RESEARCH DIRECTIONS

IN COMPUTATIONAL MECHANICS
A Report of the United States National Committee on Theoretical and Applied Mechanics
September 2000
Computational Mechanics: A Core Discipline in Com- Not surprisingly, successful research in CM is usually
putational Science and Engineering interdisciplinary in nature, reflecting a combination of
Theoretical and Applied Mechanics (TAM) is the concepts, methods, and principles that often span sev-
branch of applied science concerned with the study of eral areas of mechanics, mathematics, computer sci-
mechanical phenomena: the behavior of fluids, solids, ences, and other scientific disciplines as well. As will
and complex materials under the actions of forces. Few soon become evident in this exposition, tomorrow’s
disciplines have had a greater impact on the industrial- research in CM will be broader than ever before, span-
ized world, enabling technological developments in vir- ning many new technologies and scientific fields.
tually every area that affects our lives, security, and
well being. Our goal here is to provide a perspective of the major
research areas in CM that will be the focus of inquiry
Computational Mechanics (CM) is that sub-discipline during the next decade: what are the research directions
of TAM concerned with the use of computational meth- in the CM and what are the opportunities for industrial,
ods and devices to study events governed by the prin- governmental, university researchers, and those who
ciples of mechanics. It is the fundamentally important would implement and apply the research results in com-
part of computational science and engineering con- putational mechanics?
cerned with the use of computational approaches to
characterize, predict, and simulate physical events and Computational Mechanics: Societal Benefits
engineering systems governed by the laws of mechan- The success of CM will ultimately be judged by effec-
ics. Computational mechanics has had a profound im- tiveness in solving problems of interest to society and
pact on science and technology over the past three de- on providing deeper understanding of natural phenom-
cades. It has transformed much of classical Newtonian ena and engineering systems. The field has been enor-
theory into practical tools for prediction and understand- mously successful to date because of its unprecedented
ing of complex systems. These are used in the simula- predictive powers, making possible the simulation of
tion and design of current and future advances in tech- complex physical events and the use of these simula-
nology throughout the developed and developing world. tions to design engineering systems. This is done
These have had a pervasive impact on manufacturing, through so-called “computer modeling”: the develop-
communication, transportation, medicine, defense and ment of discretized versions of the theories of mechan-
many other areas central to modern civilization. By in- ics which are amenable to digital computation, together
corporating new models of physical and biological sys- with the complex process of manipulating these digital
tems based upon quantum, molecular and biological representations to produce abstractions of the way real
mechanics, computational mechanics has an enormous systems behave.
potential for future growth and applicability.
Today, a thriving international industry markets com- that generate a model of blood flow through veins and
putational mechanics software for the analysis, simu- arteries of the specific human subject. In this example,
lation, and design of engineering products and systems. the flow of oxygenated blood that results from several
The industry revenues are in the billions of dollars; but different options in bypass surgery are calculated and
the overall impact on goods and services affects tril- presented to the surgical team so that the best proce-
lions of dollars in commerce and product development. dure for the particular patient under treatment can be
In addition, simulation software is used extensively in obtained. Many different surgical strategies can be
medical applications, military applications, transpor- simulated and the results predicted by CM software
tation, and even as tools to complement and validate before a single step in the actual surgery is taken.
experimental research and testing.

Some of the applications of CM are well known; oth-

ers are not. One well-known area in which CM has
had dramatic success is with the simulation of crash
worthiness of automobiles. Computer-generated simu-
lations of the collision of a vehicle with walls or ob-
stacles, based on fundamental scientific principles on
the dynamics of deformable bodies, have replaced hun-
dreds of full-scale tests and saved major automobile
manufacturers millions of dollars. More importantly,
countless lives have been saved and injuries diminished
by improved safety features developed through com- Figure 2a:
puter modeling and simulation. A human aorta containing an aneurism, and diseased veins
and arteries, shaded gray in the figures, and various by-
pass surgical procedures for allowing healthy blood flow
Figure 1 depicts a computer-generated simulation of a around the affected area. The blood flow is modeled for
full-size car in oblique impact with a cylindrical pole. each surgical process using principles of fluid mechanics
The entire impact event, which takes place in a thou- and fluid-structure interactions.
sandth of a second, can be studied in detail in the CM
simulation, providing information on deformation, ac-
celeration, forces on models of passengers, energy trans-
Figure 2b:
fer, and other features of the phenomena that are useful The computed flow chan-
in designing safer and more reliable vehicles. nels and blood flow from
the surgical solutions best
suited for a particular pa-
Figure 1: tient is selected and mod-
Computer-generated simulation eled in detail using CM
of the oblique impact of a full- principles.
scale model of a van collid-
ing with a cylindrical pole.
This simulation models the Oden
damage of the vehicle
and provides informa-
tion on passenger safety
for various structural de- CM has long been used in military applications. One
signs of the vehicle. Many different design parameters can be var- example is in the analysis and design of weapons and
ied to determine the effects of each design decision before a single armor, such as were the goals of the simulations de-
physical test is done.
picted in Figure 3. There, we observe a computer-gen-
erated simulation of a highly complex phenomenon that
An exciting CM application area under development is occurs in micro-seconds: the oblique hypervelocity
predictive surgery. One example is illustrated in Fig- implates. Here a long list of interactive mechanical
ure 2, where a computer-generated image of a diseased phenomena take place, including phase changes of ma-
human aorta is given. The geometry and properties of terials, ablations and spalling, shrapnel spraying, ther-
the living tissue are deduced from MRI imaging and mal effects, and flow – all within the realm of TAM
other tests and go directly into computer subroutines and all modeled using CM technology.
SCORDIM report page 3

Figure 4b:
A model of the dispersion of
pollutants in a porous substrata
where biochemical reactions
Figure 3a: are followed. Various species
A simulation of a metallic rod ( a penetrator) impacting and concentrations shown in differ-
piercing a collection of armor plates in a design of military ent colors within a complex
systems structure.

Another example of a CM simulation in this general

area is indicated in Figures 5a, b. There, we find a
simulation of a supernova, the explosion of a star, a
billion light years from earth. Fundamental principles
Figure 3b: of fluid mechanics and gas dynamics, and thermody-
A simulation of a hypervelocity impact of a deformable pen-
etration impacting multiple plates of armor. These calcula- namics were used to construct a computational model
tions enable the better design of protective armor configu- of this extraterrestrial event, and the remarkable simu-
rations. lations give new information on the structure and the
evolution of supernovas, the mass of the system, ve-
locities of debris, and other data that suggest mecha-
Applications of CM are not limited to the engineering
nisms for the cause of the phenomena.
design of products and systems. Many are concerned
with the basic understanding of natural phenomena or
with the prediction of natural physical events. Examples
in this area include the use of CM methods to study
atmospheric changes, ocean currents, surface flow in
rivers, subsurface flows in oil reservoirs, or geological
phenomena such as the movement and evolution of
polar ice caps or the tectonic plates. One such simula-
tion is illustrated in Figures 4a, b. Shown in Figure 4a
is a computer-generated model of the topography and
substrata of an oil reservoir and the flow of oil, water,
and gas in a simulation designed to estimate well pro- Figure 5a:
Here the theories of gas dynamics and com-
duction. Figure 4b is a computer visualization of the pressible flow of multiphase media are used
flow of biological and chemical species in a simula- to model the explosion of a distant star.
tion of pollution remediation in porous media. Colors indicate variations in mass density
and temperature.

Figure 5b:
A zoom into a closer
look at the evolution
of the supernova.

Figure 4a:
Computer-generated 3D map of an oil reservoir and
the flow of oil, gas, and water through a complex
porous media
Reliability of CM opment and many open issues remain.
The question of reliability of computer-generated pre-
dictions is one of great concern to specialists in CM. During the next decade, the field of CM will undergo
Without some confidence in the accuracy of simula- dramatic changes that will require computational me-
tions, their value is obviously diminished. Today, re- chanics capabilities not available today. An unprec-
markably accurate and reliable simulations are obtained edented growth in activity and importance is expected
routinely in many application areas while others are, at that may have been unimaginable only a few years ago.
best, qualitative and capable of depicting only trends Also, in the next millennium, a paradigm shift in the
in physical events. computational sciences is essential if the needs of in-
dustry are to be met.
This concern for reliability has led to the creation of a
challenging technological area labeled simply Valida- Perhaps the most remarkable aspect of our vision of
tion and Verification: the creation, study and documen- CM in the next decade is that CM research will require
tation of tools for assessing the predictability of CM- new interactions of computational methods and devices
based methods and programs. Validation has to do with with a variety of supporting technologies, including
determining the appropriateness of the scientific prin- imaging, various tomographic modalities, visualization,
ciples and mathematical models used to develop a simu- testing, and laboratory experimentation. In the new CM,
lation tool. Verification has to do with determining if “mechanics” will interact with theories of quantum
the final tool can indeed function as it was intended mechanics, molecular dynamics, materials science, bio-
and if it can correctly produce results consistent with medical and biological systems, and other disciplines.
the models upon which it is based. It will involve the study of microscopic phenomena
taking place in pica seconds. A myriad of technologi-
Validation methods include, when possible, actual com- cal tools need to be integrated into the mechanical analy-
parisons of CM predictions with observations, physi- sis, not only for studying engineering materials, but also
cal tests and experiments. Even then, such compari- to study biological systems, submicron devices, and
sons can be made for only a limited range of param- robots. These far-reaching applications will impact
eters and generally not the full range of possibilities predictability, our life styles and personal health and
embraced by general computational modeling schemes. longevity.
Modern validation methods also seek to define the limits
of various modeling schemes and, ideally, to provide Funding Sources: Computer Versus Computational
insight as to when these limits are reached and can be Science
overcome. In general, the subtle difference between the terms
“computer science” and “computational science”, are
Verification methods involve the use of benchmark tests not recognized but the distinct differences in these dis-
developed by experienced analysts over years of work, ciplines has had a significant impact on the distribu-
or implementation of software engineering methods to tion of resources for fundamental research. Computer
minimize coding errors and optimize computational science refers to the science and technology pertinent
performance and program design. to the computer, the computational device with which
computation is done. Computational science, on the
With the rapidly expanding use of CM throughout many other hand, addresses the development of modeling
areas of science and engineering, “V and V” has be- techniques, algorithms, software, and for specific prob-
come a critical component of today’s research in CM. lems in science and engineering. There has been a ten-
New V and V tools are on the horizon that could dra- dency to promote investment in computer science as
matically increase the reliability of sophisticated com- the necessary tool with which good computational sci-
putation modeling. Some of these tools are mentioned ence can be done, but there has been no comparable
later in this exposition. support by traditional funding agencies has emerged in
the important computational sciences necessary to do
Computational Mechanics: the Next Decade and the important applications. A good example is PITAC, the
Next Millennium President’s Information Technology Advisory Commit-
It may be confidently said that virtually every aspect tee, largely populated by computer scientists as opposed
of our day-to-day lives is affected in some way by CM. to computational scientists, which after a study of over
Still, the subject is undergoing rapid change and devel- a year, recommended that federal government invest
SCORDIM report page 5

over a billion dollars into new research in “informa- and in the engine compartment, are used in designs by
tion technology”, principally pure computer science, computers. Crash and occupant protection that, in the
networking, and hardware. In explaining the value of past, required many prototypes, are also simulated in
this program to the US Congress, however, the great the design process. For Defense Department products,
benefits of such an investment for doing computational extreme environments, such as live-fire tests are also
science were stressed, but no funds were allocated to increasingly simulated.
the latter.
Although great strides have been made in simulation
In order to meet the needs of industry, renewed empha- in the past two decades, virtual prototyping is still more
sis on the computational sciences, and, particularly, CM, of an art than a science. To develop a virtual prototyping
is needed. This is not to say that some impressive in- capability, many tests must be performed since many
vestments have not been made in this discipline. In- of the physical phenomena can not be modeled on the
deed, very ambitious programs are underway that will basis first principles today. Instead, models are tuned
raise the bar in computational modeling. These include to tests, and the technology is not applicable to radi-
the ASCI program (Accelerated Scientific Computa- cally new designs. Specific obstacles to virtual
tion Initiative) which is aimed at displacing traditional prototyping include the inability to simulate problems
nuclear testing and storage procedures with highly- with multiphysics phenomena, such as burning and
tuned, computer simulations. It remains to be seen change of phase, fracture and spalling, phenomena in-
whether many of the goals of ASCI are attainable, but volving large disparities in scales, and behavior with a
the program undoubtedly will create some important significant stochastic characteristics.
advances in CM and related sciences. Such programs
also foster a broadening and generalization of all fields These capabilities are also of crucial importance to our
supporting computational science, and intrinsically defense. With the rapid development of new concepts
advance supporting technologies to make meaningful of warfare and defense, new weapons and devices must
simulations possible. be quickly designed and evaluated. Virtual design and
prototyping are essential in this process. For example,
In the pages below, we outline briefly a number of ar- with the new emphasis on the soldier and body armor,
eas with significant research opportunities in CM: various protective devices must be evaluated. How-
w Virtual Design ever, modeling of materials such as kevlar and other
w Multi-scale Phenomena, including bridging of mo- new materials in the failure range require a dynamic
lecular to atomistic to continuum models failure analysis that is beyond the state of our knowl-
w Model Selection and Adaptivity edge. These capabilities are also essential to maintain-
w Very Large-scale Parallel Computing ing our nuclear weapons stockpile without testing.
w Biomedical Applications, including predictive sur-
gery, application of mechanics to the study of cells, In order to make virtual design a reality in the next
bones, nerves, and other biological systems decade, radically new computational tools with the
w Controlling Uncertainty: Probabilistic Methods ability to handle multiscale phenomena, very hetero-
geneous materials, and discontinuous behavior, such
Virtual Design as fracture and assessment of the range of performance
One of the major factors in increasing U.S. industrial and automatic guidance to improving design, must be
competitiveness is the reduction in design cycle time. available.
Such reduction hinges critically on the availability of
virtual design, the ability to complete designs entirely Multi-scale phenomena
in the computer, without making time-consuming pro- One of the great strengths of computer simulation over
totypes. In virtual design, prototyping is bypassed and traditional methods of experimental science is its ap-
prototype tests of normal operations and extreme con- plicability to the study of a complete range of physical
ditions are simulated on the computer. For example, in phenomena at all possible spatial and temporal scales.
the automobile industry, normal operations of the en- This ability of computer simulation to slow down events
gine and the body, such as ride assessment, stresses in that take only 10-9 seconds and study them in some de-
the engine, and car body and airflow around the body tail, or to look back into time at the evolution of events
that began tens of thousands or hundreds of thousands scribe the same physical phenomena. The selection of
of years in the past, or to extrapolate to the evolution of the model, by which we ordinarily mean the selection
events, and make predictions of their structural effects of the partial differential, integral or ordinary differen-
and characteristics many millennia into the future, all tial equations, the algorithms, the physical, geometri-
are, in principle, achievable with computer simulations. cal and topological characteristics, boundary and ini-
Correspondingly, the study of mechanical events with tial conditions, etc, is quite often the single most im-
spatial scales take place at the atomistic level, the mo- portant step in obtaining valid computer simulations of
tion of dislocations, electrons and protons, the analysis physical events.
of submicron devices, the study of events with interac-
tions of many different scales, such as turbulent flow, In recent years, considerable progress has been made
to the characterization of motion of solar systems, gal- in determining theoretical and computational techniques
axies, and systems of gargantuan dimension, are also that aid in model selection. A variety of techniques are
within the scope of tomorrow’s work on CM. under study. Some of these involve embedding a given
class of models into a larger class of more sophisti-
A major challenge to CM for the future is to model cated models in which finer and more detailed repre-
events in which these remarkably varying scales are sentations of the behavior of physical system may be
significant in a single system or phenomena. It is then possible. Once such a datum is identified, a notion of
necessary to model multi-scale phenomena simulta- modeling error can be made precise, and by various
neously for predictive capability. Analysis of multi-scale measures, such modeling error can be controlled by
phenomena, while apparently beyond the horizon of adaptive modeling processes. Areas in which adaptive
contemporary capabilities, is one of the most funda- modeling have great promise include the study and char-
mental challenges of research in the next decade and acterization of composite materials, unsteady turbulent
beyond. So-called scale bridging, in which the careful flows, multiphase flows of fluids, etc. Other techniques
characterization of mechanical phenomena require that for model adaptivity involve the use and integration of
the model “bridge” the representations of events that test and imaging data, feedback from experiments and
occur at two or more scales, require the development measurements, and various combinations of these meth-
of a variety of new techniques and methods. In this odologies.
area, integration of computational methods and devices
with experimental or sensing devices is critical. High Model selection is a crucial element in automating en-
fidelity simulation and computational mechanics must gineering analysis and applications are unlimited; the
involve innovative and efficient use of a spectrum of subject could conceivably embrace classes of models
imaging modalities, including x-ray tomography, elec- including diverse spatial and temporal scales, enabling
tron microscopy, sonar tomography, and many others. the systematic and controlled simulation of events mod-
Similarly, in modeling phenomena such as climate eled using atomistic or molecular models to continuum
changes, weather conditions, and the interaction of models. Model selection, model error estimation, and
ocean and atmosphere, satellite-generated data must be model adaptivity are exciting areas of CM and prom-
incorporated seamlessly into viable computational ise to provide an active area of research for the next
models to obtain meaningful predictions. Again, the decade and beyond.
spectrum of computational mechanics must be signifi-
cantly broadened to include the use of these technolo- Controlling Uncertainty: Probabilistic Methods
gies. Once more, the intrinsically interdisciplinary na- The random nature of many features of physical events
ture of the subject will be expanded and reinforced. is widely recognized by industry and researchers. The
natural stimuli that activate physical systems may be
Model Selection and Adaptivity completely unpredictable by deterministic models: the
Throughout all the mathematical and computational randomness of a gust of wind, the characterization of
sciences, the first and most primitive step in computer forces in boundary and initial conditions on mechani-
modeling is the selection of the mathematical and com- cal systems, random microstructural features of engi-
putational model itself. Model selection is a largely neering materials, the random fluctuations in tempera-
heuristic process, based on the judgement and experi- ture, humidity, and other environmental factors, all
ence of the modeler, and on testing and experimenta- make the characterizations provided by deterministic
tion. But it is frequently purely a subjective endeavor: models of mechanics less satisfactory with respect to
different analysts may select different models to de- their predictive capabilities.
SCORDIM report page 7

Fortunately, the entire subject of uncertainty can itself simulation as any other estimate of physical quantities
be addressed in a scientific and mathematically pre- of interest. An important advance in this area has been
cise way and the random characteristics of nature can the recent discovery of methods to determine upper and
be addressed by computational models. lower bounds of local approximation error, so that in
any given simulation, once a particular model is se-
During the next decade, probabilistic modeling of lected, computable bounds giving upper and lower lim-
problems in mechanical problems will be a topic of its to computed quantities of interest could be a natural
great importance and interest. Including stochastic by-product in every simulation. This is a fertile area of
features into computational models will not only pro- research, one in which significant work will be done
vide realistic simulations of physical events but will during the next decade.
also provide the analyst with specific information on
the probabilities that can be assigned to predictions. Very Large Scale Parallel Computing
Thus, using probabilistic models of mechanics, the One of the most difficult issues facing researchers in
analyst may, for example, determine what the maxi- CM in the next decade will be purely a conceptual one:
mum stress will be in a particular machine part under the recalibration of their own education, approach, and
design, rather than upper and lower bounds on such perceptions to allow them to use efficiently the extraor-
quantities of interest can be expected in view of the dinary computational tools that will be developed dur-
uncertainty of the data. What, in particular, is the prob- ing this period. Today, mechanicians using computa-
ability of failure of a sub-system being analyzed and tional products in engineering analysis and design can
designed? New methods for treating uncertainty will routinely develop computational models involving
become important in virtually all branches of mechan- 500,000 to 10,000,000 degrees of freedom. Problems
ics: fluid mechanics, mechanics of materials, solid of this size today are being solved on contemporary
mechanics; it will also promote the development of workstations. Nevertheless, these contemporary mod-
new computational techniques to analyze uncertainty els employ rather crude characterizations of materials,
in engineering systems. geometry, boundary conditions, failure criteria, and
many other important features of the system, because
Error Estimation and Adaptivity it is taken for granted by the modeler that to include
The notion of computing estimates of numerical error these details will result in computational problems so
in computer simulations is not new; serious work in large and complex that they would exceed the capaci-
this subject began in the 1980’s, and today the notion ties of modern computational facilities.
of a posteriori error estimation is a common topic in
university research environments. Error estimation This argument is no longer correct. As the twenty-first
provides a quantitative measure for determining the century begins, computational devices capable of de-
quality of numerical simulations; it provides a basis livering five trillion operations per second and storing
for adapting characteristics of discrete models (for ex- a thousand trillion bytes of data are in use and larger
ample, meshes or approximation orders) so as to im- machines are being developed. In a decade’s time, ma-
prove the quality of results. chines with capabilities an order-of-magnitude beyond
this level may be available. It is probable that such
To date, most of a posteriori estimation has been con- terascale computation capabilities will soon be in the
fined to a fairly narrow class of problems, largely hands of most engineers and mechanicians, thus mak-
drawn from linear theory; it has not been fully uti- ing possible models with a level of detail and sophisti-
lized in the more complex computer simulations used cation completely unimaginable only a decade ago.
in industry and government laboratories.
The proper use of this extraordinary toolkit will itself
It is predicted that a posteriori error estimation and represent a significant challenge. Included in the chal-
adaptivity will become a common ingredient in all sig- lenge is the education of the next generation of engi-
nificant computer simulations in CM during the next neers and mechanicians who will be expected to not
decade. With a renewed and invigorated interest in re- only master the principles of mechanics but also the
liability of simulations, the calculation of estimates of use of the computational tools available to them.
error in simulations will be as natural a feature of the
These new capabilities, and advances in modeling and Acknowledgements
parallel computation, will ultimately have a remark- This document is one of a series prepared by SCORDIM, the
Subcommittee on Research Directions in Mechanics, a stand-
able and irreversible impact in education in science and ing committee of the U.S. National Committee for Theoretical
engineering. Simplified models and approximate theo- and Applied Mathematics. Previous volumes in this series:
ries remain important in developing understanding, but 1. Research Directions in Computational Mechanics, edited by
students need no longer rehearse only idealized situa- J.T. Oden, National Research Council, National Academy
Press, Washington, D.C, 1991.
tions: they can now tackle more realistic models. High-
2. Research Trends in Fluid Dynamics, edited by J.L. Lumley,
speed parallel computing together with the software de- Andreas Acrivos, L. Gary Leal, and Sidney Leibovich, Ameri-
velopments, alluded to elsewhere in this document, will can Institute of Physics, Woodbury, New York, 1996.
create a revolution in engineering analysis and ulti- 3. Research Directions in Solid Mechanics, edited by G. Dvorak,
mately in the way it is taught in colleges and universi- Elsevier Science Publishers, Amsterdam, 1999.
ties. Less than a decade ago many feared that access to This report was written and compiled by a subcommittee on Re-
modern computational methods and machines would search Directions in Computational Mechanics: J. Tinsley Oden,
breed overconfidence in engineers, at the expense of The University of Texas (Chair); Ted Belytschko, Northwestern
common sense, judgement and reasoning. Now, the University; Ivo Babuska, The University of Texas; and T.J.R.
Hughes, Stanford University.
new concern is one of underestimation of the power of
modern computational methods and devices and the Acknowledgement is given to the following individuals for pro-
danger of their under-utilization in important simula- viding figures:
tions, analysis and design. w Page 1: Professor Leszek Demkowicz, TICAM, The Univer-
sity of Texas
w Figure 1: Ford Motor Company
Computational Mechanics: Conclusions w Figures 2a, b: Professor Charles Taylor, Department of Me-
CM has become a central enabling discipline that has chanical Engineering, Stanford University
led to greater understanding and advances in modern w Figures 3a, b: Dr. David Littlefield, IAT, The University of
science and technology. It has been the basis of nu- Texas
w Figures 4a, b: Professors Chandrajit Bajaj and Mary F.
merous important developments in recent years and will
Wheeler, TICAM, The University of Texas
continue to be crucial to industrial development and w Figures 5a, b: Professors Chandrajit Bajaj TICAM, and Paul
competition, to safety and security, and to understand- Shapiro, Department of Astronomy, The University of Texas
ing the diverse physical and biological systems occur-
ring in nature and in the society. Many research prob-
lems in CM await resolution and will provide signifi-
cant challenges for research in the future. If “the past
in prologue”, then, in the future, we may anticipate even
greater contributions of Computational Mechanics to
the advancement of knowledge and to the benefit of a
global society.

USNC/TAM ROSTER
November 1, 1999 - October 31, 2000
OFFICERS SOCIETY REPRESENTATIVES IUTAM-BASED MEMBERS SCORDIM REPORT
Ron Adrian, Chair Earl H. Dowell, AAM L. B. Freund, Treasurer COMMITTEE
Earl H. Dowell, First Past-Chair Dr. Charles E. Harris, AIAA Bruno A. Boley, General Assembly MAL Andreas Acrivos
James C. Hill, AIChE Daniel C. Drucker, General Assembly MAL Ronald Adrian
Gary Leal, Second Past-Chair
Philip John Holmes, AMS Yu H. Ku, General Assembly MAL Ted Belytschko
James C. Hill, Vice chair Parviz Moin, APS J. Tinsley Oden, Rep Int Ass’n Comp Mech George J. Dvorak
Philip G. Hodge, Secretary Allan D. Pierce, ASA Andreas Acrivos, Congress Committee Carl Herakovich
Stein Sture, ASCS David Bogy, Congress Committee John Hutchinson
MEMBERS-AT-LARGE Carl T. Herakovich, ASME Hassan Aref, Congress Committee Stelios Kyriakides
Ilene Busch-Vishniac David McDowell, ASTM Jan D. Achenbach, Solids Symposium Panel Parviz Moin
Edmund G. Henneke William L. Fourney, SEM J. Tinsley Oden
John Hutchinson George Dvorak, SES NRC STAFF Allan D. Pierce
Wolfgang G. Knauss Daljit S. Ahluwalia, SIAM Ms. Lois Peterson
Stelios Kyriakides Michael M. Bernitsas, SNAME
Andrea Prosperetti Dr. Andrew Kraynik, SOR
Alexander Smits