Title: AFOSR/ARO/NSF/ONR/ESF Jointly Sponsored Workshop on Autonomic Structural Systems for Threat Mitigation Preliminary report of Group 2 SYSTEM DYNAMICS, CONTROL AND MECHARONICS Background review and analysis of System Dynamics, Control and Mechatronics
1AFOSR/ARO/NSF/ONR/ESF Jointly Sponsored Workshop
on Autonomic Structural Systems for Threat
MitigationPreliminary report of Group 2 SYSTEM
DYNAMICS, CONTROL AND MECHARONICS Background
review and analysis of System Dynamics, Control
and Mechatronics
-
- Group 2- SYSTEM DYNAMICS, CONTROL, AND
MECHATRONICS - Participants
- Asada, Harry asada_at_mit.edu
- Becker, Jürgen juergen.becker.external_at_eads.com,
Co-Chair - Bergman, Larry lbergman_at_uiuc.edu
- Hansen, Mark Markcocteau_at_stat.ucla.edu
- Krommer, Michael krommer_at_mechatronik.uni-linz.ac.a
t - Luber, Wolfgang Wolfgang.Luber_at_eads.com
- Percival, Donald dbp_at_apl.washington.edu
- Rodellar, Jose Jose.rodellar_at_upc.es
- Simpson, John john.simpson_at_eads.com
- Syrmakezis, Costa isaasyr_at_central.ntua.gr
- Masayoshito, Tomizuka tomizuka_at_me.berkeley.edu,
Co-Chair
2Preliminary report of Group 2Table of Contents
1. INTRODUCTION 2.DESCRIPTION OF DYNAMIC
THREATS 3. THREAT MITIGATION OF STRUCTURAL
SYSTEMS UNDER DYNAMIC LOAD CONDITIONS 4. THREAT
MITIGATION OF CONTROLLED STRUCTURAL SYSTEMS
UNDER DYNAMIC LOAD CONDITIONS 5. THREAT
MITIGATION BY AUTONOMIC STRUCTURAL SYSTEMS UNDER
DYNAMIC LOAD CONDITIONS 6. STATE OF THE ART
TECHNOLOGIES (Dyn Systems Control) 6.1 Modeling
and Identification 6.2 Control System Design
Methodologies Control Theories 6.3 Control
System Design Technologies Control Theories 6.4
Sensors/Actuators and Component Technologies
3Preliminary report of Group 2Table of Contents
(continues)
- 6.5 Applications and Mechatronics
- 6.6 Robotics and Automation e.g. mobile ME as
on-board ancillary system "R2-D2" - 6.7 ME(mechatronic elements) Catagories
- 6.8 System subordinationexpendable(sacrificial)
- 7. ROLE OF SYSTEM DYNAMICS, CONTROL
MECHATRONICS IN SMART AUTONOMIC STRUCTURES
SYSTEMS - 8. SHORT-TERM LONG-TERM GOALS APPLICTION
AREAS OF TECHNOLOGIES IN SYSTEM DYNAMICS, CONTROL
MECHATRONICS - 9. TECHNOLOGICAL BARRIERS TO REACH THE IDENTIFIED
GOALS - RECOMMENDATIONS FOR ACHIEVING THE GOALS AND
POTETIAL JOINT/COLLABORATIVE PROJECT ROAD MAP - 11. REFERENCES
4- 1. INTRODUCTION
-
- Autonomic Structural Systems Systems that can
withstand undesirable dynamic excitations and
inputs and has capability of self-xxxx including
self-healing and self-sustainable. - Challenges realization of smart and
multi-functional dynamic load bearing composite
material structures for and their integration in
civil, aerospace, surface transportation and a
variety of defence systems. - Problem areas include structures or electronic
subsystems under uncertain and stochastic natural
hazardous excitations the sensing and diagnosis
of dynamic threats, penetration prevention, load
capacity preservation, and functionality
restoration.
5- INTRODUCTION (cont.)
- The field of dynamic systems and control deals
with the modeling, design, analysis and
synthesis of functioning systems by integrating
various components. - The feedback principle is applied either
explicitly or implicitly in these processes. - Mechatronics considerations are important in the
field of dynamic systems and control. - The Intelligent Civil and Mechanical Systems
Cluster of the CMS (Civil and Mechanical Systems
Division), National Science Foundation has three
programs relevant to this workshop - Dynamic systems (Dr. Eduardo Misawa)
- Sensor Technologies for Civil and Mechanical
Systems (Dr. S.C. Liu) - Control Systems (Dr. Mario Rotea)
6Mechatronics
Synergistic integration of physical systems with
electronics/information technology and
complex-decision making in the design and
construction of functioning systems.
Modern Mechatronic Systems
7Mechatronics
- Synergy in terms of
- Dealing with complexity
- Performance
- Physical dimension, weight
- Time for development
- Cost
- Reliability
- Power efficiency ..
Synergy cannot be obtained via traditional
approach Material scientists come up with
materials Mechanical engineers design
mechananisms Control engineers design
controller..
82. DESCRIPTION OF DYNAMIC THREATS (to civil
and manned/unmanned military aircraft,
helicopters and spacecraft) Dynamic
excitations causing dynamic response of the total
structural system (vehicle, air-vehicle) may be
distinguished between external and internal
excitations. Internal excitations may result
from systems and smart systems interactions
causing dynamic loads due to servo-elastic or
aero-servo-elastic phenomena. The detection of
dynamic excitations causing threats by sensor
systems is essential for controlled system
dynamics for autonomic systems.
2.0 Dynamic Threat definition
behavioural (pilot) operationalmission
maintenance (engine, avionics,
quality) Airframe body/payload,
Multibody, Multi aircraft dynamics
9 2.1 Gust penetration Maximum gust
velocities defined by civil and military
specifications might be exceeded in reality
2.2 Stochastic excitation through air
turbulence Maximum rms velocities and power
spectral densities defined by civil and military
specifications might be exceeded in reality
2.3 Stochastic excitation through flow
separation on structural components of
air-vehicles Predicted buffet induced dynamic
loads acting on air vehicle components might be
exceeded
102. DESCRIPTION OF DYNAMIC THREATS
2.4 Wake penetration of civil and military
aircraft The wake penetration of civil or
military aircraft by another civil or military
aircraft can lead to dynamic loads exceeding the
dynamic structural design loads
2.5 Dynamic Impact loads Firing
gun firing impact, missile impact Bird
strike structure/engine intake Rough
runway, damaged runway Landing
impact Ice particle impact
Store jettison Noise etc
11 2.6 ASE (Aeroservoelastic)/ CASE
(Computer ASE hazard) Loss of control
surfaces impaired servoelastic
control designed in hazard aeroelastic/servo-
dynamics -- optimal effectiveness could be
severer mechanism in threat case
2.7 Flight Dynamics ballistically optimal EM
reflection by trajectory field of view (FOV)
2.8 Dynamic load excitance of EM
functionality 2.9 Dynamic loads HUMS (health
usage monitoring systems) interaction
12- 3. THREAT MITIGATION OF STRUCTURAL SYSTEMS UNDER
DYNAMIC LOAD CONDITIONS - Threat mitigation can be achieved to a certain
degree by a conventional design of the structure
system (not including active autonomic systems
for mitigation) by using maximum definitions of
dynamic excitations. - The classical design is based on system dynamics
modelling and experimental identification. - System safety and system certification is
achieved by a validated model. - This classical approach might also include a
process considering locally damaged structural
parts for the final purpose of safe landing
13- 4. THREAT MITIGATION OF CONTROLLED STRUCTURAL
SYSTEMS UNDER DYNAMIC LOAD CONDITIONS - Threat mitigation can be achieved to a certain
degree by a conventional design of the
controlled structure system (including active
gust load alleviation closed and open loop
systems, vibration alleviation systems, general
dynamic load suppression systems, but not
including active autonomic systems for
mitigation) by using maximum definitions of
dynamic excitations. - The classical design including the alleviation
systems is based on a coupled structural system /
control system /sensor /actuator - dynamics
modelling and experimental identification. System
safety and system certification is achieved by a
validated model trough on ground and in flight
system identification tests.
14- 5. THREAT MITIGATION BY AUTONOMIC STRUCTURAL
SYSTEMS UNDER DYNAMIC LOAD CONDITIONS - Threat mitigation can be achieved by autonomic
systems using as basis the conventional design
of the controlled structure system (including for
example active gust load alleviation closed and
open loop systems, vibration alleviation systems,
general dynamic load suppression systems) by
application of novel excitation detection
systems, health usage monitoring -, damage
monitoring- system capabilities and integrated
antenna systems.
15- 6. STATE OF THE ART TECHNOLOGIES
- Five Subfields
- Modeling and Identification (include FDI, Failure
Detection and Identification) - Control System Design Methodologies Control
Theories - Linear robust control/Nonlinear Control/Adaptive
and learning control - Actuators/Sensors and Component Technologies
- For Group 2, important issues include signal
processing, noise filtering, sensor and actuator
placement, - Applications and Mechatronics
- Robotics and Automation
- Concept design drivers
- with rigid plant assumption
- with flexible plant assumption
- with variable system design attitude
16- 6.1 Modeling and Identification
- Michael Krommer
- Smart composite structures can be characterized
as those that incorporate materials into the
structure with both sensing and actuation
capabilities. These materials are load bearing
hence, they need to be incorporated into the
modeling. - This clearly gives raise to the necessity of a
multi-field modeling approach, because many
different fields of different physical nature
will be involved. As one example the well studied
piezoelectric composite structures may be
mentioned. -
17- 6.1 Modeling and Identification (Larry Bergman)
- Identification methods developed for linear
systems generally fail when applied to damaged
structures. The underlying systems will be
strongly nonlinear and nonstationary with unknown
models. - The Hilbert-Huang transform (HHT) is the basis
for a fully nonparametric identification method.
It precisely determines the most appropriate
adaptive basis through the Empirical Mode
Decomposition (EMD) to provide the Intrinsic Mode
Functions (IMF) of the system. - Success of the method hinges on the recently
observed relationship between the slow flows of
the system, obtained from complexification and
averaging (CxA), and the IMFs.
18Figure 1 Response y(t) of oscillator (1) (a)
Comparison between exact (numerical) and
slow-flow models (b) Leading-order IMFs and
their instantaneous amplitudes.
19Figure 2 Comparisons of the results of the CxA
___________ and HHT ----------- methods for the
response y(t).
20- 6.1 Modeling and Identification
- Jose Rodellar
- In smart stucture and mechatronic systems (thus
in the new smart autonomic system), it is
expected to have materials and devices exhibiting
complex nonlinear dynamics with coupling. Two
modeling approaches are physical and
phenomenological. - Physical models are based on first principles and
constitutive relations. They support the
knowledge, design and prediction of behaviors. - Phenomenological models are approximations but
useful for control system design. An example is
the Bouc-Wen model which is often used for
describing MR dampers piezoelectric actuators,
base isolation devices for building, strctural
joints,
21- 6.2 Control System Design Methodologies Control
Theories - Michael Krommer
- CONTROL OF DISTRIBUTED PARAMETER SYSTEMS
- Modal-space control methods have become popular
- Powerful methods of modern control theory for
distributed parameter systems are available
nowadays.
- Tomizuka
- Fault Tolerant Control Methodologies such as
simultaneous stabilization (Vidyasagger) are
relevant. - If nonlinear dynamic models are linear with
respect to parameters, the adaptive control
theory may be rigorously applied (e.g. robotics). - Sensing and Precognition ? Preview control?
- Networked Sensing/Control.
22- 6.2 Control System Design Methodologies Control
Theories - Costas Syromakezis
- Incorporation of passive control elements into
the (civil) load-bearing systems (two case
studies) - A probabilistic approach on structural
vulnerability - Note added by MT Three basic approaches are 1)
passive control (dampers), 2) active control
(actuation) and 3) Semi-active control
(controllable dampers, etc)
23Fragility curves for the structure, with dampers
(continuous line) and without dampers (dashed
line)
24- 6.2 Control System Design Methodologies Control
Theories
- Harry Asada
- Distributed, stochastic control and broadcast
feedback - Control an ensemble of cellular units
stochastically based on broadcast feedback
information. - Scalable to the order of thousands and millions
of units, very robust, fault tolerant, etc.
25Control an ensemble of cellular units
stochastically
Broadcast Feedback
Aggregate error signal alone is broadcasted.
Each cellular unit makes a probabilistic decision.
State transition probability is modulated with
the error signal.
Aggregate Output is fed back
26- 6.3 Sensors/Actuators and Component Technologies
- Michael Krommer
- DISTRIBUTED SENSORS/ACTUATORS AND DENSE SENSOR
NETWORKS - In the last decade, a large number of research
studies have been devoted to continuously
distributed sensors for measuring overall
structural entities, such as natural frequencies
and modal amplitudes e.g. piezoelectric sensors
can be used for that purpose . - Similar to piezoelectric sensors, optical fibers
can measure a weighted integral over the strains
they are suffering throughout their extension,
and thus can be applied in structural control.
27 The typical strategy for the design of such
networks is based on the optimization of
performance indices based on modal and system
observability. These methods directly target at
their application in the control of vibrations,
but may not necessarily be used for detecting
localized damage or controlling local structural
entities. The corresponding methods are not yet
developed far enough to be applied to highly
redundant structures to be encountered in
sensitive facilities of civil and mechanical
engineering for the case of dense sensor
networks such methods are not available.
Similar to distributed sensors spatially
distributed actuators are widely used as modal
actuators.
28- 6.3 Sensors/Actuators and Component Technologies
(Harry Asada) - BINARY STRUCTURES
- Construct a structure with a collection of
elements with binary (ON-OFF) actuators having
bi-stable characteristics. - CELLULAR ACTUATORS
- Divide smart structure actuator materials into a
vast number of small segments, control individual
segments (cells) as finite state machines, and
coordinate the cells to control the aggregate
behavior of the entire cells. - Biologically inspired architecture, robust,
making a smart material, a linear stepping motor
29Bulk control of a smart material is difficult due
to hysteresis and nonlinearity.
Segment the actuator material into many tiny cells
Apply simple ON-OFF binary control ? Stepping
Motor, robust and simple
30Cellular Muscle Actuators Biologically Inspired,
building an actuator as an ensemble of many
cellular units.
Artificial Sarcomeres
Sarcomere
Spence, A.P., Basic Human Anatomy, 3-rd
Edition, Benjamin Cummings, 1990.
31- 6.5 Applications and Mechatronics
- contribution of J. S.
- Applications have been demonstrated by industry
and show state of the art (see J. Becker, J.
Simpson, K. Dittrich The future role of smart
structure systems in modern aircraft Smart
Structures and Systems, Vol. 1, No.2
(2005)describing electromagnetic structures
Structure integrated distributed Antennas (SIA)
health monitoring systems and vibration
alleviation - K. P. Kress Overview on the AMOS project,
International workshop on Structural Health
Monitoring, Stanford, Sept. 2003) describing
Structural health monitoring integrated antenna
manufacture monitoring - Individual Load Monitoring, Testing,
Diagnostics and Prognosis -
32- 6.6 Robotics and Automation
- Harry Asada
- Modular robots
- Build a reconfigurable structure or an active
structure by combinations of (small) modular
robots. They are reconfigurable, repairable,
flexible, robust, adaptable, distributed, etc. - 6.7 ME ( mechatronic elements) Catagories
- -non-mobile-mobile (multi-body ME)-distributed
- 6.8 System subordination
- expendable (sacrificial)
- endurance element
33Robotics and Automation
Reconfigurable modular robots
- Build a variety of structures with many modules
self assembly - Each module capable of sensing, actuation,
communication, and control
Reconfigurable, Repairable, Robust, Adaptable and
Distributed Applications to Space Mission,
Medical Robots
34- 8. SHORT-TERM LONG-TERM GOALS APPLICATION
AREAS OF TECHNOLOGIES IN SYSTEM DYNAMICS, CONTROL
MECHATRONICS. (Michael Krommer) - SHORT-TERM GOALS
- Advancement of existing methods
- system dynamics multi-field modeling, inverse
problems of dynamics for sensor and actuator
distribution and placement - control control theory of infinite-dimensional
systems - mechatronics (controlled system dynamics)
integrated design and simulation of structural
systems, including direct and inverse coupled
multi-field problems, sensors and actuators as
well as control systems - Application of available technology
- Technology transfer from fields, in which smart
material technology is accepted and already
successfully implemented to other fields e.g. to
civil engineering not only for monitoring, but
also for active control. - Application of available sensors/actuators for
the measurement/control of physical entities not
directly measured/controlled e.g. use of
strain-type sensors/actuators for the
measurement/control of stress. -
35LONG-TERM GOALS Novel multi-functional
materials Self-healing structures Development
of advanced methods of system dynamics for the
modeling and analysis of the overall dynamic
behavior of future self-healing structures in
combination with incorporating local dynamic
healing processes. Inverse problems of combined
overall and local dynamics for sensor and
actuator distribution and placement in
self-healing structures. Control system design
for self-healing structures -Taking into account
combined overall and local dynamics -Possible
need for localized control actively supporting
self-healing process.
369. TECHNOLOGICAL BARRIERS TO REACH THE IDENTIFIED
GOALS . Technological challenge (Jose
Rodellar) Develop phenomenological models for
complex behaviours able to capture essential
physical properties while keeping enough
simplicity for input-output characterization and
for enabling the design of controllers which can
be implemented in real time with the available
technology.
37- RECOMMENDATIONS FOR ACHIEVING THE GOALS AND
POTENTIAL JOINT/COLLABORATIVE PROJECT - M. K. proposes some recommendations for future
EDUCATION and RESEARCH in MECHATRONICS -
- An important goal would be the implementation of
new mechatronics curricula at universities in the
U.S. and in Europe. - From a European point of view these new
mechatronics curricula need to be implemented in
the framework of the ongoing Bologna process
including undergraduate, graduate and Ph.D.
degrees.
38- RECOMMENDATIONS FOR ACHIEVING THE GOALS AND
POTETIAL JOINT/COLLABORATIVE PROJECT (by Jose
Rodellar)
Develop joint projects with developers
of materials, sensors and actuators together with
a mix of theoretically/engineering oriented
researchers in dynamics and control. These
projects should emphasize on the integration of
disciplines and mutual approaches. For example,
simplified models for control purposes would be
better based if supervised by material developers
and by experts in complex numerical modeling in
conjunction with system concepts. Design of
sensors and actuators can benefit if
modeling/control people is cooperating at early
stages. For specific applications, experts-end
users are fundamental.
39- Challenges
- Education concerning inter-disciplinary
system design and mechatronics at Universities - Autonomic structural systems for threat
mitigation shall remain low weight,
manufacturable, cost-effective, environmentally
robust and sustainable (dynamic/static load
carrying). - Autonomic structural systems shall be
compatible with the overall vehicle system
structural requirements and all total system
control and other control system requirements - Integration of autonomic structural systems
into total systems - Modeling of excitations for example
turbulence , wake penetration including wake
turbulence - Sensing of local damage, sensing of gusts,
turbulence and wakes etc. flow field detection
systems -
40- Challenges Interdisciplinary approach for
dynamic modeling and design of the local
multilayer structures and total dynamic systems
including autonomic threat mitigation
systems (physical and phenomenological modeling)
(for instance elastic, servo-elastic,
aero-elastic, aero-servo-elastic modeling of
vehicles/air-vehicles, especially modeling of
stationary and unsteady aerodynamics at transonic
and at high incidence via elastic/aero-elastic
simulations using CFD (Navier- Stokes)
methods/predictions modeling including rigid
total vehicle/air-vehicle dynamics/ flight
mechanics). - Validation of local structural and total
structural system modeling through dynamic
experiments - Procedures and processes for system dynamic
qualification and certification for different
damage scenarios other challenges are t. b. d.
during discussion
41- Technological Barriers
- Autonomic structural load/dynamic load
carrying, low weight /small volume systems
compared to non autonomic systems - Autonomic structural load carrying, low
weight , self heeling systems - Affordable control systems for carefree
handling of systems and damaged structures,
damaged actuators and sensor systems (cycle times
of computers)other barriers t. b. d. during
discussion
42- Enablers
- Joint projects with developers of materials,
sensor and actuator and systems together with a
mix of theoretically/engineering oriented
researchers in dynamic and control expert end
users are fundamental - Joint research projects of EU/US
Universities and industry (dynamic system
modeling, system design and operation, system
strategies and technologies, parameter
identification, fault diagnostics and
identification and novel sensor technologies
etc.) Joint projects of Universities
and industry (dynamic smart and total system
modeling, system design and operation, smart
system integration into vehicle/air-vehicle and
or equipment systems, validation)
43- EnablersFor example - Conduction of joint
Univ/Industry research in the field of - Autonomic systems for local structural damage ,
partial destroyed integrated antennas or health
monitoring systems (phased array radar,
ultrasonic systems) - Autonomic flight control systems for partially
damaged structure, actuators , control surfaces
and sensors other enablers are t. b. d. during
discussion - Restrainerst. b. d.
44- Required Facilities Resources
- FacilitiesComputer software for dynamic
simulation of coupled elastic structure with
control systems (including embedded or other
sensors and electronic equipments and actuators
and without motion (vehicles) and with/without
motion induced and external excitation
aerodynamics.Test equipments for dynamic testing
of local multilayer and total structure systems
(shakers, ground resonance test facilities and
structural coupling test facilities on ground and
in flight)Resources - Funding through- EU framework research
programs- US research programs- industry
other are t. b. d. during discussion - Milestones - Short term goals t. b. d.
- - Long term goals t. b. d.