No. 10, 1981, pp. 1249–1263.
doi:10.2514/3.7859
[4] Schmit, L. A., Jr., “Structural Synthesis: Precursor and Catalyst, Recent
Experiences in Multidisciplinary Analysis and Optimization,” NASA
CP-2337, 1984.
[5] Haftka, R. T., “Automated Procedure for Design of Wing Structures to
Satisfy Strength and Flutter Requirements,” NASA Langley Research
Center TN-D-7264, Hampton, VA, 1973.
[6] Haftka, R. T., Starnes, J. H., Jr., Barton, F. W., and Dixon, S. C.,
“Comparison of Two Types of Optimization Procedures for Flutter
Requirements,” AIAA Journal, Vol. 13, No. 10, 1975, pp. 1333–1339.
doi:10.2514/3.60545
[7] Haftka, R. T., “Optimization of Flexible Wing Structures Subject to
Strength and Induced Drag Constraints,” AIAA Journal, Vol. 14, No. 8,
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
1977, pp. 1106–1977.
doi:10.2514/3.7400
[8] Haftka, R. T., and Shore, C. P., “Approximate Methods for Combined
Thermal/Structural Design,” NASA TP-1428, June 1979.
[9] Ashley, H., “On Making Things the Best: Aeronautical Uses of
Optimization,” Journal of Aircraft, Vol. 19, No. 1, 1982, pp. 5–28.
doi:10.2514/3.57350
[10] Green, J. A., “Aeroelastic Tailoring of Aft-Swept High-Aspect-Ratio
Composite Wings,” Journal of Aircraft, Vol. 24, No. 11, 1987, pp. 812–
819.
doi:10.2514/3.45525
[11] Grossman, B., Gurdal, Z., Strauch, G. J., Eppard, W. M., and Haftka, R.
T., “Integrated Aerodynamic/Structural Design of a Sailplane Wing,”
Journal of Aircraft, Vol. 25, No. 9, 1988, pp. 855–860.
doi:10.2514/3.45670
[12] Grossman, B., Haftka, R. T., Kao, P.-J., Polen, D. M., and Rais-Rohani,
M., “Integrated Aerodynamic-Structural Design of a Transport Wing,”
Journal of Aircraft, Vol. 27, No. 12, 1990, pp. 1050–1056.
doi:10.2514/3.45980
[13] Livne, E., Schmit, L., and Friedmann, P., “Towards Integrated
Multidisciplinary Synthesis of Actively Controlled Fiber Composite
Wings,” Journal of Aircraft, Vol. 27, No. 12, 1990, pp. 979–992.
doi:10.2514/3.45972
[14] Livne, E., “Integrated Aeroservoelastic Optimization: Status and
Direction,” Journal of Aircraft, Vol. 36, No. 1, 1999, pp. 122–145.
doi:10.2514/2.2419
[15] Jansen, P., Perez, R. E., and Martins, J. R. R. A., “Aerostructural
Optimization of Nonplanar Lifting Surfaces,” Journal of Aircraft,
Vol. 47, No. 5, 2010, pp. 1491–1503.
doi:10.2514/1.44727
[16] Ning, S. A., and Kroo, I., “Multidisciplinary Considerations in the
Design of Wings and Wing Tip Devices,” Journal of Aircraft, Vol. 47,
No. 2, 2010, pp. 534–543.
doi:10.2514/1.41833
[17] Kroo, I. M., Altus, S., Braun, R. D., Gage, P. J., and Sobieski, I. P.,
“Multidisciplinary Optimization Methods for Aircraft Preliminary
Design,” Proceedings of the AIAA 5th Symposium on Multidisciplinary
Analysis and Optimization, AIAA Paper 1994-4325, Panama City
Beach, FL, 1994.
[18] Manning, V. M., “Large-Scale Design of Supersonic Aircraft via
Collaborative Optimization,” Ph.D. Thesis, Stanford Univ., Stanford,
CA, 1999.
[19] Antoine, N. E., and Kroo, I. M., “Framework for Aircraft Conceptual
Design and Environmental Performance Studies,” AIAA Journal,
Vol. 43, No. 10, 2005, pp. 2100–2109.
doi:10.2514/1.13017
[20] Henderson, R. P., Martins, J. R. R. A., and Perez, R. E., “Aircraft
Conceptual Design for Optimal Environmental Performance,” The
Aeronautical Journal, Vol. 116, No. 1175, Jan. 2012, pp. 1–22.
[21] Alonso, J. J., and Colonno, M. R., “Multidisciplinary Optimization
with Applications to Sonic-Boom Minimization,” Annual Review of
Fluid Mechanics, Vol. 44, No. 1, 2012, pp. 505–526.
doi:10.1146/annurev-fluid-120710-101133
[22] Balling, R., and Rawlings, M. R., “Collaborative Optimization with
Disciplinary Conceptual Design,” Structural and Multidisciplinary
Optimization, Vol. 20, No. 3, 2000, pp. 232–241.
doi:10.1007/s001580050151
[23] Choudhary, R., Malkawi, A., and Papalambros, P. Y., “Analytic Target
Cascading in Simulation- Based Building Design,” Automation in
Construction, Vol. 14, No. 4, 2005, pp. 551–568.
doi:10.1016/j.autcon.2004.11.004
[24] Geyer, P., “Component-Oriented Decomposition for Multidisciplinary
Design Optimization in Building Design,” Advanced Engineering
Informatics, Vol. 23, No. 1, 2009, pp. 12–31.
doi:10.1016/j.aei.2008.06.008
[25] He, Y., and McPhee, J., “Multidisciplinary Optimization of Multibody
Systems with Application to the Design of Rail Vehicles,” Multibody
System Dynamics, Vol. 14, No. 2, 2005, pp. 111–135.
doi:10.1007/s11044-005-4310-0
[26] Enblom, R., “Two-Level Numerical Optimization of Ride Comfort in
Railway Vehicles,” Journal of Rail and Rapid Transit, Vol. 220, No. 1,
2006, pp. 1–11.
doi:10.1243/095440905X33279
[27] Potsaid, B., Bellouard, Y., and Wen, J. T.-Y., “A Multidisciplinary
Design and Optimization Methodology for the Adaptive Scanning
Optical Microscope (ASOM),” Proceedings of the SPIE, Vol. 6289,
2006, pp. 62890L1–62890L12.
doi:10.1117/12.680450
[28] McAllister, C. D., and Simpson, T. W., “Multidisciplinary Robust
Design Optimization of an Internal Combustion Engine,” Journal of
Mechanical Design, Vol. 125, No. 1, 2003, pp. 124–130.
doi:10.1115/1.1543978
[29] Kokkolaras, M., Louca, L. S., Delagrammatikas, G. J., Michelena, N.
F., Filipi, Z. S., Papalambros, P. Y., Stein, J. L., and Assanis, D. N.,
“Simulation-Based Optimal Design of Heavy Trucks by Model- Based
Decomposition: An Extensive Analytical Target Cascading Case
Study,” International Journal of Heavy Vehicle Systems, Vol. 11, No. 3/
4, 2004, pp. 403–433.
doi:10.1504/IJHVS.2004.005456
[30] Peri, D., and Campana, E. F., “Multidisciplinary Design Optimization
of a Naval Surface Combatant,” Journal of Ship Research, Vol. 47,
No. 1, 2003, pp. 1–12.
[31] Kalavalapally, R., Penmetsa, R., and Grandhi, R., “Multidisciplinary
Optimization of a Lightweight Torpedo Structure Subjected to an
Underwater Explosion,” Finite Elements in Analysis and Design,
Vol. 43, No. 2, 2006, pp. 103–111.
doi:10.1016/j.finel.2006.07.005
[32] Takekoshi, Y., Kawamura, T., Yamaguchi, H., Maeda, M., Ishii, N.,
Kimura, K., Taketani, T., and Fujii, A., “Study on the Design of
Propeller Blade Sections Using the Optimization Algorithm,” Journal
of Marine Science and Technology, Vol. 10, No. 2, 2005, pp. 70–81.
doi:10.1007/s00773-005-0197-y
[33] Young, Y., Baker, J., and Motley, M., “Reliability-Based Design and
Optimization of Adaptive Marine Structures,” Composite Structures,
Vol. 92, No. 2, 2010, pp. 244–253.
doi:10.1016/j.compstruct.2009.07.024
[34] Ganguli, R., “Survey of Recent Developments in Rotorcraft
Design Optimization,” Journal of Aircraft, Vol. 41, No. 3, 2004,
pp. 493–510.
doi:10.2514/1.58
[35] Glaz, B., Friedmann, P. P., and Liu, L., “Helicopter Vibration
Reduction Throughout the Entire Flight Envelope Using Surrogate-
Based Optimization,” Journal of the American Helicopter Society,
Vol. 54, No. 1, 2009, pp. 12007-1–12007-15.
doi:10.4050/JAHS.54.012007
[36] Fuglsang, P., and Madsen, H., “Optimization Method for Wind Turbine
Rotors,” Journal of Wind Engineering and Industrial Aerodynamics,
Vol. 80, Nos. 1–2, 1999, pp. 191–206.
doi:10.1016/S0167-6105(98)00191-3
[37] Fuglsang, P., Bak, C., Schepers, J. G., Bulder, B., Cockerill, T. T.,
Claiden, P., Olesen, A., and van Rossen, R., “Site-Specific Design
Optimization of Wind Turbines,” Wind Energy, Vol. 5, No. 4, 2002,
pp. 261–279.
doi:10.1002/we.61
[38] Kenway, G., and Martins, J. R. R. A., “Aerostructural Shape
Optimization of Wind Turbine Blades Considering Site-Specific
Winds,” 12th AIAA/ISSMO Multidisciplinary Analysis and Optimiza-
tion Conference, Sept. 2008.
doi:10.2514/6.2008-6025
[39] Braun, R. D., Moore, A. A., and Kroo, I. M., “Collaborative Approach
to Launch Vehicle Design,” Journal of Spacecraft and Rockets, Vol. 34,
No. 4, 1997, pp. 478–486.
doi:10.2514/2.3237
[40] Cai, G., Fang, J., Zheng, Y., Tong, X., Chen, J., and Wang, J.,
“Optimization of System Parameters for Liquid Rocket Engines with
Gas-Generator Cycles,” Journal of Propulsion and Power, Vol. 26,
 MARTINS AND LAMBE
No. 1, 2010, pp. 113–119.
doi:10.2514/1.40649
[41] Hwang, J. T., Lee, D. Y., Cutler, J., and Martins, J. R. R. A., “Large-
Scale MDO of a Small Satellite Using a Novel Framework for the
Solution of Coupled Systems and Their Derivatives,” Proceedings of
the 54th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynam-
ics, and Materials Conference, Boston, MA, April 2013.
doi:10.2514/6.2013-1599
[42] Zingg, D. W., Nemec, M., and Pulliam, T. H., “A Comparative
Evaluation of Genetic and Gradient-Based Algorithms Applied to
Aerodynamic Optimization,” European Journal of Computational
Mechanics, Vol. 17, Nos. 1–2, Jan. 2008, pp. 103–126.
doi:10.2514/1.40649
[43] Nocedal, J., and Wright, S. J., Numerical Optimization, 2nd ed.,
Springer–Verlag, New York, 2006, p. 321.
[44] Tappeta, R. V., and Renaud, J. E., “Multiobjective Collaborative
Optimization,” Journal of Mechanical Design, Vol. 119, No. 3, 1997,
pp. 403–411.
doi:10.1115/1.2826362
[45] Lewis, K., and Mistree, F., “The Other Side of Multidisciplinary
Design Optimization: Accommodating a Multiobjective, Uncertain
and Non-Deterministic World,” Engineering Optimization, Vol. 31,
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
No. 2, 1998, pp. 161–189.
doi:10.1080/03052159808941369
[46] Miettinen, K. M., Nonlinear Multiobjective Optimization, Kluwer
Academic, Norwell, MA, 1998.
[47] Haftka, R. T., and Watson, L. T., “Decomposition Theory for
Multidisciplinary Design Optimization Problems with Mixed Integer
Quasiseparable Subsystems,” Optimization and Engineering, Vol. 7,
No. 2, 2006, pp. 135–149.
doi:10.1007/s11081-006-6836-2
[48] Michalek, J. J., and Papalambros, P. Y., “BB-ATC: Analytical Target
Cascading Using Branch and Bound for Mixed Integer Nonlinear
Programming,” Proceedings of the ASME Design Engineering
Technical Conference, Philadelphia, PA, DET2006/DAC-99040, 2006.
[49] Zadeh, P. M., Toropov, V. V., and Wood, A. S., “Metamodel-Based
Collaborative Optimization Framework,” Structural and Multidisci-
plinary Optimization, Vol. 38, No. 2, 2009, pp. 103–115.
doi:10.1007/s00158-008-0286-8
[50] Conn, A. R., Schienburg, K., and Vicente, L. N., Introduction to
Derivative-Free Optimization, Society for Industrial and Applied
Mathematics, Philadelphia, 2009.
[51] Rao, S. S., “Game Theory Approach for Multiobjective Structural
Optimization,” Computers and Structures, Vol. 25, No. 1, 1987,
pp. 119–127.
doi:10.1016/0045-7949(87)90223-9
[52] Lewis, K., and Mistree, F., “Modeling Interactions in Multidisciplinary
Design: A Game Theoretic Approach,” AIAA Journal, Vol. 35, No. 8,
1997, pp. 1387–1392.
doi:10.2514/2.248
[53] Honda, T., Cucci, F., and Yang, M. C., “Achieving Pareto Optimality in
a Decentralized Design Environment,” Proceedings of the International
Conference on Engineering Design, Stanford, CA, Aug. 2009.
[54] Sobieszczanski–Sobieski, J., Altus, T. D., Phillips, M., and Sandusky,
R. R., Jr., “Bilevel Integrated System Synthesis for Concurrent and
Distributed Processing,” AIAA Journal, Vol. 41, No. 10, 2003,
pp. 1996–2003.
doi:10.2514/2.1889
[55] Tosserams, S., Kokkolaras, M., Etman, L. F. P., and Rooda, J. E., “A
Nonhierarchical Formulation of Analytical Target Cascading,” Journal
of Mechanical Design, Vol. 132, No. 5, 2010, pp. 051002-1–051002-13.
doi:10.1115/1.4001346
[56] Roth, B., and Kroo, I., “Enhanced Collaborative Optimization:
Application to an Analytic Test Problem and Aircraft Design,” 12th
AIAA/ISSMO Multidisciplinary Analysis and Optimization
Conference, AIAA Paper 2008-5841, Sept. 2008.
[57] Agte, J., de Weck, O., Sobieszczanski–Sobieski, J., Arendsen, P.,
Morris, A., and Spieck, M., “MDO: Assessment and Direction for
Advancement, An Opinion of One International Group,” Structural and
Multidisciplinary Optimization, Vol. 40, Nos. 1–6, 2010, pp. 17–33.
doi:10.1007/s00158-009-0381-5
[58] Kim, H. M., Michelena, N. F., Papalambros, P. Y., and Jiang, T., “Target
Cascading in Optimal System Design,” Journal of Mechanical Design,
Vol. 125, No. 3, 2003, pp. 474–480.
doi:10.1115/1.1582501
[59] Michelena, N. F., Park, H., and Papalambros, P. Y., “Convergence
Properties of Analytical Target Cascading,” AIAA Journal, Vol. 41,
No. 5, 2003, pp. 897–905.
doi:10.2514/2.2025
2071
[60] Michalek, J. J., and Papalambros, P. Y., “An Efficient Weighting
Update Method to Achieve Acceptable Consistency Deviation in
Analytical Target Cascading,” Journal of Mechanical Design, Vol. 127,
No. 2, 2005, pp. 206–214.
doi:10.1115/1.1830046
[61] Cramer, E. J., Dennis, J. E., Jr., Frank, P. D., Lewis, R. M., and Shubin,
G. R., “Problem Formulation for Multidisciplinary Optimization,”
SIAM Journal on Optimization, Vol. 4, No. 4, 1994, pp. 754–776.
doi:10.1137/0804044
[62] Alexandrov, N. M., and Lewis, R. M., “Analytical and Computational
Aspects of Collaborative Optimization for Multidisciplinary Design,”
AIAA Journal, Vol. 40, No. 2, 2002, pp. 301–309.
doi:10.2514/2.1646
[63] Alexandrov, N. M., and Lewis, R. M., “Reconfigurability in MDO
Problem Synthesis, Part 1,” 10th AIAA/ISSMO Multidisciplinary
Analysis and Optimization Conference, AIAA Paper 2004-4307,
Sept. 2004.
[64] Alexandrov, N. M., and Lewis, R. M., “Reconfigurability in MDO
Problem Synthesis, Part 2,” 10th AIAA/ISSMO Multidisciplinary
Analysis and Optimization Conference, AIAA Paper 2004-4308,
Sept. 2004.
[65] Wujek, B. A., Renaud, J. E., Batill, S. M., and Brockman, J. B.,
“Concurrent Subspace Optimization Using Design Variable Sharing in
a Distributed Computing Environment,” Concurrent Engineering:
Research and Applications, Vol. 4, No. 4, 1996, pp. 361–377.
doi:10.1177/1063293X9600400405
[66] Haftka, R. T., and Watson, L. T., “Multidisciplinary Design
Optimization with Quasiseparable Subsystems,” Optimization and
Engineering, Vol. 6, No. 1, 2005, pp. 9–20.
doi:10.1023/B:OPTE.0000048534.58121.93
[67] Sobieszczanski-Sobieski, J., and Haftka, R. T., “Multidisciplinary
Aerospace Design Optimization: Survey of Recent Developments,”
Structural Optimization, Vol. 14, No. 1, 1997, pp. 1–23.
doi:10.1007/BF01197554
[68] Kodiyalam, S., and Sobieszczanski–Sobieski, J., “Bilevel Integrated
System Synthesis with Response Surfaces,” AIAA Journal, Vol. 38,
No. 8, 2000, pp. 1479–1485.
doi:10.2514/2.1126
[69] Sellar, R. S., Batill, S. M., and Renaud, J. E., “Response Surface Based,
Concurrent Subspace Optimization for Multidisciplinary System
Design,” 34th AIAA Aerospace Sciences and Meeting Exhibit, AIAA
Paper 1996-0714, Jan. 1996.
[70] Sobieszczanski–Sobieski, J., Agte, J. S., and Sandusky, R. R., Jr.,
“Bilevel Integrated System Synthesis,” AIAA Journal, Vol. 38, No. 1,
2000, pp. 164–172.
doi:10.2514/2.937
[71] Shin, M.-K., and Park, G.-J., “Multidisciplinary Design Optimization
Based on Independent Subspaces,” International Journal for
Numerical Methods in Engineering, Vol. 64, 2005, pp. 599–617.
doi:10.1002/(ISSN)1097-0207
[72] DeMiguel, V., and Murray, W., “A Local Convergence Analysis of
Bilevel Decomposition Algorithms,” Optimization and Engineering,
Vol. 7, No. 2, 2006, pp. 99–133.
doi:10.1007/s11081-006-6835-3
[73] Braun, R. D., and Kroo, I. M., “Development and Application of
the Collaborative Optimization Architecture in a Multidisciplinary
Design Environment,” Multidisciplinary Design Optimization:
State-of-the-Art, edited by Alexandrov, N., and Hussaini, M. Y.,
Society for Industrial and Applied Mathematics, Philadelphia, 1997,
pp. 98–116.
[74] Kroo, I. M., “MDO for Large-Scale Design,” Multidisciplinary Design
Optimization: State-of-the-Art, edited by Alexandrov, N., and
Hussaini, M. Y., Society for Industrial and Applied Mathematics,
Philadelphia, 1997, pp. 22–44.
[75] Sobieski, I. P., and Kroo, I. M., “Collaborative Optimization Using
Response Surface Estimation,” AIAA Journal, Vol. 38, No. 10, 2000,
pp. 1931–1938.
doi:10.2514/2.847
[76] Chittick, I. R., and Martins, J. R. R. A., “Aero-Structural Optimiza-
tion Using Adjoint Coupled Post Optimality Sensitivities,” Structural
and Multidisciplinary Optimization, Vol. 36, No. 1, 2008,
pp. 59–70.
doi:10.1007/s00158-007-0200-9
[77] Haftka, R. T., Sobieszczanski–Sobieski, J., and Padula, S. L., “On
Options for Interdisciplinary Analysis and Design Optimization,”
Structural Optimization, Vol. 4, No. 2, 1992, pp. 65–74.
doi:10.1007/BF01759919
[78] Balling, R. J., and Sobieszczanski–Sobieski, J., “Optimization of
Coupled Systems: A Critical Overview of Approaches,” AIAA Journal,
 2072 MARTINS AND LAMBE
Vol. 34, No. 1, 1996, pp. 6–17.
doi:10.2514/3.13015
[79] Alexandrov, N., and Hussaini, M. Y. (eds.), Multidisciplinary Design
Optimization: State-of-the-Art, Society for Industrial and Applied
Mathematics, Philadelphia, 1997, pp. 22–44.
[80] Alexandrov, N. M., “Multilevel Methods for MDO,” Multidisciplinary
Design Optimization: State-of-the-Art, edited by Alexandrov, N. M.,
and Hussaini, M. Y., Society for Industrial and Applied Mathematics,
Philadelphia, 1997, pp. 79–89.
[81] Balling, R. J., “Approaches to MDO Which Support Disciplinary
Autonomy,” Multidisciplinary Design Optimization: State-of-the-Art,
edited by Alexandrov, N. M., and Hussaini, M. Y., Society for
Industrial and Applied Mathematics, Philadelphia, 1997, pp. 90–97.
[82] Willcox, K., and Wakayama, S., “Simultaneous Optimization of a
Multiple-Aircraft Family,” Journal of Aircraft, Vol. 40, No. 4, 2003,
pp. 616–622.
doi:10.2514/2.3156
[83] Piperni, P., Abdo, M., Kafyeke, F., and Isikveren, A. T., “Preliminary
Aerostructural Optimization of a Large Business Jet,” Journal of
Aircraft, Vol. 44, No. 5, 2007, pp. 1422–1438.
doi:10.2514/1.26989
[84] Simpson, T. W., and Martins, J. R. R. A., “Multidisciplinary Design
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
Optimization for Complex Engineered Systems Design: Report from
an NSF Workshop,” Journal of Mechanical Design, Vol. 133, No. 10,
Oct. 2011, pp. 101002-1–101002-10.
doi:10.1115/1.4004465
[85] Tosserams, S., Etman, L. F. P., and Rooda, J. E., “A Classification of
Methods for Distributed System Optimization Based on Formulation
Structure,” Structural and Multidisciplinary Optimization, Vol. 39,
No. 5, 2009, pp. 503–517.
doi:10.1007/s00158-008-0347-z
[86] Lambe, A. B., and Martins, J. R. R. A., “Extensions to the Design
Structure Matrix for the Description of Multidisciplinary Design,
Analysis, and Optimization Processes,” Structural and Multidiscipli-
nary Optimization, Vol. 46, No. 2, 2012, pp. 273–284.
doi:10.1007/s00158-012-0763-y
[87] Steward, D. V., “The Design Structure Matrix: A Method for Managing
the Design of Complex Systems,” IEEE Transactions on Engineering
Management, Vol. 28, 1981, pp. 71–74.
doi:10.1109/TEM.1981.6448589
[88] Browning, T. R., “Applying the Design Structure Matrix to System
Decomposition and Integration Problems: A Review and New
Directions,” IEEE Transactions on Engineering Management, Vol. 48,
No. 3, 2001, pp. 292–306.
doi:10.1109/17.946528
[89] Haftka, R. T., “Simultaneous Analysis and Design,” AIAA Journal,
Vol. 23, No. 7, 1985, pp. 1099–1103.
doi:10.2514/3.9043
[90] Schmit, L. A., Jr., and Ramanathan, R. K., “Multilevel Approach to
Minimum Weight Design Including Buckling Constraints,” AIAA
Journal, Vol. 16, No. 2, 1978, pp. 97–104.
doi:10.2514/3.60867
[91] Biegler, L. T., Ghattas, O., Heinkenschloss, M., and van Bloemen
Waanders, B. (eds.), Large-Scale PDE-Constrained Optimization,
Springer–Verlag, New York, 2003.
[92] Squire, W., and Trapp, G., “Using Complex Variables to Estimate
Derivatives of Real Functions,” SIAM Review, Vol. 40, No. 1, 1998,
pp. 110–112.
doi:10.1137/S003614459631241X
[93] Martins, J. R. R. A., Sturdza, P., and Alonso, J. J., “The Complex-Step
Derivative Approximation,” ACM Transactions on Mathematical
Software, Vol. 29, No. 3, 2003, pp. 245–262.
doi:10.1145/838250
[94] Adelman, H. M., and Haftka, R. T., “Sensitivity Analysis of
Discrete Structural Systems,” AIAA Journal, Vol. 24, No. 5, 1986,
pp. 823–832.
doi:10.2514/3.48671
[95] Jameson, A., “Aerodynamic Design via Control Theory,” Journal of
Scientific Computing, Vol. 3, No. 3, 1988, pp. 233–260.
doi:10.1007/BF01061285
[96] Giles, M. B., Duta, M. C., Muller, J.-D., and Pierce, N. A., “Algorithm
Developments for Discrete Adjoint Methods,” AIAA Journal, Vol. 41,
No. 2, 2003, pp. 198–205.
doi:10.2514/2.1961
[97] van Keulen, F., Haftka, R., and Kim, N., “Review of Options for
Structural Design Sensitivity Analysis. Part 1: Linear Systems,”
Computer Methods in Applied Mechanics and Engineering, Vol. 194,
Nos. 30–33, 2005, pp. 3213–3243.
doi:10.1016/j.cma.2005.02.002
[98] Mader, C. A., Martins, J. R. R. A., Alonso, J. J., and van der Weide, E.,
“ADjoint: An Approach for the Rapid Development of Discrete
Adjoint Solvers,” AIAA Journal, Vol. 46, No. 4, 2008, pp. 863–873.
doi:10.2514/1.29123
[99] Martins, J. R. R. A., Alonso, J. J., and Reuther, J. J., “A Coupled-
Adjoint Sensitivity Analysis Method for High-Fidelity Aero-Structural
Design,” Optimization and Engineering, Vol. 6, No. 1, 2005,
pp. 33–62.
doi:10.1023/B:OPTE.0000048536.47956.62
[100] Barcelos, M., Bavestrello, H., and Maute, K., “A Schur-Newton-
Krylov Solver for Steady-State Aeroelastic Analysis and Design
Sensitivity Analysis,” Computer Methods in Applied Mechanics and
Engineering, Vol. 195, Nos. 17–18, 2006, pp. 2050–2069.
doi:10.1016/j.cma.2004.09.013
[101] Hicken, J., and Zingg, D., “Aerodynamic Optimization Algorithm with
Integrated Geometry Parameterization and Mesh Movement,” AIAA
Journal, Vol. 48, No. 2, Feb. 2009, pp. 400–413.
doi:10.2514/1.44033
[102] Martins, J. R. R. A., and Hwang, J. T., “Review and Unification of
Methods for Computing Derivatives of Multidisciplinary Computa-
tional Models,” AIAA Journal, 2013.
doi:10.2514/1.J052184.
[103] Bloebaum, C., “Coupling Strength-Based System Reduction for
Complex Engineering Design,” Structural Optimization, Vol. 10,
No. 2, 1995, pp. 113–121.
doi:10.1007/BF01743538
[104] Kennedy, G. J., and Martins, J. R. R. A., “Parallel Solution Methods for
Aerostructural Analysis and Design Optimization,” 13th AIAA/ISSMO
Multidisciplinary Analysis Optimization Conference, AIAA Paper
2010-9308, Sept. 2010.
[105] Vanderplaats, G. N., “An Efficient Feasible Directions Algorithm for
Design Synthesis,” AIAA Journal, Vol. 22, No. 11, 1984, pp. 1633–
1640.
doi:10.2514/3.8829
[106] Gill, P. E., Murray, W., and Saunders, M. A., “SNOPT: An SQP
Algorithm for Large-Scale Constrained Optimization,” SIAM Review,
Vol. 47, No. 1, 2005, pp. 99–131.
doi:10.1137/S0036144504446096
[107] Wright, S. J., Primal-Dual Interior Point Methods, Society for
Industrial and Applied Mathematics, Philadelphia, 1997.
[108] Sobieszczanski–Sobieski, J., “Sensitivity of Complex, Internally
Coupled Systems,” AIAA Journal, Vol. 28, No. 1, 1990, pp. 153–160.
doi:10.2514/3.10366
[109] Kenway, G. K. W., Kennedy, G. J., and Martins, J. R. R. A., “A Scalable
Parallel Approach for High-Fidelity Steady-State Aeroelastic Analysis
and Derivative Computations,” AIAA Journal, 2013 (in press).
[110] Kenway, G. K. W., and Martins, J. R. R. A., “Multi-Point High-Fidelity
Aerostructural Optimization of a Transport Aircraft Configuration,”
Journal of Aircraft, 2013 (in press).
[111] Marriage, C. J., and Martins, J. R. R. A., “Reconfigurable
Semi-Analytic Sensitivity Methods and MDO Architectures Within
the πMDO Framework,” 12th AIAA/ISSMO Multidisciplinary
Analysis and Optimization Conference, AIAA Paper 2008-5956,
Sept. 2008.
[112] Martins, J. R. R. A., and Hwang, J. T., “Review and Unification of
Discrete Methods for Computing Derivatives of Single- and Multi-
Disciplinary Computational Models,” AIAA Journal (in press).
[113] Lasdon, L. S., Optimization Theory for Large Systems, Macmillan,
New York, 1970, pp. 104–143, Chap. 2.
[114] Conejo, A. J., Nogales, F. J., and Prieto, F. J., “A Decomposition
Procedure Based on Approximate Newton Directions,” Mathematical
Programming, Vol. 93, No. 3, 2002, pp. 495–515.
doi:10.1007/s10107-002-0304-3
[115] Dantzig, G. B., and Wolfe, P., “Decomposition Principle for Linear
Programs,” Operations Research, Vol. 8, No. 1, 1960, pp. 101–111.
doi:10.1287/opre.8.1.101
[116] Benders, J. F., “Partitioning Procedures for Solving Mixed Variables
Programming Problems,” Numerische Mathematik, Vol. 4, No. 1,
1962, pp. 238–252.
doi:10.1007/BF01386316
[117] Michelena, N. F., and Papalambros, P. Y., “A Network Reliability
Approach to Optimal Decomposition of Design Problems,” Journal of
Mechanical Design, Vol. 117, No. 3, 1995, pp. 433–440.
doi:10.1115/1.2826697
[118] Michelena, N. F., and Papalambros, P. Y., “A Hypergraph Framework
for Optimal Model-Based Decomposition of Design Problems,”
Computational Optimization and Applications, Vol. 8, No. 2, 1997,
pp. 173–196.
doi:10.1023/A:1008673321406
 MARTINS AND LAMBE
[119] Tosserams, S., Hofkamp, A. T., Etman, L. F. P., and Rooda, J. E., “A
Specification Language for Problem Partitioning in Decomposition-
Based Design Optimization,” Structural and Multidisciplinary
Optimization, Vol. 42, No. 5, 2010, pp. 707–723.
doi:10.1007/s00158-010-0512-z
[120] Bertsekas, D. P., and Tsitsiklis, J. N., Parallel and Distributed
Computation: Numerical Methods, Athena Scientific, Belmont, MA,
1997, pp. 88–104.
[121] Zadeh, P. M., and Toropov, V. V., “Multi-Fidelity Multidisciplinary
Design Optimization Based on Collaborative Optimization Frame-
work,” 9th AIAA/ISSMO Symposium on Multidisciplinary Analysis
and Optimization, AIAA Paper 2002-5504, Sept. 2002.
[122] Robinson, T. D., Willcox, K. E., Eldred, M. S., and Haimes, R.,
“Multifidelity Optimization for Variable Complexity Design,” 11th
AIAA/ISSMO Multidisciplinary Analysis and Optimization
Conference, AIAA Paper 2006-7114, Sept. 2006.
[123] Chittick, I. R., and Martins, J. R. R. A., “An Asymmetric Sub-
optimization Approach to Aerostructural Optimization,” Optimization
and Engineering, Vol. 10, No. 1, 2009, pp. 133–152.
doi:10.1007/s11081-008-9046-2
[124] Alexandrov, N. M., and Lewis, R. M., “Comparative Properties of
Collaborative Optimization and Other Approaches to MDO,” NASA
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
Technical Rept. CR-1999-209354, Hampton, VA, July 1999.
[125] De Wit, A. J., and Van Keulen, F., “Overview of Methods for Multilevel
and/or Multidisciplinary Optimization,” 51st AIAA/ASME/ASCE/AHS/
ASC Structures, Structural Dynamics, and Materials Conference,
AIAA Paper 2010-2914, April 2010.
[126] Geethaikrishnan, C., Mujumdar, P. M., Sudhakar, K., and Adimurthy,
V., “A Hybrid MDO Architecture for Launch Vehicle Conceptual
Design,” 51st AIAA/ASME/ASCE/AHS/ASC Structures, Structural
Dynamics, and Materials Conference, AIAA Paper 2010-2757,
April 2010.
[127] Sobieszczanski–Sobieski, J., “Optimization by Decomposition: A Step
from Hierarchic to Non- Hierarchic Systems,” NASA Langley
Research Center TR-CP-3031, Hampton, VA, Sept. 1988.
[128] Bloebaum, C. L., Hajela, P., and Sobieszczanski–Sobieski, J., “Non-
Hierarchic System Decomposition in Structural Optimization,”
Engineering Optimization, Vol. 19, No. 3, 1992, pp. 171–186.
doi:10.1080/03052159208941227
[129] Shankar, J., Ribbens, C. J., Haftka, R. T., and Watson, L. T.,
“Computational Study of a Nonhierarchical Decomposition Algo-
rithm,” Computational Optimization and Applications, Vol. 2, No. 3,
1993, pp. 273–293.
doi:10.1007/BF01299452
[130] Renaud, J. E., and Gabriele, G. A., “Approximation in Non-Hierarchic
System Optimization,” AIAA Journal, Vol. 32, No. 1, 1994, pp. 198–205.
doi:10.2514/3.11967
[131] Renaud, J. E., and Gabriele, G. A., “Improved Coordination in
Nonhierarchic System Optimization,” AIAA Journal, Vol. 31, No. 12,
1993, pp. 2367–2373.
doi:10.2514/3.11938
[132] Parashar, S., and Bloebaum, C. L., “Multi-Objective Genetic
Algorithm Concurrent Subspace Optimization (MOGACSSO) for
Multidisciplinary Design,” 11th AIAA/ISSMO Multidisciplinary
Analysis and Optimization Conference, AIAA Paper 2006-2047,
May 2006.
[133] Huang, C.-H., Galuski, J., and Bloebaum, C. L., “Multi-Objective
Pareto Concurrent Subspace Optimization for Multidisciplinary
Design,” AIAA Journal, Vol. 45, No. 8, 2007, pp. 1894–1906.
doi:10.2514/1.19972
[134] Zhang, K.-S., Han, Z.-H., Li, W.-J., and Song, W.-P., “Bilevel
Adaptive Weighted Sum Method for Multidisciplinary Multi-
Objective Optimization,” AIAA Journal, Vol. 46, No. 10, 2008,
pp. 2611–2622.
doi:10.2514/1.36853
[135] Parashar, S., and Bloebaum, C. L., “Robust Multi-Objective Genetic
Algorithm Concurrent Subspace Optimization (R-MOGACSSO) for
Multidisciplinary Design,” 11th AIAA/ISSMO Multidisciplinary
Analysis and Optimization Conference, AIAA Paper 2006-6907,
Sept. 2006.
[136] Tappeta, R. V., Nagendra, S., and Renaud, J. E., “A Multidisciplinary
Design Optimization Approach for High Temperature Aircraft Engine
Components,” Structural Optimization, Vol. 18, Nos. 2–3, 1999,
pp. 134–145.
doi:10.1007/BF01195988
[137] Perez, R. E., Liu, H. H. T., and Behdinan, K., “Evaluation of
Multidisciplinary Optimization Approaches for Aircraft Conceptual
Design,” 10th AIAA/ISSMO Muldisciplinary Analysis and Optimiza-
tion Conference, AIAA Paper 2004-4537, Aug. 2004.
2073
[138] Yi, S. I., Shin, J. K., and Park, G. J., “Comparison of MDO Methods
with Mathematical Examples,” Structural and Multidisciplinary
Optimization, Vol. 39, No. 5, 2008, pp. 391–402.
doi:10.1007/s00158-007-0150-2
[139] Tedford, N. P., and Martins, J. R. R. A., “Benchmarking
Multidisciplinary Design Optimization Algorithms,” Optimization
and Engineering, Vol. 11, No. 1, 2010, pp. 159–183.
doi:10.1007/s11081-009-9082-6
[140] de Wit, A. J., and van Keulen, F., “Numerical Comparison of Multi-
Level Optimization Techniques,” 48th AIAA/ASME/ASCE/AHS/ASC
Structures, Structural Dynamics, and Materials Conference, AIAA
Paper 2007-1895, April 2007.
[141] Braun, R. D., “Collaborative Optimization: An Architecture for
Large-Scale Distributed Design,” Ph.D. Thesis, Stanford Univ.,
Stanford, CA, 1996.
[142] Braun, R. D., Gage, P. J., Kroo, I. M., and Sobieski, I. P.,
“Implementation and Performance Issues in Collaborative Optimiza-
tion,” 6th AIAA/USAF/NASA/ISSMO Multidisciplinary Analysis and
Optimization Symposium, AIAA Paper 1996-4017, Sept. 1996.
[143] DeMiguel, A.-V., and Murray, W., “An Analysis of Collaborative
Optimization Methods,” 8th AIAA/USAF/NASA/ISSMO Symposium on
Multidisciplinary Analysis and Optimization, AIAA Paper 2000-4720,
2000.
[144] Lin, J. G., “Analysis and Enhancement of Collaborative Optimization
for Multidisciplinary Design,” AIAA Journal, Vol. 42, No. 2, 2004,
pp. 348–360.
doi:10.2514/1.9098
[145] Fiacco, A. V., and McCormick, G. P., Nonlinear Programming:
Sequential Unconstrained Minimization Techniques, Society for
Industrial and Applied Mathematics, Philadelphia, 1990, pp. 53–72.
[146] Brown, N. F., and Olds, J. R., “Evaluation of Multidisciplinary
Optimization Techniques Applied to a Reusable Launch Vehicle,”
Journal of Spacecraft and Rockets, Vol. 43, No. 6, 2006, pp. 1289–
1300.
doi:10.2514/1.16577
[147] Perez, R. E., “A Multidisciplinary Optimization Framework for Flight
Dynamics and Control Integration in Aircraft Design,” Ph.D. Thesis,
Univ. of Toronto, Toronto, 2007.
[148] Marriage, C., “Automatic Implementation of Multidisciplinary Design
Optimization Architectures Using πMDO,” M.S. Thesis, Univ. of
Toronto, Toronto, 2008.
[149] Li, X., Li, W., and Liu, C., “Geometric Analysis of Collaborative
Optimization,” Structural and Multidisciplinary Optimization, Vol. 35,
No. 4, 2008, pp. 301–313.
doi:10.1007/s00158-007-0127-1
[150] Budianto, I. A., and Olds, J. R., “Design and Deployment of a Satellite
Constellation Using Collaborative Optimization,” Journal of Space-
craft and Rockets, Vol. 41, No. 6, 2004, pp. 956–963.
doi:10.2514/1.14254
[151] Ledsinger, L. A., and Olds, J. R., “Optimized Solutions for Kistler K-1
Branching Trajectories Using Multidisciplinary Design Optimization
Techniques,” Journal of Spacecraft and Rockets, Vol. 39, No. 3, 2002,
pp. 420–429.
doi:10.2514/2.3825
[152] Perez, R. E., Liu, H. H. T., and Behdinan, K., “Multidisciplinary
Optimization Framework for Control-Configuration Integration in
Aircraft Conceptual Design,” Journal of Aircraft, Vol. 43, No. 6, 2006,
pp. 1937–1948.
doi:10.2514/1.22263
[153] Allison, J. T., Roth, B., Kokkolaras, M., Kroo, I. M., and Papalambros,
P. Y., “Aircraft Family Design Using Decomposition-Based Methods,”
11th AIAA/ISSMO Multidisciplinary Analysis and Optimization
Conference, AIAA Paper 2006-6950, Sept. 2006.
[154] McAllister, C. D., Simpson, T. W., Hacker, K., Lewis, K., and Messac,
A., “Integrating Linear Physical Programming Within Collaborative
Optimization for Multiobjective Multidisciplinary Design Optimiza-
tion,” Structural and Multidisciplinary Optimization, Vol. 29, No. 3,
2005, pp. 178–189.
doi:10.1007/s00158-004-0481-1
[155] Gu, X., Renaud, J. E., Ashe, L. M., Batill, S. M., Budhiraja, A. S., and
Krajewski, L. J., “Decision-Based Collaborative Optimization,”
Journal of Mechanical Design, Vol. 124, No. 1, 2002, pp. 1–13.
doi:10.1115/1.1432991
[156] Gu, X. S., Renaud, J. E., and Penninger, C. L., “Implicit Uncertainty
Propagation for Robust Col-laborative Optimization,” Journal of
Mechanical Design, Vol. 128, No. 4, 2006, pp. 1001–1013.
doi:10.1115/1.2205869
[157] Huang, H.-Z., Tao, Y., and Liu, Y., “Multidisciplinary Collaborative
Optimization Using Fuzzy Satisfaction Degree and Fuzzy Sufficiency
 2074 MARTINS AND LAMBE
Degree Model,” Soft Computing, Vol. 12, No. 10, 2008, pp. 995–1005.
doi:10.1007/s00500-007-0268-6
[158] Roth, B. D., “Aircraft Family Design Using Enhanced Collaborative
Optimization,” Ph.D. Thesis, Stanford Univ., Stanford, CA, 2008.
[159] Kodiyalam, S., “Evaluation of Methods for Multidisciplinary
Design Optimization (MDO), Phase I,” NASA CR-1998-208716,
Sept. 1998.
[160] Kodiyalam, S., and Yuan, C., “Evaluation of Methods for Multi-
disciplinary Design Optimization (MDO), Part 2,” NASA CR-2000-
210313, Nov. 2000.
[161] Conn, A. R., Gould, N. I. M., and Toint, P. L., Trust Region Methods,
Society for Industrial and Applied Mathematics, Philadelphia, 2000.
[162] Kim, H., Ragon, S., Soremekun, G., Malone, B., and Sobieszczanski–
Sobieski, J., “Flexible Approximation Model Approach for Bi-Level
Integrated System Synthesis,” 10th AIAA/ISSMO Multidisciplinary
Analysis and Optimization Conference, AIAA Paper 2004-4545,
Aug. 2004.
[163] Ahn, J., and Kwon, J. H., “An Efficient Strategy for Reliability-Based
Multidisciplinary Design Optimization Using BLISS,” Structural and
Multidisciplinary Optimization, Vol. 31, No. 5, 2006, pp. 363–372.
doi:10.1007/s00158-005-0565-6
[164] LeGresley, P. A., and Alonso, J. J., “Improving the Performance of
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
Design Decomposition Methods with POD,” 10th AIAA/ISSMO
Multidisciplinary Analysis and Optimization Conference, AIAA Paper
2004-4465, Aug. 2004.
[165] Berkooz, G., Holmes, P., and Lumley, J. L., “The Proper Orthogonal
Decomposition in the Analysis of Turbulent Flows,” Annual Review of
Fluid Mechanics, Vol. 25, 1993, pp. 539–575.
doi:10.1146/annurev.fl.25.010193.002543
[166] Sobieszczanski–Sobieski, J., “Integrated System-of-Systems Synthe-
sis,” AIAA Journal, Vol. 46, No. 5, 2008, pp. 1072–1080.
doi:10.2514/1.27953
[167] Kim, H. M., “Target Cascading in Optimal System Design,” Ph.D.
Thesis, Univ. of Michigan, Ann Arbor, MI, 2001.
[168] Tosserams, S., Etman, L. F. P., and Rooda, J. E., “Augmented
Lagrangian Coordination for Distributed Optimal Design in MDO,”
International Journal for Numerical Methods in Engineering, Vol. 73,
2008, pp. 1885–1910.
doi:10.1002/(ISSN)1097-0207
[169] Kim, H. M., Chen, W., and Wiecek, M. M., “Lagrangian Coordination
for Enhancing the Convergence of Analytical Target Cascading,” AIAA
Journal, Vol. 44, No. 10, 2006, pp. 2197–2207.
doi:10.2514/1.15326
[170] Tosserams, S., Etman, L. F. P., Papalambros, P. Y., and Rooda, J. E.,
“An Augmented Lagrangian Relaxation for Analytical Target
Cascading Using the Alternating Direction Method of Multipliers,”
Structural and Multidisciplinary Optimization, Vol. 31, No. 3, 2006,
pp. 176–189.
doi:10.1007/s00158-005-0579-0
[171] Bertsekas, D. P., Constrained Optimization and Lagrange Multiplier
Methods, Athena Scientific, Belmont, MA, 1996, pp. 104–125.
[172] Li, Y., Lu, Z., and Michalek, J. J., “Diagonal Quadratic Approximation
for Parallelization of Analytical Target Cascading,” Journal of
Mechanical Design, Vol. 130, No. 5, 2008, pp. 051402-1–051402-11.
doi:10.1115/1.2838334
[173] Ruszczynski, A., “On Convergence of an Augmented Lagrangian
Decomposition Method for Sparse Convex Optimization,” Mathemat-
ics of Operations Research, Vol. 20, No. 3, 1995, pp. 634–656.
doi:10.1287/moor.20.3.634
[174] Han, J., and Papalambros, P. Y., “A Sequential Linear Programming
Coordination Algorithm for Analytical Target Cascading,” Journal of
Mechanical Design, Vol. 132, No. 3, 2010, pp. 021003-1–021003-8.
doi:10.1115/1.4000758
[175] Bazaara, M. S., Sherali, H. D., and Shetty, C. M., Nonlinear
Programming: Theory and Algorithms, Wiley, New York, 2006,
pp. 568–576.
[176] Chen, T.-Y., “Calculation of the Move Limits for the Sequential Linear
Programming Method,” International Journal for Numerical Methods
in Engineering, Vol. 36, No. 15, 1993, pp. 2661–2679.
doi:10.1002/(ISSN)1097-0207
[177] Han, J., and Papalambros, P. Y., “A Note on the Convergence of
Analytical Target Cascading with Infinite Norms,” Journal of
Mechanical Design, Vol. 132, No. 3, 2010, pp. 034502-1–034502-6.
doi:10.1115/1.4001001
[178] Kokkolaras, M., Fellini, R., Kim, H. M., Michelena, N. F., and
Papalambros, P. Y., “Extension of the Target Cascading Formulation to
the Design of Product Families,” Structural and Multidisciplinary
Optimization, Vol. 24, No. 4, 2002, pp. 293–301.
doi:10.1007/s00158-002-0240-0
[179] Kim, H. M., Kokkolaras, M., Louca, L. S., Delagrammatikas, G. J.,
Michelena, N. F., Filipi, Z. S., Papalambros, P. Y., Stein, J. L., and
Assanis, D. N., “Target Cascading in Vehicle Redesign: A Class VI
Truck Study,” International Journal of Vehicle Design, Vol. 29, No. 3,
2002, pp. 199–225.
doi:10.1504/IJVD.2002.002010
[180] Kim, H. M., Rideout, D. G., Papalambros, P. Y., and Stein, J. L.,
“Analytical Target Cascading in Automotive Vehicle Design,” Journal
of Mechanical Design, Vol. 125, No. 3, 2003, pp. 481–490.
doi:10.1115/1.1586308
[181] Blouin, V. Y., Fadel, G. M., Haque, I. U., Wagner, J. R., and Samuels,
H. B., “Continuously Variable Transmission Design for Optimum
Vehicle Performance by Analytical Target Cascading,” Interna-
tional Journal of Heavy Vehicle Systems, Vol. 11, No. 3/4, 2004,
pp. 327–348.
doi:10.1504/IJHVS.2004.005454
[182] Tajima, J., Momiyama, F., and Yuhara, N., “A New Solution for
Two-Bag Air Suspension System with Leaf Spring for Heavy-
Duty Vehicle,” Vehicle System Dynamics, Vol. 44, No. 2, 2006,
pp. 107–138.
doi:10.1080/00423110500385907
[183] Chen, Y., Chen, X., and Lin, Y., “The Application of Analytical Target
Cascading in Parallel Hybrid Electric Vehicle,” IEEE Vehicle Power
and Propulsion Conference, IEEE Publ., Piscataway, NJ, 2009,
pp. 1602–1607.
[184] Han, J., and Papalambros, P. Y., “Optimal Design of Hybrid Electric
Fuel Cell Vehicles Under Uncer-tainty and Enterprise Considerations,”
Journal of Fuel Cell Science and Technology, Vol. 7, No. 2, 2010,
pp. 021020-1–021020-9.
doi:10.1115/1.3179762
[185] Allison, J. T., Walsh, D., Kokkolaras, M., Papalambros, P. Y., and
Cartmell, M., “Analytical Target Cascading in Aircraft Design,” 44th
AIAA Aerospace Sciences Meeting and Exhibit, AIAA Paper 2006-
1325, Jan. 2006.
[186] Li, Z., Kokkolaras, M., Papalambros, P. Y., and Hu, S. J., “Product and
Process Tolerance Allocation in Multistation Compliant Assembly
Using Analytical Target Cascading,” Journal of Mechanical Design,
Vol. 130, No. 9, 2008, pp. 091701-1–091701-9.
doi:10.1115/1.2943296
[187] Huang, G. Q., and Qu, T., “Extending Analytical Target Cascading for
Optimal Configuration of Supply Chains with Alternative
Autonomous Suppliers,” International Journal of Production
Economics, Vol. 115, No. 1, 2008, pp. 39–54.
doi:10.1016/j.ijpe.2008.04.008
[188] Michalek, J. J., Feinberg, F. M., and Papalambros, P. Y., “Linking
Marketing and Engineering Product Design Decisions via Analytical
Target Cascading,” Journal of Product Innovation Management,
Vol. 22, No. 1, 2005, pp. 42–62.
doi:10.1111/jpim.2005.22.issue-1
[189] Huang, G. Q., Qu, T., Cheung, D. W. L., and Liang, L., “Extensible
Multi-Agent System for Optimal Design of Complex Systems Using
Analytical Target Cascading,” International Journal of Advanced
Manufacturing Technology, Vol. 30, Nos. 1–10, 2006, pp. 917–926.
doi:10.1007/s00170-005-0064-3
[190] Etman, L. F. P., Kokkolaras, M., Hofkamp, A. T., Papalambros, P. Y.,
and Rooda, J. E., “Coordination Specification in Distributed Optimal
Design of Multilevel Systems Using the χ Language,” Structural and
Multidisciplinary Optimization, Vol. 29, No. 3, 2005, pp. 198–212.
doi:10.1007/s00158-004-0467-z
[191] Kokkolaras, M., Mourelatos, Z. P., and Papalambros, P. Y., “Design
Optimization of Hierarchically Decomposed Multilevel Systems
Under Uncertainty,” Journal of Mechanical Design, Vol. 128, No. 2,
2006, pp. 503–508.
doi:10.1115/1.2168470
[192] Liu, H., Chen, W., Kokkolaras, M., Papalambros, P. Y., and Kim, H. M.,
“Probabilistic Analytical Target Cascading: A Moment Matching
Formulation for Multilevel Optimization Under Uncertainty,” Journal
of Mechanical Design, Vol. 128, No. 4, 2006, pp. 991–1000.
doi:10.1115/1.2205870
[193] Tosserams, S., Etman, L. F. P., and Rooda, J. E., “Block-Separable
Linking Constraints in Augmented Lagrangian Coordination in
MDO,” Structural and Multidisciplinary Optimization, Vol. 37, No. 5,
2009, pp. 521–527.
doi:10.1007/s00158-008-0244-5
[194] DeMiguel, A.-V., “Two Decomposition Algorithms for Nonconvex
Optimization Problems with Global Variables,” Ph.D. Thesis, Stanford
Univ., Stanford, CA, 2001.
[195] Tosserams, S., Etman, L. F. P., and Rooda, J. E., “An Augmented
Lagrangian Decomposition Method for Quasiseparable Problems in
 MARTINS AND LAMBE
MDO,” Structural and Multidisciplinary Optimization, Vol. 34, No. 3,
2007, pp. 211–227.
doi:10.1007/s00158-006-0077-z
[196] Shin, M. K., Kang, B. S., and Park, G. J., “Application of the [205]
Multidisciplinary Design Optimization Algorithm to the Design of a
Belt-Integrated Seat While Considering Crashworthiness,” Journal of
Automobile Engineering, Vol. 219, No. 11, 2005, pp. 1281–1292.
doi:10.1243/095440705X34928 [206]
[197] Liu, B., Haftka, R. T., and Watson, L. T., “Global-Local Structural
Optimization Using Response Surfaces of Local Optimization
Margins,” Structural and Multidisciplinary Optimization, Vol. 27,
No. 5, 2004, pp. 352–359.
doi:10.1007/s00158-004-0393-0 [207]
[198] Martins, J. R. R. A., Alonso, J. J., and Reuther, J. J., “High-Fidelity
Aerostructural Design Optimization of a Supersonic Business Jet,”
Journal of Aircraft, Vol. 41, No. 3, 2004, pp. 523–530.
doi:10.2514/1.11478 [208]
[199] Kennedy, G., Martins, J. R. R. A., and Hansen, J. S., “Aerostructural
Optimization of Aircraft Structures Using Asymmetric Subspace
Optimization,” 12th AIAA/ISSMO Multidisciplinary Analysis and
Optimization Conference, AIAA Paper 2008-5847, Sept. 2008.
[200] Kaufman, M., Balabanov, V., Burgee, S. L., Giunta, A. A., Grossman, [209]
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895
B., Haftka, R. T., Mason, W. H., and Watson, L. T., “Variable-
Complexity Response Surface Approximations for Wing Structural
Weight in HSCT Design,” Computational Mechanics, Vol. 18, No. 2,
1996, pp. 112–126. [210]
doi:10.1007/BF00350530
[201] Hosder, S., Watson, L. T., Grossman, B., Mason, W. H., Kim, H.,
Haftka, R. T., and Cox, S. E., “Polynomial Response Surface
Approximations for the Multidisciplinary Design Optimization of a [211]
High Speed Civil Transport,” Optimization and Engineering, Vol. 2,
No. 4, 2001, pp. 431–452.
doi:10.1023/A:1016094522761
[202] Hulme, K. F., and Bloebaum, C. L., “A Simulation-Based Comparison
of Multidisciplinary Design Optimization Solution Strategies Using [212]
CASCADE,” Structural and Multidisciplinary Optimization, Vol. 19,
No. 1, 2000, pp. 17–35.
doi:10.1007/s001580050083
[203] Allison, J. T., Kokkolaras, M., and Papalambros, P. Y., “On Selecting
Single-Level Formulations for Complex System Design Optimization,”
Journal of Mechanical Design, Vol. 129, No. 9, 2007, pp. 898–906.
doi:10.1115/1.2747632
[204] Martins, J. R. R. A., Marriage, C., and Tedford, N., “pyMDO: An
Object-Oriented Framework for Multidisciplinary Design Optimiza-
2075
tion,” ACM Transactions on Mathematical Software, Vol. 36, No. 4,
2009, pp. 20:1–20:25.
doi:10.1145/1555386
Gray, J., Moore, K. T., and Naylor, B. A., “OpenMDAO: An Open
Source Framework for Multidisciplinary Analysis and Optimization,”
13th AIAA/ISSMO Multidisciplinary Analysis Optimization
Conference, AIAA Paper 2010-9101, Sept. 2010.
Lu, Z., and Martins, J. R. R. A., “Graph Partitioning-Based
Coordination Methods for Large-Scale Multidisciplinary Design
Optimization Problems,” Proceedings of the 14th AIAA/ISSMO
Multidisciplinary Analysis Optimization Conference, Indianapolis, IN,
AIAA Paper 2012-5522, 2012.
Padula, S. L., and Gillian, R. E., “Multidisciplinary Environments: A
History of Engineering Framework Development,” 11th AIAA/ISSMO
Multidisciplinary Analysis and Optimization Conference, AIAA Paper
2006-7083, Sept. 2006.
Kodiyalam, S., and Sobieszczanski–Sobieski, J., “Multidisciplinary
Design Optimization: Somemal Methods, Framework Requirements,
and Application to Vehicle Design,” International Journal of Vehicle
Design, Vol. 25, No. 1/2, 2001, pp. 3–22.
doi:10.1504/IJVD.2001.001904
Kodiyalam, S., Kremenetsky, M., and Posey, S., “Balanced HPC
Infrastructure for CFD and Associated Multi-Discipline Simulations of
Engineering Systems,” Journal of Aerospace Sciences and
Technologies, Vol. 61, No. 3, 2009, pp. 434–443.
Tosserams, S., Etman, L. F. P., and Rooda, J. E., “A Micro-
Accelerometer MDO Benchmark Problem,” Structural and Multidis-
ciplinary Optimization, Vol. 41, No. 2, 2010, pp. 255–275.
doi:10.1007/s00158-009-0422-0
Tosserams, S., Etman, L. F. P., and Rooda, J. E., “Multi-Modality in
Augmented Lagrangian Co-ordination for Distributed Optimal
Design,” Structural and Multidisciplinary Optimization, Vol. 40,
Nos. 1–6, 2010, pp. 329–352.
doi:10.1007/s00158-009-0371-7
Padula, S. L., Alexandrov, N., and Green, L. L., “MDO Test Suite at
NASA Langley Research Center,” 6th AIAA/NASA/ISSMO Symposium
on Multidisciplinary Analysis and Optimization, AIAA Paper 1996-
4028, Sept. 1996.
P. Givi
Associate Editor
 This article has been cited by:

1. Imco van Gent, Gianfranco La Rocca. 2019. Formulation and integration of MDAO systems for collaborative design: A
graph-based methodological approach. Aerospace Science and Technology 90, 410-433. [Crossref]
2. Tingting Xia, Mian Li. 2019. A Sequential Approach for Robust Multidisciplinary Design Optimization Under Mixed
Interval and Probabilistic Uncertainties. ASCE-ASME J. Risk and Uncert. in Engrg. Sys., Part B: Mech. Engrg. 5:2, 020905.
[Crossref]
3. T. Hirschler, R. Bouclier, A. Duval, T. Elguedj, J. Morlier. 2019. The embedded isogeometric Kirchhoff–Love shell: From
design to shape optimization of non-conforming stiffened multipatch structures. Computer Methods in Applied Mechanics
and Engineering 349, 774-797. [Crossref]
4. Anh Tran, Minh Tran, Yan Wang. 2019. Constrained mixed-integer Gaussian mixture Bayesian optimization and its
applications in designing fractal and auxetic metamaterials. Structural and Multidisciplinary Optimization 59:6, 2131-2154.
[Crossref]
5. E. Sepulveda, H. Smith, D. Sziroczak. 2019. Multidisciplinary analysis of subsonic stealth unmanned combat aerial vehicles.
CEAS Aeronautical Journal 10:2, 431-442. [Crossref]
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

6. Thanh Nguyen, Snehasis Mukhopadhyay, Meghna Babbar-Sebens. 2019. Why the ‘selfish’ optimizing agents could solve
the decentralized reinforcement learning problems. AI Communications 32:2, 143-159. [Crossref]
7. G. Pagliuca, T. Kipouros, M. A. Savill. 2019. Surrogate Modelling for Wing Planform Multidisciplinary Optimisation
Using Model-Based Engineering. International Journal of Aerospace Engineering 2019, 1-15. [Crossref]
8. Janet L. Ellzey, Erica L. Belmont, Colin H. Smith. 2019. Heat recirculating reactors: Fundamental research and
applications. Progress in Energy and Combustion Science 72, 32-58. [Crossref]
9. Xiaojian Sun, Jianquan Ge, Tao Yang, Qiangqiang Xu, Bin Zhang. 2019. Multifidelity Multidisciplinary Design
Optimization of Integral Solid Propellant Ramjet Supersonic Cruise Vehicles. International Journal of Aerospace Engineering
2019, 1-18. [Crossref]
10. Haowei Gu, Ximin Lyu, Zexiang Li, Fu Zhang. 2019. Coordinate Descent Optimization for Winged-UAV Design. Journal
of Intelligent & Robotic Systems 24. . [Crossref]
11. N Cimmino, SS Ponnusamy, A Garcia Garriga, L Mainini. 2019. A modelling and simulation framework for the integrated
design of aircraft systems. Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering
2, 095441001983572. [Crossref]
12. Sinan Eyi, Kyle M. Hanquist, Iain D. Boyd. 2019. Aerothermodynamic Design Optimization of Hypersonic Vehicles.
Journal of Thermophysics and Heat Transfer 33:2, 392-406. [Abstract] [Full Text] [PDF] [PDF Plus]
13. Justin S. Gray, John T. Hwang, Joaquim R. R. A. Martins, Kenneth T. Moore, Bret A. Naylor. 2019. OpenMDAO:
an open-source framework for multidisciplinary design, analysis, and optimization. Structural and Multidisciplinary
Optimization 59:4, 1075-1104. [Crossref]
14. Samuel Vogel. 2019. An application-independent continuum mechanics interface for virtual engineering. Engineering with
Computers 35:2, 551-565. [Crossref]
15. Wei Zhao, Rakesh K. Kapania. Bilevel Programming Weight Minimization of Composite Flying-Wing Aircraft with
Curvilinear Spars and Ribs. AIAA Journal, ahead of print1-15. [Abstract] [Full Text] [PDF] [PDF Plus] [Supplementary
Material]
16. Debiao Meng, Yan Li, Shun-Peng Zhu, Gang Lv, José Correia, Abílio de Jesus. 2019. An Enhanced Reliability Index
Method and Its Application in Reliability-Based Collaborative Design and Optimization. Mathematical Problems in
Engineering 2019, 1-10. [Crossref]
17. Wei Li, Liang Gao, Mi Xiao. 2019. Multidisciplinary robust design optimization under parameter and model uncertainties.
Engineering Optimization 40, 1-20. [Crossref]
18. Gang Du, Yi Xia, Roger J. Jiao, Xiaojie Liu. 2019. Leader-follower joint optimization problems in product family design.
Journal of Intelligent Manufacturing 30:3, 1387-1405. [Crossref]
19. Renhe Shi, Li Liu, Teng Long, Yufei Wu, Yifan Tang. 2019. Filter-Based Sequential Radial Basis Function Method for
Spacecraft Multidisciplinary Design Optimization. AIAA Journal 57:3, 1019-1031. [Abstract] [Full Text] [PDF] [PDF
Plus]
 20. Nicolas P. Bons, Xiaolong He, Charles A. Mader, Joaquim R. R. A. Martins. 2019. Multimodality in Aerodynamic Wing
Design Optimization. AIAA Journal 57:3, 1004-1018. [Abstract] [Full Text] [PDF] [PDF Plus]
21. Shuvodeep De, Mohamed Jrad, Rakesh K. Kapania. 2019. Structural Optimization of Internal Structure of Aircraft Wings
with Curvilinear Spars and Ribs. Journal of Aircraft 56:2, 707-718. [Abstract] [Full Text] [PDF] [PDF Plus]
22. T. Hirschler, R. Bouclier, A. Duval, T. Elguedj, J. Morlier. 2019. Isogeometric sizing and shape optimization of thin
structures with a solid-shell approach. Structural and Multidisciplinary Optimization 59:3, 767-785. [Crossref]
23. Daniel Hulse, Christopher Hoyle, Kai Goebel, Irem Y. Tumer. 2019. Quantifying the Resilience-Informed Scenario Cost
Sum: A Value-Driven Design Approach for Functional Hazard Assessment. Journal of Mechanical Design 141:2, 021403.
[Crossref]
24. Daniel Hulse, Kagan Tumer, Christopher Hoyle, Irem Tumer. 2019. Modeling multidisciplinary design with multiagent
learning. Artificial Intelligence for Engineering Design, Analysis and Manufacturing 33:1, 85-99. [Crossref]
25. John T. Hwang, John P. Jasa, Joaquim R. R. A. Martins. High-Fidelity Design-Allocation Optimization of a Commercial
Aircraft Maximizing Airline Profit. Journal of Aircraft, ahead of print1-15. [Abstract] [Full Text] [PDF] [PDF Plus]
26. Daniel J. Poole, Christian B. Allen, T. Rendall. Efficient Aero-Structural Wing Optimization Using Compact Aerofoil
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

Decomposition . [Citation] [PDF] [PDF Plus]

27. Joaquim Martins, Graeme Kennedy. Enabling Large-scale Multidisciplinary Design Optimization through Adjoint
Sensitivity Analysis . [Citation] [PDF] [PDF Plus]
28. Tonghui Cui, Pingfeng Wang, Koki Ho. Staged-Deployment Design for Resilient Expansion Planning of Large Scale
Complex Systems . [Citation] [PDF] [PDF Plus]
29. Hao Xu, Yu Sheng. Propeller Blades Design for the Bidirectional Rotor Systems of Fault Tolerant Multirotors . [Citation]
[PDF] [PDF Plus]
30. Alexander D. Taylor, Alexandre D. Payot, Christian B. Allen, T. Rendall. Structural Topology Optimisation with R-
Snakes Volume of Solid . [Citation] [PDF] [PDF Plus]
31. Giulia E. Palma, Bryan Mesmer, Amy Guerin, Kristin Weger. Developing the “Trading Places” Boot Camp: Sharing
Knowledge Between Theatre and Engineering . [Citation] [PDF] [PDF Plus]
32. Carina Marcus. Balancing Antenna Performance vs. Radar Cross Section for a Passive Radar-Detecting Sensor on an
Aircraft . [Citation] [PDF] [PDF Plus]
33. Ping He, Charles A. Mader, Joaquim Martins, Kevin Maki. An Object-oriented Framework for Rapid Discrete Adjoint
Development using OpenFOAM . [Citation] [PDF] [PDF Plus]
34. Nathaniel S. Gates, Kevin R. Moore, Andrew Ning, John D. Hedengren. Combined Trajectory and Propulsion
Optimization for Solar-Regenerative High-Altitude Long Endurance Unmanned Aircraft . [Citation] [PDF] [PDF Plus]
35. Brian W. Chell, Steven Hoffenson, Mark R. Blackburn. A Comparison of Multidisciplinary Design Optimization
Architectures with an Aircraft Case Study . [Citation] [PDF] [PDF Plus]
36. Randa Ammar, Moncef Hammadi, Jean-Yves Choley, Maher Barkallah, Jamel Louati, Mohamed Haddar. The Design
and Modeling of an Optimized Mechatronic System Using a Set Based Concurrent Engineering 111-120. [Crossref]
37. M. Secanell, A. Jarauta, A. Kosakian, M. Sabharwal, J. Zhou. PEM Fuel Cells: Modeling 235-293. [Crossref]
38. Parviz Mohammad Zadeh, Mohsen Sayadi, Amirreza Kosari. 2019. An efficient metamodel-based multi-objective
multidisciplinary design optimization framework. Applied Soft Computing 74, 760-782. [Crossref]
39. Foozieh Morovat, Ali Mozaffari, Jafar Roshanian, Hadi Zare. 2019. A novel aspect of composite sandwich fairing structure
optimization of a two-stage launch vehicle (Safir) using multidisciplinary design optimization independent subspace
approach. Aerospace Science and Technology 84, 865-879. [Crossref]
40. R. El Maani, B. Radi, A. El Hami. 2019. Multiobjective backtracking search algorithm: application to FSI. Structural and
Multidisciplinary Optimization 59:1, 131-151. [Crossref]
41. Yongsheng Yi, Wei Li, Mi Xiao, Liang Gao. 2019. A set strategy approach for multidisciplinary robust design optimization
under interval uncertainty. Advances in Mechanical Engineering 11:1, 168781401882038. [Crossref]
42. Daniel Hulse, Kagan Tumer, Chris Hoyle, Irem Tumer. Modeling Collaboration in Parameter Design Using Multiagent
Learning 577-593. [Crossref]
 43. Nitin Garg, Bryce W. Pearce, Paul A. Brandner, Andrew W. Phillips, Joaquim R.R.A. Martins, Yin L. Young. 2019.
Experimental investigation of a hydrofoil designed via hydrostructural optimization. Journal of Fluids and Structures 84,
243-262. [Crossref]
44. Nickolay Jelev, Andy Keane, Carren Holden. 2019. Pattern Search Algorithm for Blackboard-Based Multidisciplinary
Design Optimization Frameworks. Journal of Aircraft 56:1, 121-136. [Abstract] [Full Text] [PDF] [PDF Plus]
45. Mads H. Aa. Madsen, Frederik Zahle, Niels N. Sørensen, Joaquim R. R. A. Martins. 2019. Multipoint high-fidelity CFD-
based aerodynamic shape optimization of a 10&thinsp;MW wind turbine. Wind Energy Science 4:2, 163-192. [Crossref]
46. François M. Hemez, Kendra L. Van Buren. Info-Gap (IG): Robust Design of a Mechanical Latch 201-222. [Crossref]
47. Justin S. Gray, Joaquim R. R. A. Martins. 2019. Coupled aeropropulsive design optimisation of a boundary-layer ingestion
propulsor. The Aeronautical Journal 123:1259, 121-137. [Crossref]
48. Stefano Caloni, Shahrokh Shahpar, Vassili Toropov. 2018. Multi-Disciplinary Design Optimisation of the Cooled Squealer
Tip for High Pressure Turbines. Aerospace 5:4, 116. [Crossref]
49. Imco van Gent, Gianfranco La Rocca, Maurice F. M. Hoogreef. 2018. CMDOWS: a proposed new standard to store and
exchange MDO systems. CEAS Aeronautical Journal 9:4, 607-627. [Crossref]
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

50. Stefania Gemma, Franco Mastroddi. 2018. Multi-disciplinary and multi-objective optimization of an over-wing-nacelle
aircraft concept. CEAS Aeronautical Journal 26. . [Crossref]
51. Hua Su, Chun-lin Gong, Liang-xian Gu. 2018. A Universal MDO Framework Based on the Adaptive Discipline Surrogate
Model. International Journal of Aerospace Engineering 2018, 1-14. [Crossref]
52. Renhe Shi, Li Liu, Teng Long, Yufei Wu, G. Gary Wang. 2018. Multidisciplinary modeling and surrogate assisted
optimization for satellite constellation systems. Structural and Multidisciplinary Optimization 58:5, 2173-2188. [Crossref]
53. Xu Chen, Peng Wang, Daiyu Zhang, Huachao Dong. 2018. Gradient-based multidisciplinary design optimization of an
autonomous underwater vehicle. Applied Ocean Research 80, 101-111. [Crossref]
54. O. Stodieck, J. E. Cooper, S. A. Neild, M. H. Lowenberg, L. Iorga. 2018. Slender-Wing Beam Reduction Method for
Gradient-Based Aeroelastic Design Optimization. AIAA Journal 56:11, 4529-4545. [Abstract] [Full Text] [PDF] [PDF
Plus]
55. Martin A. York, Berk Öztürk, Edward Burnell, Warren W. Hoburg. 2018. Efficient Aircraft Multidisciplinary Design
Optimization and Sensitivity Analysis via Signomial Programming. AIAA Journal 56:11, 4546-4561. [Abstract] [Full Text]
[PDF] [PDF Plus]
56. Kanishk Bhadani, Gauti Asbjörnsson, Erik Hulthén, Magnus Evertsson. 2018. Application of multi-disciplinary
optimization architectures in mineral processing simulations. Minerals Engineering 128, 27-35. [Crossref]
57. Luca Bonfiglio, Paris Perdikaris, Jose del Águila, George E. Karniadakis. 2018. A probabilistic framework for
multidisciplinary design: Application to the hydrostructural optimization of supercavitating hydrofoils. International
Journal for Numerical Methods in Engineering 116:4, 246-269. [Crossref]
58. David J. Munk, Timoleon Kipouros, Gareth A. Vio. 2018. Multi-physics bi-directional evolutionary topology optimization
on GPU-architecture. Engineering with Computers 20. . [Crossref]
59. Daniel R. Herber, James T. Allison. 2018. Nested and Simultaneous Solution Strategies for General Combined Plant and
Control Design Problems. Journal of Mechanical Design 141:1, 011402. [Crossref]
60. David A. Burdette, Joaquim R.R.A. Martins. 2018. Design of a transonic wing with an adaptive morphing trailing edge
via aerostructural optimization. Aerospace Science and Technology 81, 192-203. [Crossref]
61. J. D. Sinsay. 2018. Re-imagining rotorcraft advanced design. The Aeronautical Journal 122:1256, 1497-1521. [Crossref]
62. Hugo Aliaga-Aguilar, Cristina Cuerno-Rejado. 2018. Generic parameter penalty architecture. Structural and
Multidisciplinary Optimization 58:4, 1559-1569. [Crossref]
63. Siyu Tao, Kohei Shintani, Guang Yang, Herb Meingast, Daniel W. Apley, Wei Chen. 2018. Enhanced Collaborative
Optimization Using Alternating Direction Method of Multipliers. Structural and Multidisciplinary Optimization 58:4,
1571-1588. [Crossref]
64. Mohsen Bidoki, Mehdi Mortazavi, Mehdi Sabzehparvar. 2018. A new multidisciplinary robust design optimization
framework for an autonomous underwater vehicle in system and tactic design. Proceedings of the Institution of Mechanical
Engineers, Part M: Journal of Engineering for the Maritime Environment 51, 147509021879108. [Crossref]
65. Marco Berci, Rauno Cavallaro. 2018. A Hybrid Reduced-Order Model for the Aeroelastic Analysis of Flexible Subsonic
Wings—A Parametric Assessment. Aerospace 5:3, 76. [Crossref]
 66. Bin Li, Yongjie Pang, Xiaomeng Zhu, Yanxue Cheng. 2018. Collaborative optimization and 6σ design for composite
pressure hull of underwater vehicle based on lamination parameters. Journal of Marine Science and Technology 23:3, 557-566.
[Crossref]
67. J. Roshanian, A. A. Bataleblu, M. Ebrahimi. 2018. A novel evolution control strategy for surrogate-assisted design
optimization. Structural and Multidisciplinary Optimization 58:3, 1255-1273. [Crossref]
68. Shamsheer S. Chauhan, John T. Hwang, Joaquim R. R. A. Martins. 2018. An automated selection algorithm for nonlinear
solvers in MDO. Structural and Multidisciplinary Optimization 58:2, 349-377. [Crossref]
69. Benjamin J. Brelje, Joaquim R. Martins. Development of a Conceptual Design Model for Aircraft Electric Propulsion
with Efficient Gradients . [Citation] [PDF] [PDF Plus]
70. Qian Li, Rolando Burgos, Dushan Boroyevich. Hierarchical Weight Optimization Design of Aircraft Power Electronics
Systems Using Metaheuristic Optimization Methods . [Citation] [PDF] [PDF Plus]
71. Taylan G. Topcu, Bryan L. Mesmer. 2018. Incorporating end-user models and associated uncertainties to investigate
multiple stakeholder preferences in system design. Research in Engineering Design 29:3, 411-431. [Crossref]
72. Christian Alba, Ali Elham, Brian J. German, Leo L.L.M. Veldhuis. 2018. A surrogate-based multi-disciplinary design
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

optimization framework modeling wing–propeller interaction. Aerospace Science and Technology 78, 721-733. [Crossref]
73. Athanasios Papageorgiou, Mehdi Tarkian, Kristian Amadori, Johan Ölvander. 2018. Multidisciplinary Optimization of
Unmanned Aircraft Considering Radar Signature, Sensors, and Trajectory Constraints. Journal of Aircraft 55:4, 1629-1640.
[Abstract] [Full Text] [PDF] [PDF Plus]
74. Silvia Volpi, Matteo Diez, Frederick Stern. Multidisciplinary Design Optimization of a 3D Composite Hydrofoil via Variable
Accuracy Architecture . [Citation] [PDF] [PDF Plus]
75. Giulio Mose' Mancuso, Fabio Cremona, Leonardo Mangeruca, Alessandro Ulisse, Alessandro Mignogna, Stefano
Boccabella, Alberto Ferrari, Luigi Di Guglielmo. A computational interface for Steady-State Models: An FMI Extension .
[Citation] [PDF] [PDF Plus]
76. Gustavo E. Fujiwara, Nhan T. Nguyen, Eli Livne, Michael B. Bragg. Aerostructural Design Optimization of a Flexible
Wing Aircraft with Continuous Morphing Trailing Edge . [Citation] [PDF] [PDF Plus]
77. Thierry Lefebvre, Nathalie Bartoli, Sylvain Dubreuil, Marco Panzeri, Riccardo Lombardi, Wim Lammen, Mengmeng
Zhang, Imco van Gent, Pier Davide Ciampa. A clustered and surrogate-based MDA use case for MDO scenarios in AGILE
project . [Citation] [PDF] [PDF Plus]
78. Bert De Wit, Wim Lammen, Jos Vankan, Huub Timmermans, Dominique Charbonnier, Ton van der Laan, Pier Ciampa.
A collaborative design method for the aircraft supply chain: multi-level optimization . [Citation] [PDF] [PDF Plus]
79. Alp Dener, Jason E. Hicken, Gaetan K. Kenway, Joaquim Martins. Enabling Modular Aerostructural Optimization:
Individual Discipline Feasible without the Jacobians . [Citation] [PDF] [PDF Plus]
80. Imco van Gent, Benedikt Aigner, Bastiaan Beijer, Gianfranco La Rocca. A Critical Look at Design Automation Solutions
for Collaborative MDO in the AGILE Paradigm . [Citation] [PDF] [PDF Plus]
81. Ping He, Charles A. Mader, Joaquim Martins, Kevin Maki. Aerothermal Optimization of Internal Cooling Passages Using
a Discrete Adjoint Method . [Citation] [PDF] [PDF Plus]
82. Sinan Eyi, Kyle M. Hanquist, Iain D. Boyd. Aerothermodynamic Design Optimization of Hypersonic Vehicles . [Citation]
[PDF] [PDF Plus]
83. Wei Zhao, Rakesh K. Kapania. Multiobjective Optimization of Composite Flying-wings with SpaRibs and Multiple
Control Surfaces . [Citation] [PDF] [PDF Plus]
84. Michael J. Burton, Mark Drela, Christopher Courtin, Dorian Colas, Vishvas S. Suryakumar, Nicholas H. Roberts. Solar
Aircraft Design Trade Studies Using Geometric Programming . [Citation] [PDF] [PDF Plus]
85. Ana Garcia Garriga, Sangeeth S. Ponnusamy, Laura Mainini. A multi-fidelity framework to support the design of More-
Electric Actuation . [Citation] [PDF] [PDF Plus]
86. John T. Hwang, Joaquim R.R.A. Martins. 2018. A Computational Architecture for Coupling Heterogeneous Numerical
Models and Computing Coupled Derivatives. ACM Transactions on Mathematical Software 44:4, 1-39. [Crossref]
87. Ricardo Baptista, Youssef Marzouk, Karen Willcox, Benjamin Peherstorfer. 2018. Optimal Approximations of Coupling
in Multidisciplinary Models. AIAA Journal 56:6, 2412-2428. [Abstract] [Full Text] [PDF] [PDF Plus]
88. S Sanchez Perez-Moreno, K Dykes, K O Merz, M B Zaaijer. 2018. Multidisciplinary design analysis and optimisation of
a reference offshore wind plant. Journal of Physics: Conference Series 1037, 042004. [Crossref]
 89. Siva Chaitanya Chaduvula, Mikhail J. Atallah, Jitesh H. Panchal. 2018. Secure Codesign: Achieving Optimality Without
Revealing. Journal of Computing and Information Science in Engineering 18:2, 021007. [Crossref]
90. M. Cid Montoya, S. Hernández, F. Nieto. 2018. Shape optimization of streamlined decks of cable-stayed bridges considering
aeroelastic and structural constraints. Journal of Wind Engineering and Industrial Aerodynamics 177, 429-455. [Crossref]
91. M. Nosratollahi, M.A. Ghapanvary. 2018. A novel algorithm for conceptual design and optimisation of an affordable gliding
airdrop platform using TCOMOGA. The Aeronautical Journal 122:1252, 933-959. [Crossref]
92. Ping He, Charles A. Mader, Joaquim R.R.A. Martins, Kevin J. Maki. 2018. An aerodynamic design optimization
framework using a discrete adjoint approach with OpenFOAM. Computers & Fluids 168, 285-303. [Crossref]
93. Martin A. York, Warren W. Hoburg, Mark Drela. 2018. Turbofan Engine Sizing and Tradeoff Analysis via Signomial
Programming. Journal of Aircraft 55:3, 988-1003. [Abstract] [Full Text] [PDF] [PDF Plus]
94. Philippe G. Kirschen, Martin A. York, Berk Ozturk, Warren W. Hoburg. 2018. Application of Signomial Programming
to Aircraft Design. Journal of Aircraft 55:3, 965-987. [Abstract] [Full Text] [PDF] [PDF Plus]
95. Sjonnie Boonstra, Koen van der Blom, Hèrm Hofmeyer, Michael T.M. Emmerich, Jos van Schijndel, Pieter de Wilde. 2018.
Toolbox for super-structured and super-structure free multi-disciplinary building spatial design optimisation. Advanced
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

Engineering Informatics 36, 86-100. [Crossref]

96. John P. Jasa, John T. Hwang, Joaquim R. R. A. Martins. 2018. Open-source coupled aerostructural optimization using
Python. Structural and Multidisciplinary Optimization 57:4, 1815-1827. [Crossref]
97. Zhao Liu, Ping Zhu, Liwei Wang, Ching-Hung Chuang. 2018. Multidisciplinary Optimization of Auto-Body Lightweight
Design Using Hybrid Metamodeling Technique and Particle Swarm Optimizer. SAE International Journal of Materials
and Manufacturing 11:4, 373-384. [Crossref]
98. Emile Glorieux, Pasquale Franciosa, Darek Ceglarek. 2018. End-effector design optimisation and multi-robot motion
planning for handling compliant parts. Structural and Multidisciplinary Optimization 57:3, 1377-1390. [Crossref]
99. Pedro F. Albuquerque, Pedro V. Gamboa, Miguel A. Silvestre. 2018. Mission-Based Multidisciplinary Aircraft Design
Optimization Methodology Tailored for Adaptive Technologies. Journal of Aircraft 55:2, 755-770. [Abstract] [Full Text]
[PDF] [PDF Plus]
100. Demetrios Kellari, Edward F. Crawley, Bruce G. Cameron. 2018. Architectural Decisions in Commercial Aircraft from the
DC-3 to the 787. Journal of Aircraft 55:2, 792-804. [Abstract] [Full Text] [PDF] [PDF Plus]
101. M. Ripepi, M. J. Verveld, N. W. Karcher, T. Franz, M. Abu-Zurayk, S. Görtz, T. M. Kier. 2018. Reduced-order models
for aerodynamic applications, loads and MDO. CEAS Aeronautical Journal 9:1, 171-193. [Crossref]
102. Anirban Chaudhuri, John Jasa, Joaquim Martins, Karen E. Willcox. Multifidelity Optimization Under Uncertainty for a
Tailless Aircraft . [Citation] [PDF] [PDF Plus]
103. Francois Gallard, Charlie Vanaret, Damien Guenot, Vincent Gachelin, Rémi Lafage, Benoit Pauwels, Pierre-Jean Barjhoux,
Anne Gazaix. GEMS: A Python Library for Automation of Multidisciplinary Design Optimization Process Generation .
[Citation] [PDF] [PDF Plus]
104. Yogesh Bile, Atif Riaz, Marin D. Guenov, Arturo Molina-Cristobal. Towards Automating the Sizing Process in Conceptual
(Airframe) Systems Architecting . [Citation] [PDF] [PDF Plus]
105. Luis G. Bahamonde, Brian Tatting, Ramy Harik, Zafer Gurdal, Adriana Blom-Schieber, Mostafa Rassaian, Steven P.
Wanthal. An Analysis Framework for Topology Optimization of 3D Printed Reinforced Composites . [Citation] [PDF]
[PDF Plus]
106. Spyridon G. Kontogiannis, T. Kipouros, A Mark Savill. Multiobjective aerostructural optimization for efficient transport
wing conceptual design . [Citation] [PDF] [PDF Plus]
107. Giorgio Enrico Bona, Kristof Risse. A Functional Approach to a Collaborative Iintegrated Wing and Movables Design .
[Citation] [PDF] [PDF Plus]
108. Sean D. Vermillion, Richard J. Malak. A Game Theoretical Perspective on Incentivizing Collaboration in System Design
845-855. [Crossref]
109. Saeed Azad, Mohammad Behtash, Arian Houshmand, Michael Alexander-Ramos. Comprehensive PHEV Powertrain Co-
design Performance Studies Using MDSDO 83-97. [Crossref]
110. Shamsheer S. Chauhan, John T. Hwang, Joaquim R. R. A. Martins. Benchmarking Approaches for the Multidisciplinary
Analysis of Complex Systems Using a Taylor Series-Based Scalable Problem 98-116. [Crossref]
 111. Vahid Ghaffari Mejlej, Paul Falkenberg, Eiko Türck, Thomas Vietor. Multidisciplinary Structural Optimization Using of
NSGA-II and ɛ-Constraint Method in Lightweight Application 573-589. [Crossref]
112. Haitao Liu, Yew-Soon Ong, Jianfei Cai. 2018. A survey of adaptive sampling for global metamodeling in support of
simulation-based complex engineering design. Structural and Multidisciplinary Optimization 57:1, 393-416. [Crossref]
113. Anirban Chaudhuri, Remi Lam, Karen Willcox. 2018. Multifidelity Uncertainty Propagation via Adaptive Surrogates in
Coupled Multidisciplinary Systems. AIAA Journal 56:1, 235-249. [Abstract] [Full Text] [PDF] [PDF Plus]
114. Tobias Wunderlich, Lars Reimer. Integrated Process Chain for Aerostructural Wing Optimization and Application to an
NLF Forward Swept Composite Wing 3-33. [Crossref]
115. Michael Seibel. Design Procedure for Optimum Fiber Composite Airframe Structures Within an Automated
Multidisciplinary Design and Optimization Process 49-63. [Crossref]
116. Athanasios Papageorgiou, Mehdi Tarkian, Kristian Amadori, Johan Ölvander. 2018. Multidisciplinary Design Optimization
of Aerial Vehicles: A Review of Recent Advancements. International Journal of Aerospace Engineering 2018, 1-21. [Crossref]
117. Gilberto Reynoso-Meza, Helem Sabina Sánchez. 2018. Multidisciplinary optimisation and controller tuning: an analysis
with multi-objective techniques ⁎ 1This work is under the research initiative Multi-objective optimisation design
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

(MOOD) procedures for engineering systems: Industrial applications, unmanned aerial systems and mechatronic devices,
supported by the National Council of Scientific and Technological Development of Brazil (CNPq) through the grant
PQ-2/304066/2016-8 and by MINECO and FEDER through the project CICYT HARCRICS (ref.DPI2014-58104-R).
IFAC-PapersOnLine 51:4, 280-285. [Crossref]
118. Mohsen Bidoki, Mehdi Mortazavi, Mehdi Sabzehparvar. 2018. A new approach in system and tactic design optimization
of an autonomous underwater vehicle by using Multidisciplinary Design Optimization. Ocean Engineering 147, 517-530.
[Crossref]
119. Gabriel Weaver, Lavanya Marla. 2018. Cyber-Physical Simulation and Optimal Mitigation for Shipping Port Operations.
SSRN Electronic Journal . [Crossref]
120. Alparslan Emrah Bayrak, Arianne X. Collopy, Panos Y. Papalambros, Bogdan I. Epureanu. 2017. Multiobjective
optimization of modular design concepts for a collection of interacting systems. Structural and Multidisciplinary
Optimization 25. . [Crossref]
121. Tobias Wunderlich, Sascha Dähne, Lars Heinrich, Lars Reimer. 2017. Multidisciplinary optimization of an NLF forward
swept wing in combination with aeroelastic tailoring using CFRP. CEAS Aeronautical Journal 8:4, 673-690. [Crossref]
122. Bastien Talgorn, Michael Kokkolaras. 2017. Compact implementation of non-hierarchical analytical target cascading for
coordinating distributed multidisciplinary design optimization problems. Structural and Multidisciplinary Optimization
56:6, 1597-1602. [Crossref]
123. Thanh Nguyen, Snehasis Mukhopadhyay. Multidisciplinary Optimization in Decentralized Reinforcement Learning
779-784. [Crossref]
124. Gang Lei, Jianguo Zhu, Youguang Guo, Chengcheng Liu, Bo Ma. 2017. A Review of Design Optimization Methods for
Electrical Machines. Energies 10:12, 1962. [Crossref]
125. Wenqiang Yuan, Yusheng Liu, Hongwei Wang, Xiaoping Ye. 2017. A serialization-based partial decoupling approach for
multidisciplinary design optimization of complex systems. Proceedings of the Institution of Mechanical Engineers, Part B:
Journal of Engineering Manufacture 231:14, 2608-2621. [Crossref]
126. Carina Marcus, Kent Andersson, Christina Åkerlind. 2017. Balancing the radar and long wavelength infrared signature
properties in concept analysis of combat aircraft – A proof of concept. Aerospace Science and Technology 71, 733-741.
[Crossref]
127. Lili Yang, Deyu Wang. 2017. Modified collaborative optimization for feasibility problem of final solution. Structural and
Multidisciplinary Optimization 56:5, 1109-1123. [Crossref]
128. Alan Tan Wei Min, Ramon Sagarna, Abhishek Gupta, Yew-Soon Ong, Chi Keong Goh. 2017. Knowledge Transfer
Through Machine Learning in Aircraft Design. IEEE Computational Intelligence Magazine 12:4, 48-60. [Crossref]
129. Moncef Hammadi, Stanislao Patalano. Automatic generation of simulation workflows for system verification using XDSM
representation 1-6. [Crossref]
130. Xu Chen, Peng Wang, Daiyu Zhang. Surrogate-Based Multidisciplinary Design Optimization of an Autonomous
Underwater Vehicle Hull 191-194. [Crossref]
 131. Andrzej Iwaniuk, Witold Wiśniowski. 2017. Optimization of small aircraft parameters in the initial phase of the project.
Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering 231:12, 2248-2258. [Crossref]
132. Tomasz Goetzendorf-Grabowski. 2017. Multi-disciplinary optimization in aeronautical engineering. Proceedings of the
Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering 231:12, 2305-2313. [Crossref]
133. Justin Gray, Jeffrey Chin, Tristan Hearn, Eric Hendricks, Thomas Lavelle, Joaquim R. R. A. Martins. 2017. Chemical-
Equilibrium Analysis with Adjoint Derivatives for Propulsion Cycle Analysis. Journal of Propulsion and Power 33:5,
1041-1052. [Abstract] [Full Text] [PDF] [PDF Plus]
134. Salvatore Maraniello, Rafael Palacios. 2017. Optimal Rolling Maneuvers with Very Flexible Wings. AIAA Journal 55:9,
2964-2979. [Abstract] [Full Text] [PDF] [PDF Plus]
135. Bijan Mohammadi. 2017. Controlling first four moments for robust optimization. Optimization and Engineering 18:3,
561-585. [Crossref]
136. Huanwei Xu, Wei Li, Liudong Xing, Shun-Peng Zhu. 2017. Multidisciplinary design optimization under correlated
uncertainties. Concurrent Engineering 25:3, 262-275. [Crossref]
137. J. Mieloszyk, T. Goetzendorf-Grabowski. 2017. Introduction of full flight dynamic stability constraints in aircraft
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

multidisciplinary optimization. Aerospace Science and Technology 68, 252-260. [Crossref]

138. Dylan Temple, Matthew Collette. 2017. Understanding lifecycle cost trade-offs for naval vessels: minimising production,
maintenance, and resistance. Ships and Offshore Structures 12:6, 756-766. [Crossref]
139. Parviz Mohammad Zadeh, Mohadeseh Sadat Shirazi. 2017. Multidisciplinary design optimization architecture to
concurrent design of satellite systems. Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace
Engineering 231:10, 1898-1916. [Crossref]
140. Kleio Avrithi, Harrison Hyung Min Kim. 2017. Optimization of Piping Supports and Supporting Structure. Journal of
Pressure Vessel Technology 139:4, 044503. [Crossref]
141. Peter Davison, Bruce G. Cameron, Edward F. Crawley. 2017. Tradespace exploration of in-space communications network
architectures. Technology Analysis & Strategic Management 29:6, 583-599. [Crossref]
142. Jacek Mieloszyk. 2017. Practical problems of numerical optimization in aerospace sciences. Aircraft Engineering and
Aerospace Technology 89:4, 570-578. [Crossref]
143. Bin Li, Yong-jie Pang, Yan-xue Cheng, Xiao-meng Zhu. 2017. Collaborative optimization for ring-stiffened composite
pressure hull of underwater vehicle based on lamination parameters. International Journal of Naval Architecture and Ocean
Engineering 9:4, 373-381. [Crossref]
144. Jun Zou, Weixing Yao, Piao Li. 2017. Structural Optimization of Aircraft Families with Sensitivity-Based Coordination
Method. Journal of Aircraft 54:4, 1580-1586. [Citation] [Full Text] [PDF] [PDF Plus]
145. Yangyang Wang, Minghong Han. Research on multi-objective multidisciplinary design optimization based on particle
swarm optimization 1-8. [Crossref]
146. Christian Alba, Ali Elham, Brian German, Leo L. Veldhuis. A Surrogate-Based Multi-Disciplinary Design Optimization
Framework Exploiting Wing-Propeller Interaction . [Citation] [PDF] [PDF Plus]
147. Richard W. Fenrich, Juan J. Alonso. Reliable Multidisciplinary Design of a Supersonic Nozzle Using Multifidelity
Surrogates . [Citation] [PDF] [PDF Plus]
148. Imco van Gent, Pier Davide Ciampa, Benedikt Aigner, Jonas Jepsen, Gianfranco La Rocca, Joost Schut. Knowledge
architecture supporting collaborative MDO in the AGILE paradigm . [Citation] [PDF] [PDF Plus]
149. Pier Davide Ciampa, Björn Nagel. The AGILE Paradigm: the next generation of collaborative MDO . [Citation] [PDF]
[PDF Plus]
150. Pier Davide Ciampa, Erwin Moerland, Doreen Seider, Erik Baalbergen, Riccardo Lombardi, Roberto D'Ippolito. A
Collaborative Architecture supporting AGILE Design of Complex Aeronautics Products . [Citation] [PDF] [PDF Plus]
151. Charlie Vanaret, Francois Gallard, Joaquim Martins. On the Consequences of the "No Free Lunch" Theorem for
Optimization on the Choice of an Appropriate MDO Architecture . [Citation] [PDF] [PDF Plus]
152. Athanasios Papageorgiou, Johan Ölvander, Kristian Amadori. Development of a Multidisciplinary Design Optimization
Framework Applied on UAV Design by Considering Models for Mission, Surveillance, and Stealth Performance . [Citation]
[PDF] [PDF Plus]
153. Imco van Gent, Gianfranco La Rocca, Leo L. Veldhuis. Composing MDAO symphonies: graph-based generation and
manipulation of large multidisciplinary systems . [Citation] [PDF] [PDF Plus]
 154. Nicolas Bons, Xiaolong He, Charles A. Mader, Joaquim Martins. Multimodality in Aerodynamic Wing Design
Optimization . [Citation] [PDF] [PDF Plus]
155. Erwin Moerland, Till Pfeiffer, Daniel Böhnke, Jonas Jepsen, Sebastian Freund, Carsten M. Liersch, Gabriel Pinho
Chiozzotto, Carsten Klein, Julian Scherer, Yasim J. Hasan, Jan Flink. On the Design of a Strut-Braced Wing Configuration
in a Collaborative Design Environment . [Citation] [PDF] [PDF Plus]
156. Jan E.K. Hoogervorst, Ali Elham. 2017. Wing aerostructural optimization using the Individual Discipline Feasible
Architecture. Aerospace Science and Technology 65, 90-99. [Crossref]
157. Z. Zhu, G. La Rocca, M. J. L. van Tooren. 2017. A methodology to enable automatic 3D routing of aircraft Electrical
Wiring Interconnection System. CEAS Aeronautical Journal 8:2, 287-302. [Crossref]
158. Gaetan K. W. Kenway, Joaquim R. R. A. Martins. 2017. Buffet-Onset Constraint Formulation for Aerodynamic Shape
Optimization. AIAA Journal 55:6, 1930-1947. [Abstract] [Full Text] [PDF] [PDF Plus]
159. Jung-Sun Choi, Gyung-Jin Park. 2017. Multidisciplinary design optimization of the flapping wing system for forward
flight. International Journal of Micro Air Vehicles 9:2, 93-110. [Crossref]
160. A. Aranake, K. Duraisamy. 2017. Aerodynamic optimization of shrouded wind turbines. Wind Energy 20:5, 877-889.
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

[Crossref]
161. Tristan Dhert, Turaj Ashuri, Joaquim R. R. A. Martins. 2017. Aerodynamic shape optimization of wind turbine blades
using a Reynolds-averaged Navier-Stokes model and an adjoint method. Wind Energy 20:5, 909-926. [Crossref]
162. Luis Bahamonde Jacome, Ali Elham. 2017. Wing Aerostructural Optimization Under Uncertain Aircraft Range and
Payload Weight. Journal of Aircraft 54:3, 1109-1120. [Abstract] [Full Text] [PDF] [PDF Plus]
163. Daiyu Zhang, Baowei Song, Peng Wang. 2017. A parallel bi-level multidisciplinary design optimization architecture with
convergence proof for general problem. Engineering Optimization 49:4, 654-674. [Crossref]
164. A. Esmaeili, J.M.M. Sousa. 2017. Power density ratio optimization of bimorph piezocomposite energy harvesters using a
Multidisciplinary Design Feasible method. Composite Structures 165, 171-179. [Crossref]
165. Xiling Yao, Seung Ki Moon, Guijun Bi. 2017. Multidisciplinary design optimization to identify additive manufacturing
resources in customized product development. Journal of Computational Design and Engineering 4:2, 131-142. [Crossref]
166. Daiyu Zhang, Baowei Song, Peng Wang, Xu Chen. 2017. Multidisciplinary optimization design of a new underwater vehicle
with highly efficient gradient calculation. Structural and Multidisciplinary Optimization 55:4, 1483-1502. [Crossref]
167. Samee Ur Rehman, Matthijs Langelaar. 2017. Adaptive efficient global optimization of systems with independent
components. Structural and Multidisciplinary Optimization 55:4, 1143-1157. [Crossref]
168. Maxime Moret, Alexandre Delecourt, Hany Moustapha, Acher-Igal Abenhaim, Francois Garnier. 2017. Automated
thermal and stress preliminary analyses applied to a turbine rotor. Aerospace Science and Technology 63, 123-131. [Crossref]
169. Austin J. Herrema, Nelson M. Wiese, Carolyn N. Darling, Baskar Ganapathysubramanian, Adarsh Krishnamurthy, Ming-
Chen Hsu. 2017. A framework for parametric design optimization using isogeometric analysis. Computer Methods in Applied
Mechanics and Engineering 316, 944-965. [Crossref]
170. Raghu Chaitanya Munjulury, Hossein Nadali Najafabadi, Edris Safavi, Johan Ölvander, Petter Krus, Matts Karlsson. 2017.
A comprehensive computational multidisciplinary design optimization approach for a tidal power plant turbine. Advances
in Mechanical Engineering 9:3, 168781401769517. [Crossref]
171. Timothy W. Simpson, Simon Miller, Elliott B. Tibor, Michael A. Yukish, Gary Stump, Hanumanthrao Kannan, Bryan
Mesmer, Eliot H. Winer, Christina L. Bloebaum. 2017. Adding Value to Trade Space Exploration When Designing
Complex Engineered Systems. Systems Engineering 20:2, 131-146. [Crossref]
172. L. Swartjes, L.F.P. Etman, J.M. van de Mortel-Fronczak, J.E. Rooda, L.J.A.M. Somers. 2017. Simultaneous analysis and
design based optimization for paper path and timing design of a high-volume printer. Mechatronics 41, 82-89. [Crossref]
173. Xiaolong He, Junqiang Bai, Liang Xin. Multidisciplinary optimization of subsonic wing planform using free form
deformation 205-209. [Crossref]
174. Ali Elham, Michael J. Van Tooren. Toward Wing Aerostructural Optimization Using Simultaneous Analysis and Design
Strategy . [Citation] [PDF] [PDF Plus]
175. Joan Mas Colomer, Nathalie Bartoli, Thierry Lefebvre, Sylvain Dubreuil, Joaquim Martins, Emmanuel Benard, Joseph
Morlier. Similarity Maximization of a Scaled Aeroelastic Flight Demonstrator via Multidisciplinary Optimization .
[Citation] [PDF] [PDF Plus]
 176. Julius M. Vegh, Timothy MacDonald, Andrew Wendorff, Juan J. Alonso. Sizing Methods for Aircraft of Variable
Propulsion System Complexity . [Citation] [PDF] [PDF Plus]
177. Harry H. Hilton, Steven J. D’Urso, Noe Wiener. Designer Systems of Systems: A Rational Integrated Approach of System
Engineering to Tailored Aerodynamics, Aeroelasticity, Aero-viscoelasticity, Stability, Control, Geometry, Materials,
Structures, Propulsion, Performance, Sizing, Weight, Cost 49-84. [Crossref]
178. Gilberto Reynoso Meza, Xavier Blasco Ferragud, Javier Sanchis Saez, Juan Manuel Herrero Durá. Background on
Multiobjective Optimization for Controller Tuning 23-58. [Crossref]
179. Gabriella Bolzon. 2017. Complementarity Problems in Structural Engineering: An Overview. Archives of Computational
Methods in Engineering 24:1, 23-36. [Crossref]
180. Jonathan Ollar, Vassili Toropov, Royston Jones. 2017. Sub-space approximations for MDO problems with disparate
disciplinary variable dependence. Structural and Multidisciplinary Optimization 55:1, 279-288. [Crossref]
181. Daiyu Zhang, Baowei Song, Peng Wang, Yanru He. 2017. Performance Evaluation of MDO Architectures within a Variable
Complexity Problem. Mathematical Problems in Engineering 2017, 1-9. [Crossref]
182. Melih Cakmakci, Gullu Kiziltas Sendur, Umut Durak. Simulation-Based Engineering 39-73. [Crossref]
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

183. Satoshi YAMAUCHI, Shintaro YAMASAKI, Kentaro YAJI, Kikuo FUJITA. 2017. Multidisciplinary layout design
optimization method using multi-fidelity analysis model inspired by the explicit method (Application to layout design
problem considering a heat dissipation characteristic). Transactions of the JSME (in Japanese) 83:855, 17-00320-17-00320.
[Crossref]
184. Rong Yuan, Haiqing Li. 2017. A multidisciplinary coupling relationship coordination algorithm using the hierarchical
control methods of complex systems and its application in multidisciplinary design optimization. Advances in Mechanical
Engineering 9:1, 168781401668522. [Crossref]
185. Qian Yin, Yizhong Wu, Boxing Wang. 2017. A Method for Multidisciplinary System Analysis Based on Minimal Feedback
Variables. Mathematical Problems in Engineering 2017, 1. [Crossref]
186. Hanumanthrao Kannan, Bryan L. Mesmer, Christina L. Bloebaum. 2017. Increased System Consistency through
Incorporation of Coupling in Value-Based Systems Engineering. Systems Engineering 20:1, 21-44. [Crossref]
187. M. Secanell, A. Jarauta, A. Kosakian, M. Sabharwal, J. Zhou. PEM Fuel Cells, Modeling 1-61. [Crossref]
188. Asitav Mishra, Dimitri Mavriplis, Jay Sitaraman. 2016. Time-Dependent Aeroelastic Adjoint-Based Aerodynamic Shape
Optimization of Helicopter Rotors in Forward Flight. AIAA Journal 54:12, 3813-3827. [Abstract] [Full Text] [PDF] [PDF
Plus]
189. Johannes Gross, Stephan Rudolph. 2016. Rule-based spacecraft design space exploration and sensitivity analysis. Aerospace
Science and Technology 59, 162-171. [Crossref]
190. Craig Bakker, Geoffrey T. Parks. 2016. Riemannian optimization and multidisciplinary design optimization. Optimization
and Engineering 17:4, 663-693. [Crossref]
191. Qiang Liu, Mohamed Jrad, Sameer B. Mulani, Rakesh K. Kapania. 2016. Global/Local Optimization of Aircraft Wing
Using Parallel Processing. AIAA Journal 54:11, 3338-3348. [Abstract] [Full Text] [PDF] [PDF Plus]
192. T. Ashuri, J. R. R. A. Martins, M. B. Zaaijer, G. A. M. van Kuik, G. J. W. van Bussel. 2016. Aeroservoelastic design
definition of a 20 MW common research wind turbine model. Wind Energy 19:11, 2071-2087. [Crossref]
193. S. Dubreuil, N. Bartoli, C. Gogu, T. Lefebvre. 2016. Propagation of Modeling Uncertainty by Polynomial Chaos Expansion
in Multidisciplinary Analysis. Journal of Mechanical Design 138:11, 111411. [Crossref]
194. Tingting Xia, Mian Li, Jianhua Zhou. 2016. A Sequential Robust Optimization Approach for Multidisciplinary Design
Optimization With Uncertainty. Journal of Mechanical Design 138:11, 111406. [Crossref]
195. Ali Elham, Michel J. L. van Tooren. 2016. Coupled adjoint aerostructural wing optimization using quasi-three-
dimensional aerodynamic analysis. Structural and Multidisciplinary Optimization 54:4, 889-906. [Crossref]
196. Debiao Meng, Hua Zhang, Tao Huang. 2016. A concurrent reliability optimization procedure in the earlier design phases
of complex engineering systems under epistemic uncertainties. Advances in Mechanical Engineering 8:10, 168781401667397.
[Crossref]
197. Terence Macquart. Future research in multi-step composite optimisation 1-5. [Crossref]
198. Parviz Mohammad Zadeh, Ali Mehmani, Achille Messac. 2016. High fidelity multidisciplinary design optimization of a
wing using the interaction of low and high fidelity models. Optimization and Engineering 17:3, 503-532. [Crossref]
 199. S. Maraniello, R. Palacios. 2016. Optimal vibration control and co-design of very flexible actuated structures. Journal of
Sound and Vibration 377, 1-21. [Crossref]
200. D. B. Schwinn, D. Kohlgrüber, J. Scherer, M. H. Siemann. 2016. A parametric aircraft fuselage model for preliminary
sizing and crashworthiness applications. CEAS Aeronautical Journal 7:3, 357-372. [Crossref]
201. S Sanchez Perez-Moreno, M B Zaaijer, C L Bottasso, K Dykes, K O Merz, P-E Réthoré, F Zahle. 2016. Roadmap to
the multidisciplinary design analysis and optimisation of wind energy systems. Journal of Physics: Conference Series 753,
062011. [Crossref]
202. T Ashuri, C Ponnurangam, J Zhang, M Rotea. 2016. Integrated Layout and Support Structure Optimization for Offshore
Wind Farm Design. Journal of Physics: Conference Series 753, 092011. [Crossref]
203. Cecilia Leotardi, Andrea Serani, Umberto Iemma, Emilio F. Campana, Matteo Diez. 2016. A variable-accuracy metamodel-
based architecture for global MDO under uncertainty. Structural and Multidisciplinary Optimization 54:3, 573-593.
[Crossref]
204. Graeme J. Kennedy. 2016. A full-space barrier method for stress-constrained discrete material design optimization.
Structural and Multidisciplinary Optimization 54:3, 619-639. [Crossref]
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

205. Turaj Ashuri, Michiel B. Zaaijer, Joaquim R.R.A. Martins, Jie Zhang. 2016. Multidisciplinary design optimization of large
wind turbines—Technical, economic, and design challenges. Energy Conversion and Management 123, 56-70. [Crossref]
206. Edris Safavi, Mehdi Tarkian, Johan Ölvander, Hossein Nadali Najafabadi, Raghu Chaitanya Munjulury. 2016.
Implementation of collaborative multidisciplinary design optimization for conceptual design of a complex engineering
product. Concurrent Engineering 24:3, 251-265. [Crossref]
207. Nicolas Morales, Dinesh Manocha. 2016. Efficient wave-based acoustic material design optimization. Computer-Aided
Design 78, 83-92. [Crossref]
208. Daniel Rohacs, Jozsef Rohacs. 2016. Magnetic levitation assisted aircraft take-off and landing (feasibility study – GABRIEL
concept). Progress in Aerospace Sciences 85, 33-50. [Crossref]
209. Marco Caboni, M. Sergio Campobasso, Edmondo Minisci. 2016. Wind Turbine Design Optimization Under
Environmental Uncertainty. Journal of Engineering for Gas Turbines and Power 138:8, 082601. [Crossref]
210. Bo Yang Yu, Tomonori Honda, Syed M. Zubair, Mostafa H. Sharqawy, Maria C. Yang. 2016. A maintenance-focused
approach to complex system design. Artificial Intelligence for Engineering Design, Analysis and Manufacturing 30:3, 263-276.
[Crossref]
211. Jun Zou, Wei-Xing Yao, Tian-Xiang Xia. 2016. A sensitivity-based coordination method for optimization of product
families. Engineering Optimization 48:7, 1145-1163. [Crossref]
212. Paul Reverdy. Gaussian multi-armed bandit problems with multiple objectives 5263-5269. [Crossref]
213. Sayan Ghosh, Dimitri N. Mavris. A Methodology for Probabilistic Analysis of Distributed Multidisciplinary Architecture
(PADMA) . [Citation] [PDF] [PDF Plus]
214. Andrea Viti, Thierry Druot, Antoine Dumont. Aero-structural approach coupled with direct operative cost optimization
for new aircraft concept in preliminary design . [Citation] [PDF] [PDF Plus]
215. Jan E. Hoogervorst, Ali Elham. Wing Aerostructural Optimization Using the Individual Discipline Feasible Strategy .
[Citation] [PDF] [PDF Plus]
216. Jichao Li, Kun Qu, Jinsheng Cai, Changqiang Cao. Adjoint Approach based on Reduced-order Model for Steady PDE
Systems . [Citation] [PDF] [PDF Plus]
217. Robert D. Falck, Justin S. Gray, Bret Naylor. Parallel Aircraft Trajectory Optimization with Analytic Derivatives .
[Citation] [PDF] [PDF Plus]
218. B. Talgorn, M. Kokkolaras, A. DeBlois, P. Piperni. 2016. Numerical investigation of non-hierarchical coordination
for distributed multidisciplinary design optimization with fixed computational budget. Structural and Multidisciplinary
Optimization . [Crossref]
219. Sylvain Arreckx, Andrew Lambe, Joaquim R. R. A. Martins, Dominique Orban. 2016. A matrix-free augmented lagrangian
algorithm with application to large-scale structural design optimization. Optimization and Engineering 17:2, 359-384.
[Crossref]
220. Dylan Temple, Matthew Colette. 2016. A goal-programming enhanced collaborative optimization approach to reducing
lifecyle costs for naval vessels. Structural and Multidisciplinary Optimization 53:6, 1261-1275. [Crossref]
 221. E. C. M. de Borst, L. F. P. Etman, A. W. J. Gielen, A. T. Hofkamp, J. E. Rooda. 2016. Decomposition analysis of the
multidisciplinary coupling in LED System-in-Package design using a DSM and a specification language. Structural and
Multidisciplinary Optimization 53:6, 1395-1411. [Crossref]
222. Moncef Hammadi, Amir Guizani, Jean-Yves Choley, Andreas Kellner, Peter Hehenberger. 2016. An agent-supported
approach for the collaborative design of complex systems. International Journal of Modeling, Simulation, and Scientific
Computing 07:02, 1642001. [Crossref]
223. Oliviu Şugar Gabor, Andreea Koreanschi, Ruxandra Mihaela Botez, Mahmoud Mamou, Youssef Mebarki. 2016. Numerical
simulation and wind tunnel tests investigation and validation of a morphing wing-tip demonstrator aerodynamic
performance. Aerospace Science and Technology 53, 136-153. [Crossref]
224. Philip Odonkor, Kemper Lewis, Jin Wen, Teresa Wu. 2016. Adaptive Energy Optimization Toward Net-Zero Energy
Building Clusters. Journal of Mechanical Design 138:6, 061405. [Crossref]
225. Joaquim R.R.A Martins. Fuel Burn Reduction Through Wing Morphing 1-7. [Crossref]
226. J. N. Stander, G. Venter, M. J. Kamper. 2016. High fidelity multidisciplinary design optimisation of an electromagnetic
device. Structural and Multidisciplinary Optimization 53:5, 1113-1127. [Crossref]
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

227. Shreyas Vathul Subramanian, Daniel A. DeLaurentis. 2016. Application of Multidisciplinary Systems-of-Systems
Optimization to an Aircraft Design Problem. Systems Engineering 19:3, 235-251. [Crossref]
228. Enrico Ampellio, Francesco Bertini, Andrea Ferrero, Francesco Larocca, Luca Vassio. 2016. Turbomachinery design by
a swarm-based optimization method coupled with a CFD solver. Advances in aircraft and spacecraft science 3:2, 149-170.
[Crossref]
229. Paolo Guarneri, Margaret M. Wiecek. 2016. Pareto-based negotiation in distributed multidisciplinary design. Structural
and Multidisciplinary Optimization 53:4, 657-671. [Crossref]
230. Andrew Gilbert, Bryan Mesmer, Michael Watson. Exergy based optimization of rocket system staging times 1-7. [Crossref]
231. Björn Persson, Peter Bull. 2016. Empirical Study of Flight-Dynamic Influences on Radar Cross-Section Models. Journal
of Aircraft 53:2, 463-474. [Abstract] [Full Text] [PDF] [PDF Plus]
232. Andrew B. Lambe, Joaquim R. R. A. Martins. 2016. Matrix-free aerostructural optimization of aircraft wings. Structural
and Multidisciplinary Optimization 53:3, 589-603. [Crossref]
233. Debiao Meng, Xiaoling Zhang, Yuan-Jian Yang, Huanwei Xu, Hong-Zhong Huang. 2016. Interaction balance optimization
in multidisciplinary design optimization problems. Concurrent Engineering 24:1, 48-57. [Crossref]
234. Kybeom Kwon, Gregory Lantoine, Ryan P. Russell, Dimitri N. Mavris. 2016. A study on simultaneous design of a Hall
Effect Thruster and its low-thrust trajectory. Acta Astronautica 119, 34-47. [Crossref]
235. M.J. Montazeri, R. Ebrahimi. 2016. Multidisciplinary optimization of a pump–fed system in a cryogenic LPE using a
systematic approach based on genetic algorithm. Aerospace Science and Technology 49, 185-196. [Crossref]
236. Anirban Chaudhuri, Karen E. Willcox. Multifidelity Uncertainty Propagation in Coupled Multidisciplinary Systems .
[Citation] [PDF] [PDF Plus]
237. Gabrielle E. Wroblewski, Phillip J. Ansell. Prediction and Experimental Evaluation of Planar Wing Spanloads for
Minimum Drag . [Citation] [PDF] [PDF Plus]
238. Philippe G. Kirschen, Edward Burnell, Warren Hoburg. Signomial Programming Models for Aircraft Design . [Citation]
[PDF] [PDF Plus]
239. Tim Carroll, Issi-Rae E. George, Goetz Bramesfeld. Design Optimization of Small Rotors in Quad-Rotor Configuration .
[Citation] [PDF] [PDF Plus]
240. Ali Elham, Luis Bahamonde Jacome. Wing Aerostructural Optimization under Uncertain Payload Weight and Aircraft
Range . [Citation] [PDF] [PDF Plus]
241. Loïc Brevault, Mathieu Balesdent, Nicolas Bérend, Rodolphe Le Riche. 2016. Decoupled Multidisciplinary Design
Optimization Formulation for Interdisciplinary Coupling Satisfaction Under Uncertainty. AIAA Journal 54:1, 186-205.
[Abstract] [Full Text] [PDF] [PDF Plus]
242. Xuan Sun, Ulf Sellgren, Kjell Andersson. 2016. Situated Design Optimization of Haptic Devices. Procedia CIRP 50,
293-298. [Crossref]
243. M. Collu, M. Borg. Design of floating offshore wind turbines 359-385. [Crossref]
 244. Zhaoke Xu, Jian Xia. 2016. Aerodynamic Optimization Based on Continuous Adjoint Method for a Flexible Wing.
International Journal of Aerospace Engineering 2016, 1-16. [Crossref]
245. Xin-Jia Meng, Shi-Kai Jing, Li-Xiang Zhang, Ji-Hong Liu, Hai-Cheng Yang. 2016. Linear approximation filter strategy
for collaborative optimization with combination of linear approximations. Structural and Multidisciplinary Optimization
53:1, 49-66. [Crossref]
246. G. Bernardini, E. Piccione, A. Anobile, J. Serafini, M. Gennaretti. 2016. Optimal Design and Acoustic Assessment of
Low-Vibration Rotor Blades. International Journal of Rotating Machinery 2016, 1-17. [Crossref]
247. Koen van der Blom, Sjonnie Boonstra, Hèrm Hofmeyer, Michael T. M. Emmerich. Multicriteria Building Spatial Design
with Mixed Integer Evolutionary Algorithms 453-462. [Crossref]
248. Andrey Savelyev, Kirill Anisimov, Egor Kazhan, Innocentiy Kursakov, Alexandr Lysenkov. Computational study of engine
external aerodynamics as a part of multidisciplinary optimization procedure 020018. [Crossref]
249. Anand P. Deshmukh, James T. Allison. 2016. Multidisciplinary dynamic optimization of horizontal axis wind turbine
design. Structural and Multidisciplinary Optimization 53:1, 15-27. [Crossref]
250. Mathieu Balesdent, Loïc Brevault, Nathaniel B. Price, Sébastien Defoort, Rodolphe Le Riche, Nam-Ho Kim, Raphael
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

T. Haftka, Nicolas Bérend. Advanced Space Vehicle Design Taking into Account Multidisciplinary Couplings and Mixed
Epistemic/Aleatory Uncertainties 1-48. [Crossref]
251. Emilia Silvas, Theo Hofman, Nikolce Murgovski, Pascal Etman, Maarten Steinbuch. 2016. Review of Optimization
Strategies for System-Level Design in Hybrid Electric Vehicles. IEEE Transactions on Vehicular Technology 1-1. [Crossref]
252. . Multidisciplinary Design Optimization Architectures 155-207. [Crossref]
253. W. Li, Y Wen, L X Li. 2015. A collaborative optimization framework for parametric and parameter-free variables.
Engineering Computations 32:8, 2491-2503. [Crossref]
254. Ann-Britt Ryberg, Rebecka Domeij Bäckryd, Larsgunnar Nilsson. 2015. A metamodel-based multidisciplinary design
optimization process for automotive structures. Engineering with Computers 31:4, 711-728. [Crossref]
255. Fengnian Tian, Mark Voskuijl. 2015. Automated generation of multiphysics simulation models to support multidisciplinary
design optimization. Advanced Engineering Informatics 29:4, 1110-1125. [Crossref]
256. Erwin Moerland, Richard-Gregor Becker, Björn Nagel. 2015. Collaborative understanding of disciplinary correlations
using a low-fidelity physics-based aerospace toolkit. CEAS Aeronautical Journal 6:3, 441-454. [Crossref]
257. Tobias F. Wunderlich. 2015. Multidisciplinary wing optimization of commercial aircraft with consideration of static
aeroelasticity. CEAS Aeronautical Journal 6:3, 407-427. [Crossref]
258. Patrick L Cross, Mike Mulford. 2015. Realizing collaborative systems design for missile seekers by combining design
margin analysis with multi-disciplinary optimization. Concurrent Engineering 23:3, 226-235. [Crossref]
259. Yuchen Bai, Robert Farquhar, Hongcheng Xu, Xilun Ding. A Conceptual Design of a Comet Explorer Performing both
Penetrating and Surface Roving Missions . [Citation] [PDF] [PDF Plus]
260. Weilin Luo, Wenjing Lyu. 2015. An application of multidisciplinary design optimization to the hydrodynamic
performances of underwater robots. Ocean Engineering 104, 686-697. [Crossref]
261. Hyeong-Uk Park, Jae-Woo Lee, Joon Chung, Kamran Behdinan. 2015. Uncertainty-based MDO for aircraft conceptual
design. Aircraft Engineering and Aerospace Technology 87:4, 345-356. [Crossref]
262. Craig Bakker, Geoffrey T. Parks. 2015. Differential geometry tools for multidisciplinary design optimization, part II:
application to QSD. Structural and Multidisciplinary Optimization 52:1, 39-53. [Crossref]
263. Craig Bakker, Geoffrey T. Parks. 2015. Differential geometry tools for multidisciplinary design optimization, Part I:
Theory. Structural and Multidisciplinary Optimization 52:1, 27-38. [Crossref]
264. Asitav Mishra, Karthik Mani, Dimitri Mavriplis, Jay Sitaraman. 2015. Time dependent adjoint-based optimization for
coupled fluid–structure problems. Journal of Computational Physics 292, 253-271. [Crossref]
265. Maurice Hoogreef, Gianfranco La Rocca. An MDO advisory system supported by knowledge-based technologies .
[Citation] [PDF] [PDF Plus]
266. Stefania Gemma, Franco Mastroddi. Multi-Disciplinary and Multi-Objective Optimization of an Unconventional Aircraft
Concept . [Citation] [PDF] [PDF Plus]
267. Peter Schmollgruber, Nathalie Bartoli, Yves Gourinat. Virtual flight testing in an aircraft sizing and optimization process .
[Citation] [PDF] [PDF Plus]
 268. Pedro F. Albuquerque, Pedro V. Gamboa, Miguel �. Silvestre. Multidisciplinary and Multilevel Aircraft Design
Methodology Using Enhanced Collaborative Optimization . [Citation] [PDF] [PDF Plus]
269. Min Zhao, Wei-cheng Cui, Xiang Li. 2015. Multidisciplinary design optimization of a human occupied vehicle based on
Bi-Level integrated system collaborative optimization. China Ocean Engineering 29:4, 599-610. [Crossref]
270. Rhea P. Liem, Charles A. Mader, Joaquim R.R.A. Martins. 2015. Surrogate models and mixtures of experts in aerodynamic
performance prediction for aircraft mission analysis. Aerospace Science and Technology 43, 126-151. [Crossref]
271. Bijan Mohammadi. 2015. Ensemble Kalman Filters and geometric characterization of sensitivity spaces for uncertainty
quantification in optimization. Computer Methods in Applied Mechanics and Engineering 290, 228-249. [Crossref]
272. . Bibliography 299-308. [Crossref]
273. Tomonori Honda, Francesco Ciucci, Kemper E. Lewis, Maria C. Yang. 2015. Comparison of Information Passing Strategies
in System-Level Modeling. AIAA Journal 53:5, 1121-1133. [Abstract] [Full Text] [PDF] [PDF Plus]
274. Ding Yang, Michela Turrin, Sevil Sariyildiz, Yimin Sun. Sports building envelope optimization using multi-objective
multidisciplinary design optimization (M-MDO) techniques: Case of indoor sports building project in China 2269-2278.
[Crossref]
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

275. Ana Paula Curty Cuco, Fabiano L. de Sousa, Antônio J. Silva Neto. 2015. A multi-objective methodology for spacecraft
equipment layouts. Optimization and Engineering 16:1, 165-181. [Crossref]
276. Darcy L. Allison, Craig C. Morris, Joseph A. Schetz, Rakesh K. Kapania, Layne T. Watson, Joshua D. Deaton. 2015.
Development of a multidisciplinary design optimization framework for an efficient supersonic air vehicle. Advances in
aircraft and spacecraft science 2:1, 17-44. [Crossref]
277. Henric Pettersson, Maria Rodiouchkina, Gerald Micklow, Hamid Hefazi. 2015. Development of Synthesis Level Design
Model in Automobile Application Suitable for MDO using CO Approach. SAE International Journal of Materials and
Manufacturing 8:2, 344-356. [Crossref]
278. Qiang Liu, Mohamed Jrad, Sameer B. Mulani, Rakesh K. Kapania. Integrated Global Wing and Local Panel Optimization
of Aircraft Wing . [Citation] [PDF] [PDF Plus]
279. Asitav Mishra, Karthik Mani, Dimitri J. Mavriplis, Jayanarayanan Sitaraman. Time-dependent Aero-elastic Adjoint-based
Aerodynamic Shape Optimization of Helicopter Rotors in Forward Flight . [Citation] [PDF] [PDF Plus]
280. John Hwang, Satadru Roy, Jason Kao, Joaquim Martins, William A. Crossley. Simultaneous aircraft allocation and mission
optimization using a modular adjoint approach . [Citation] [PDF] [PDF Plus]
281. Rhea P. Liem, Gaetan K. W. Kenway, Joaquim R. R. A. Martins. 2015. Multimission Aircraft Fuel-Burn Minimization
via Multipoint Aerostructural Optimization. AIAA Journal 53:1, 104-122. [Abstract] [Full Text] [PDF] [PDF Plus]
282. Yutaka NOMAGUCHI, Hiroyuki NAKAYAMA, Haruki INOUE, Koutaro TODA, Kikuo FUJITA. 2015. Analysis of
empirical knowledge acquisition process in optimal design of multi-disciplinary system. Transactions of the JSME (in
Japanese) 81:830, 15-00244-15-00244. [Crossref]
283. Peter Davison, Bruce Cameron, Edward F. Crawley. 2015. Technology Portfolio Planning by Weighted Graph Analysis of
System Architectures. Systems Engineering 18:1, 45-58. [Crossref]
284. Alexey M. Nazarenko, Alexander A. Prokhorov. 2015. Hierarchical Dataflow Model with Automated File Management
for Engineering and Scientific Applications. Procedia Computer Science 66, 496-505. [Crossref]
285. Johan A. Persson, Xiaolong Feng, Daniel Wappling, Johan Ölvander. 2015. A Framework for Multidisciplinary
Optimization of a Balancing Mechanism for an Industrial Robot. Journal of Robotics 2015, 1-8. [Crossref]
286. Cees Bil. Multidisciplinary Design Optimization: Designed by Computer 421-454. [Crossref]
287. Zhenguo Wang, Wei Huang, Li Yan. 2014. Multidisciplinary design optimization approach and its application to aerospace
engineering. Chinese Science Bulletin 59:36, 5338-5353. [Crossref]
288. Warren Hoburg, Pieter Abbeel. 2014. Geometric Programming for Aircraft Design Optimization. AIAA Journal 52:11,
2414-2426. [Abstract] [Full Text] [PDF] [PDF Plus]
289. Wenrui Duan, Ling Tian, Zhansong Wang, Wenbin Han. 2014. Multidisciplinary optimization of etch process chamber
on the basis of MCDM. Journal of Mechanical Science and Technology 28:11, 4621-4633. [Crossref]
290. Sourobh Ghosh, Erich Devendorf, Kemper Lewis. 2014. Exploring the effectiveness of parallel systems in distributed
design processes subjected to stochastic disruptions. Artificial Intelligence for Engineering Design, Analysis and Manufacturing
28:4, 399-412. [Crossref]
 291. Jessica L. Chapman, Lu Lu, Christine M. Anderson-Cook. 2014. Incorporating response variability and estimation
uncertainty into Pareto front optimization. Computers & Industrial Engineering 76, 253-267. [Crossref]
292. Bijan Mohammadi. 2014. Uncertainty quantification by geometric characterization of sensitivity spaces. Computer Methods
in Applied Mechanics and Engineering 280, 197-221. [Crossref]
293. Christopher Perullo, Dimitri Mavris. 2014. A review of hybrid-electric energy management and its inclusion in vehicle
sizing. Aircraft Engineering and Aerospace Technology 86:6, 550-557. [Crossref]
294. Zhoujie Lyu, Joaquim R. R. A. Martins. 2014. Aerodynamic Design Optimization Studies of a Blended-Wing-Body
Aircraft. Journal of Aircraft 51:5, 1604-1617. [Abstract] [Full Text] [PDF] [PDF Plus]
295. Jia Xu, Ilan Kroo. 2014. Aircraft Design with Active Load Alleviation and Natural Laminar Flow. Journal of Aircraft 51:5,
1532-1545. [Abstract] [Full Text] [PDF] [PDF Plus]
296. John T. Hwang, Dae Young Lee, James W. Cutler, Joaquim R. R. A. Martins. 2014. Large-Scale Multidisciplinary
Optimization of a Small Satellite’s Design and Operation. Journal of Spacecraft and Rockets 51:5, 1648-1663. [Abstract]
[Full Text] [PDF] [PDF Plus]
297. Yuchen Bai, Robert Farquhar, Hongcheng Xu, Xilun Ding. A Conceptual Design of a Comet Explorer Performing both
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

Penetrating and Surface Roving Missions . [Citation] [PDF] [PDF Plus]

298. T. Ashuri, M.B. Zaaijer, J.R.R.A. Martins, G.J.W. van Bussel, G.A.M. van Kuik. 2014. Multidisciplinary design
optimization of offshore wind turbines for minimum levelized cost of energy. Renewable Energy 68, 893-905. [Crossref]
299. Pero Prebeg, Vedran Zanic, Bozo Vazic. 2014. Application of a surrogate modeling to the ship structural design. Ocean
Engineering 84, 259-272. [Crossref]
300. Gilberto Reynoso-Meza, Xavier Blasco, Javier Sanchis, Miguel Martínez. 2014. Controller tuning using evolutionary multi-
objective optimisation: Current trends and applications. Control Engineering Practice 28, 58-73. [Crossref]
301. Xiaojia Zhao, Richard Curran, Wim J. C. Verhagen. Aircraft Component Multidisciplinary Design Optimization
Considering Cost Performance . [Citation] [PDF] [PDF Plus]
302. Derek J. Dalle, Krzysztof Fidkowski. Multifidelity Airfoil Shape Optimization Using Adaptive Meshing . [Citation] [PDF]
[PDF Plus]
303. Paul B. Reverdy, Akhil S. Reddy, Luigi Martinelli, Naomi E. Leonard. Integrating a human designer's preferences in
multidisciplinary design optimization . [Citation] [PDF] [PDF Plus]
304. Loic Brevault, Mathieu Balesdent, Nicolas Berend, Rodolphe Le Riche. Decoupled UMDO formulation for
interdisciplinary coupling satisfaction under uncertainty . [Citation] [PDF] [PDF Plus]
305. Andrew Lambe, Graeme Kennedy, Joaquim Martins. Multidisciplinary Design Optimization of an Aircraft Wing via a
Matrix-Free Approach . [Citation] [PDF] [PDF Plus]
306. Ryan P. Starkey. MDO for Hypersonic Scramjet Vehicle Development (Invited) . [Citation] [PDF] [PDF Plus]
307. Justin S. Gray, Tristan A. Hearn, Kenneth T. Moore, John Hwang, Joaquim Martins, Andrew Ning. Automatic Evaluation
of Multidisciplinary Derivatives Using a Graph-Based Problem Formulation in OpenMDAO . [Citation] [PDF] [PDF
Plus]
308. Gaetan K. W. Kenway, Graeme J. Kennedy, Joaquim R. R. A. Martins. 2014. Scalable Parallel Approach for High-Fidelity
Steady-State Aeroelastic Analysis and Adjoint Derivative Computations. AIAA Journal 52:5, 935-951. [Abstract] [Full
Text] [PDF] [PDF Plus]
309. David J. Pate, Justin Gray, Brian J. German. 2014. A graph theoretic approach to problem formulation for multidisciplinary
design analysis and optimization. Structural and Multidisciplinary Optimization 49:5, 743-760. [Crossref]
310. James T. Allison, Daniel R. Herber. 2014. Special Section on Multidisciplinary Design Optimization: Multidisciplinary
Design Optimization of Dynamic Engineering Systems. AIAA Journal 52:4, 691-710. [Abstract] [Full Text] [PDF] [PDF
Plus]
311. Kai A. James, Graeme J. Kennedy, Joaquim R.R.A. Martins. 2014. Concurrent aerostructural topology optimization of a
wing box. Computers & Structures 134, 1-17. [Crossref]
312. Joaquim Martins, Graeme Kennedy, Gaetan K. Kenway. High Aspect Ratio Wing Design: Optimal Aerostructural
Tradeoffs for the Next Generation of Materials . [Citation] [PDF] [PDF Plus]
313. . Design Rule Extraction Using Multiobjective Design Exploration 643-667. [Citation] [PDF] [PDF Plus]
314. . Computational Intelligence in Aerospace Sciences . [Abstract] [PDF] [PDF Plus]
 315. Gaetan K. W. Kenway, Joaquim R. R. A. Martins. 2014. Multipoint High-Fidelity Aerostructural Optimization of a
Transport Aircraft Configuration. Journal of Aircraft 51:1, 144-160. [Abstract] [Full Text] [PDF] [PDF Plus]
316. . Multidisciplinary Design Optimization of Aerospace Transportation Systems 491-528. [Citation] [PDF] [PDF Plus]
317. Katharina Morawietz, Michael Hermann. 2014. Integrated Development and Modeling of Heat Pipe Solar Collectors.
Energy Procedia 48, 157-162. [Crossref]
318. Xiaoling Zhang, Debiao Meng, Ruan-Jian Yang, Zhonglai Wang, Hong-Zhong Huang. 2014. Bounded Target Cascading
in Hierarchical Design Optimization. Advances in Mechanical Engineering 6, 790620. [Crossref]
319. Jean-Antoine Désidéri. Multiple-gradient Descent Algorithm for Pareto-Front Identification 41-58. [Crossref]
320. Joaquim R. R. A. Martins, John T. Hwang. 2013. Review and Unification of Methods for Computing Derivatives of
Multidisciplinary Computational Models. AIAA Journal 51:11, 2582-2599. [Abstract] [Full Text] [PDF] [PDF Plus]
321. Justin Gray, Kenneth T. Moore, Tristan A. Hearn, Bret A. Naylor. 2013. Standard Platform for Benchmarking
Multidisciplinary Design Analysis and Optimization Architectures. AIAA Journal 51:10, 2380-2394. [Abstract] [Full Text]
[PDF] [PDF Plus]
322. Dawei Du, Dan Simon. 2013. Complex System Optimization Using Biogeography-Based Optimization. Mathematical
Downloaded by UNIVERSITY OF ILLINOIS on May 27, 2019 | http://arc.aiaa.org | DOI: 10.2514/1.J051895

Problems in Engineering 2013, 1-17. [Crossref]