Download textbook Sensors And Instrumentation Aircraft Aerospace And Energy Harvesting Volume 8 Proceedings Of The 36Th Imac A Conference And Exposition On Structural Dynamics 2018 Evro Wee Sit ebook all chapter pdf

Sensors and Instrumentation,
Aircraft/Aerospace and Energy

Harvesting, Volume 8 Proceedings of
the 36th IMAC, A Conference and
Exposition on Structural Dynamics
2018 Evro Wee Sit
Visit to download the full and correct content document:
https://textbookfull.com/product/sensors-and-instrumentation-aircraft-aerospace-and-
energy-harvesting-volume-8-proceedings-of-the-36th-imac-a-conference-and-expositi
on-on-structural-dynamics-2018-evro-wee-sit/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Sensors and Instrumentation, Aircraft/Aerospace, Energy

Harvesting & Dynamic Environments Testing, Volume 7:
Proceedings of the 37th IMAC, A Conference and
Exposition on Structural Dynamics 2019 Chad Walber
https://textbookfull.com/product/sensors-and-instrumentation-
aircraft-aerospace-energy-harvesting-dynamic-environments-
testing-volume-7-proceedings-of-the-37th-imac-a-conference-and-
exposition-on-structural-dynamics-2019-chad/
Sensors and Instrumentation Aircraft Aerospace Energy

Harvesting Dynamic Environments Testing Volume 7
Proceedings of the 38th IMAC A Conference and
Exposition on Structural Dynamics 2020 1st Edition Chad
Walber
https://textbookfull.com/product/sensors-and-instrumentation-
aircraft-aerospace-energy-harvesting-dynamic-environments-
exposition-on-structural-dynamics-2020-1st-edition/
Special Topics in Structural Dynamics, Volume 5:

Exposition on Structural Dynamics 2018 Nikolaos
Dervilis
https://textbookfull.com/product/special-topics-in-structural-
dynamics-volume-5-proceedings-of-the-36th-imac-a-conference-and-
exposition-on-structural-dynamics-2018-nikolaos-dervilis/
Biota Grow 2C gather 2C cook Loucas
https://textbookfull.com/product/biota-grow-2c-gather-2c-cook-
loucas/
Rotating Machinery, Vibro-Acoustics & Laser Vibrometry,
Volume 7: Proceedings of the 36th IMAC, A Conference
and Exposition on Structural Dynamics 2018 Dario Di
Maio
https://textbookfull.com/product/rotating-machinery-vibro-
acoustics-laser-vibrometry-volume-7-proceedings-of-the-36th-imac-
a-conference-and-exposition-on-structural-dynamics-2018-dario-di-
maio/
Dynamics of Civil Structures, Volume 2: Proceedings of

the 37th IMAC, A Conference and Exposition on
Structural Dynamics 2019 Shamim Pakzad
https://textbookfull.com/product/dynamics-of-civil-structures-
volume-2-proceedings-of-the-37th-imac-a-conference-and-
exposition-on-structural-dynamics-2019-shamim-pakzad/
Topics in Modal Analysis & Testing, Volume 8:

Exposition on Structural Dynamics 2019 Michael L. Mains
https://textbookfull.com/product/topics-in-modal-analysis-
exposition-on-structural-dynamics-2019-michael-l-mains/
Special Topics in Structural Dynamics & Experimental

Techniques, Volume 5: Proceedings of the 37th IMAC, A
Conference and Exposition on Structural Dynamics 2019
Nikolaos Dervilis
https://textbookfull.com/product/special-topics-in-structural-
dynamics-experimental-techniques-volume-5-proceedings-of-
the-37th-imac-a-conference-and-exposition-on-structural-
dynamics-2019-nikolaos-dervilis/
Topics in Modal Analysis Testing Volume 8 Proceedings

of the 38th IMAC A Conference and Exposition on
Structural Dynamics 2020 1st Edition Brandon Dilworth
https://textbookfull.com/product/topics-in-modal-analysis-
exposition-on-structural-dynamics-2020-1st-edition-brandon-
Conference Proceedings of the Society for Experimental Mechanics Series
Evro Wee Sit · Chad Walber · Patrick Walter

Alfred Wicks · Steve Seidlitz Editors

Aircraft/Aerospace and
Energy Harvesting,
Volume 8
Proceedings of the 36th IMAC, A Conference and Exposition
on Structural Dynamics 2018
Series Editor
Kristin B. Zimmerman, Ph.D.
Society for Experimental Mechanics, Inc.,
Bethel, CT, USA
More information about this series at http://www.springer.com/series/8922
Evro Wee Sit • Chad Walber • Patrick Walter • Alfred Wicks • Steve Seidlitz
Editors

Aircraft/Aerospace and Energy Harvesting,
Volume 8
Proceedings of the 36th IMAC, A Conference and Exposition
on Structural Dynamics 2018
123
Editors
Evro Wee Sit Chad Walber
Svcommunity com PCB Piezotronics, Inc
Hermosa Beach, CA, USA Depew, NY, USA
Patrick Walter Alfred Wicks

Texas Christian University Virginia Polytechnic Institute & State University
Fort Worth, TX, USA Blacksburg, VA, USA
Steve Seidlitz
Cummins (United States)
St. Paul, MN, USA
ISSN 2191-5644 ISSN 2191-5652 (electronic)

ISBN 978-3-319-74641-8 ISBN 978-3-319-74642-5 (eBook)
https://doi.org/10.1007/978-3-319-74642-5
Library of Congress Control Number: 2018941837
© The Society for Experimental Mechanics, Inc. 2019

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights
of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or
information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific
statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date
of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for
any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional
affiliations.
Printed on acid-free paper
This Springer imprint is published by the registered company Springer International Publishing AG part of Springer Nature.
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
Sensors and Instrumentation, Aircraft/Aerospace and Energy Harvesting represents one of nine volumes of technical
papers presented at the 36th IMAC, A Conference and Exposition on Structural Dynamics, organized by the Society for
Experimental Mechanics, and held in Orlando, Florida, February 12–15, 2018. The full proceedings also include volumes on
Nonlinear Dynamics; Dynamics of Civil Structures; Model Validation and Uncertainty Quantification; Dynamics of Coupled
Structures; Special Topics in Structural Dynamics; Structural Health Monitoring, Photogrammetry and DIC; Rotating
Machinery, Vibro-Acoustics and Laser Vibrometry; and Topics in Modal Analysis and Testing.
Each collection presents early findings from experimental and computational investigations on an important area within
sensors and instrumentation and other structural dynamics areas. Topics represent papers on calibration, smart sensors,
practical issues improving energy harvesting measurements, shock calibration and shock environment synthesis, and
applications for aircraft/aerospace structures.
The organizers would like to thank the authors, presenters, session organizers, and session chairs for their participation in
this track.
Hermosa Beach, CA, USA Evro Wee Sit

Depew, NY, USA Chad Walber
Fort Worth, TX, USA Patrick Walter
Blacksburg, VA, USA Alfred Wicks
St. Paul, MN, USA Steve Seidlitz
v
Contents
1 Broadband Energy Harvesting Performance of a Piezoelectrically Generated Bistable Laminate . . . . . . . . . . . . 1

Andrew J. Lee and Daniel J. Inman
2 Performance Assessment of Several Low-Cost Consumer-Grade Analog-to-Digital Conversion Devices. . . . . 15
G. R. Wetherington Jr
3 High-g Shock Acceleration Measurement Using Martlet Wireless Sensing System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Xi Liu, Xinjun Dong, Yang Wang, Lauren Stewart, Jacob Dodson, and Bryan Joyce
4 Effect of Piezoelectric Material in Mitigation of Aerodynamic Forces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Gholamreza Amirinia, Sungmoon Jung, and Grzegorz Kakareko
5 A Theoretical Description of a Multi-source Energy Harvester . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
J. Gosliga and D. J. Wagg
6 Lumped Mass Model of a 1D Metastructure with Vibration Absorbers with Varying Mass . . . . . . . . . . . . . . . . . . . 49
Katherine K. Reichl and Daniel J. Inman
7 Dynamic Behavior and Performance Analysis of Piezoelastic Energy Harvesters Under Model and
Parameter Uncertainties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Paulo S. Varoto
8 Experimental Test of Spacecraft Parachute Deployment Using Real-Time Hybrid Substructuring. . . . . . . . . . . 67
Michael J. Harris and Richard E. Christenson
9 Experimental and Analytical Approaches in a Virtual Shaker Testing Simulation Environment for
Numerical Prediction of a Spacecraft Vibration Test. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
S. Waimer, S. Manzato, B. Peeters, M. Wagner, and P. Guillaume
10 Direct Reference-Free Dynamic Deflection Measurement of Railroad Bridge under Service Load . . . . . . . . . . . 83
Bideng Liu, Ali Ozdagli, and Fernando Moreu
11 A Low-Cost Modular Impact-Based Experimental Setup for Evaluation of EMI Based Structural
Health Monitoring at High Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Ekramul Haque Ehite and Steven R. Anton
12 Real-Time Low-Cost Wireless Reference-Free Displacement Sensing of Railroad Bridges. . . . . . . . . . . . . . . . . . . . . 103
Ali Ozdagli, Bideng Liu, and Fernando Moreu
13 Multi-Tonal Based Impedance Measurements for Microsecond State Detection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Ryan A. Kettle and Steven R. Anton
14 Design and Installation Considerations of Dynamic Strain Gages for Thermo-Acoustic Aerospace
Structures Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Matthew S. Stefanski and William E. Boles
15 TESS Vibration Testing: A Boundary Condition Case Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
Alexandra Karlicek and Allison Pinosky
vii
viii Contents
16 Performing a Large-Scale Modal Test on the B2 Stand Crane at NASA’s Stennis Space Center . . . . . . . . . . . . . . 135
Eric C. Stasiunas and Russel A. Parks
17 Study on the Soft Suspension Behavior for Aircraft Ground Vibration Test Set-Up . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
Antonio Almeida Giacomin, Airton Nabarrete, Marcelo Camilo Alves Costa, and Tatiana Chloe Digou
18 A Review of the Vibration Environment Onboard Small Unmanned Aircraft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
William H. Semke and Matthew D. Dunlevy
Chapter 1
Broadband Energy Harvesting Performance of a Piezoelectrically
Generated Bistable Laminate
Andrew J. Lee and Daniel J. Inman
Abstract The vibration based energy harvesting performance of a piezoelectrically generated bistable laminate consisting of
only Macro Fiber Composites (MFC) is experimentally characterized. Conventionally, piezoelectric transducers are bonded
onto thermally induced bistable composite laminates and exhibit broadband cross-well dynamics that are exploited for
improved power generation over linear resonant harvesters. Recently, a novel method of inducing bistability was proposed
by bonding two actuated MFCs in a Œ0MFC =90MFC T layup and releasing the voltage post cure to create in-plane residual
stresses and yield two cylindrically stable configurations. Forward and backward frequency sweeps at multiple acceleration
levels across the first two observed modes of the laminate’s two states are performed to identify all dynamic regimes
and the corresponding voltages produced by each MFC. Besides single-well oscillations, snap throughs are observed
in intermittencies, subharmonic, chaotic, and limit cycle oscillations across wide frequency ranges. Resistor sweeps are
conducted for each regime to determine maximum power outputs, and single and multi-frequency performance metrics
accounting for laminate volume, mass, input accelerations, and frequencies are evaluated for the laminate. A performance
comparison with conventional bistable composite harvesters demonstrate the laminate’s viability for energy harvesting,
allowing it to be multi-functional in combination with its snap through morphing capability.
1.1 Introduction
Broadband energy harvesting applications enabled by nonlinear adaptive structures under vibration have been and continue
attracting significant attention from researchers. This is due to the harvesters retaining efficiency when ambient vibrational
energy is distributed over a wide spectrum where the spectral density may be variable over time and dominant at low
frequencies [1]. In contrast, linear harvesters are suited only for stationary and narrowband excitation at their fundamental
resonant frequencies which limit their usefulness in more realistic environments. To broaden the usable bandwidth during
operation, nonlinearities inherent in bistable oscillators have been exploited by numerous past works [2–4] as an efficient and
robust mechanism to convert vibrations into electrical energy. These structures have two stable equilibrium states with each
corresponding to a distinct geometric configuration, and transition between states occur through buckling, or a nonlinear
jump phenomenon known as snap through. Depending on the excitation frequency and amplitude, bistable structures exhibit
multiple dynamic regimes that are distinct. This includes single-well vibrations around either of the stable configurations,
or cross-well vibrations causing both periodic and aperiodic snap through between both stable equilibria. In certain cases,
multiple regimes may coexist under the same vibrational input, but only one is physically realizable depending on the initial
conditions. Cross-well regimes such as high amplitude limit cycle oscillations have shown to significantly improve power
generation over linear resonance as the harvesters displace between states under high velocities [5]. The attraction to bistable
harvesters derive from being able to trigger cross-well vibrations associated with large deformations across a wide range of
excitation conditions, which alleviate the performance limitations experienced by their linear counterparts.
Various mechanisms to induce bistability for the purpose of broadband energy harvesting have been explored in past
investigations. Popular concepts include obtaining bistability in a cantilevered ferromagnetic beam with magnetic attraction
[6, 7], destabilizing a cantilevered beam with a magnetic tip mass from its neutral position with magnetic repulsion [8–10],
buckling a clamped-clamped beam with an axial load to obtain a mechanically induced bistable structure [11, 12], and
generating internal thermal stresses within a unsymmetric composite laminate during cooldown in its cure cycle which result
A. J. Lee () · D. J. Inman

Department of Aerospace Engineering, University of Michigan, Ann Arbor, MI, USA
e-mail: ajle@umich.edu
© The Society for Experimental Mechanics, Inc. 2019 1

E. Wee Sit et al. (eds.), Sensors and Instrumentation, Aircraft/Aerospace and Energy Harvesting, Volume 8,
Conference Proceedings of the Society for Experimental Mechanics Series, https://doi.org/10.1007/978-3-319-74642-5_1
2 A. J. Lee and D. J. Inman
in two stable configurations [13–22]. As for the energy harvesting method, piezoelectric materials are attached to bistable
structures to convert vibration induced strains to electrical energy due to their large power densities, ease of fabrication and
application over other methods such as electrostatic, electromagnetic, and magnetostrictive transduction. Recently, a novel
means of generating bistability with piezoelectric actuation was investigated by the authors [23, 24]. By bonding two actuated
Macro Fiber Composites (MFC) in a cross ply layup and shutting off the power post cure, two stable states are generated
through the production of in-plane residual stresses. This is due to the applied electric field causing anisotropic strains within
the MFCs from the mismatch of their effective piezoelectric constants. The mechanics are analogous to how a bistable
composite laminate is manufactured through the mismatch of its coefficients of thermal expansion between plies [25].
The efficiency gained from simultaneously being the harvester and the primary structure may alleviate challenges inherent
in conventional bistable composite harvesters. A design conflict exists where larger piezoelectric patches could generate
more energy, but reduces the curvature of the host structure and thus limits its response amplitudes and strains along the
patch’s polarization direction. In addition, the increased stiffness from larger piezoelectric layers raise the vibrational energy
requirement for cross-well dynamics. To mitigate this issue, Betts et al. [26] determined optimal layups, aspect ratios, and
piezoelectric areas for a bistable composite harvester. This conflict is avoided by the entirely piezoelectric harvester since it
lacks the stiffness of the host composite laminate while retaining the curvatures generated by the MFCs. Its lower stiffness
implies that the vibrational energies required for maintaining high energy orbits and broadening cross-well bandwidths are
lower than those of conventional bistable composites, and may not require proof masses utilized by many of these harvesters
to aid snap through. The proposed harvester however consists of two P1 type MFCs which have a d33 poling direction
along the piezoceramic fibers with inter-digitated electrodes, and are not ideal for energy harvesting applications due to their
low capacitance which limits the current output [27, 28]. In all studies utilizing MFCs as the harvesting unit, the P2 type
is chosen which has a d31 poling direction through its thickness in a parallel plate configuration, which results in higher
capacitance and lower impedance characteristics [15–17, 19, 20]. Despite this, only the P1 type MFCs operate with a high
enough piezoelectric coefficient and voltage range necessary for generating bistability [24].
The purpose of this paper is to investigate the energy harvesting performance of the piezoelectrically generated bistable
laminate by characterizing its dynamic regimes under harmonic excitation and implementing a simple resistive circuit for
energy conversion across the observed modes. Single and multi-frequency metrics available in literature [4, 29] will be
utilized to compare the performance of the proposed harvester’s regimes and those of bistable composite harvesters due
to their similarity in mechanism. These metrics aim to address the issue of establishing fair and appropriate measures
of effectiveness for the harvesters by accounting for the excitation input and their physical characteristics. This paper
experimentally confirms the laminate’s viability for energy harvesting with comparable performance to existing concepts.
The authors already demonstrated the bistable laminate’s ability to snap through between either stable states with no
external assistance by actuating one MFC at a time within its operating limit [24]. The combined morphing and energy
harvesting capabilities enable the bistable laminate to be multi-functional with the potential for self-powered morphing and
full configuration control.
1.2 Experimental Methodology
1.2.1 Design and Manufacturing
The design of the harvester is catered towards creating the most favorable conditions for inducing bistability and maximizing
the range of motion between the stable configurations. The latter condition will enable larger strains to be imparted on the
laminate during cross-well vibrations, which will raise the subsequent power output. The simple layup of Œ0MFC =90MFC T is
chosen to minimize the total thickness and allow access to the electrodes leads of each MFC, while the cross ply configuration
will maximize the out-of-plane displacements of each state [30]. The P1 type MFC utilizing the d33 effect is chosen for the
voltage actuation required for bistability, and strains in the piezoceramic fiber direction under an operating voltage of 500
to 1500 V. The MFCs are simultaneously actuated at 1500 V while being bonded under vacuum in order to yield the most
post cure delta voltage and resulting piezoelectric strains. This also provides the most margin for bifurcation from one to
two stable states to occur post cure when the voltage is released. According to previous analysis [23], the critical side length
for generating bistability is 85 mm when the aspect ratio of the laminate is 1 and the actuation voltage during bonding is
1500 V. However at this length, the bifurcation voltage is 0 V and this leaves no margin for manufacturing imperfections
1 Broadband Energy Harvesting Performance of a Piezoelectrically Generated Bistable Laminate 3
Fig. 1.1 (a) Stable states of the 200 200 mm2 Œ0MFC =90MFC T bistable laminate and (b) experimental test setup
while completely limiting the deformation differences between the two resulting configurations. With MFC manufacturing
limitations considered, the side length of 200 mm is chosen to ensure sufficiently large displacements and adequate margin
for producing bistability.
The 200 200 mm2 Œ0MFC =90MFC T laminate is manufactured by actuating two P1 type MFCs at 1500 V in parallel with
a BK Precision 9130 DC power supply connected to a Trek 2220 high voltage amplifier and bonding them together on a flat
tool under vacuum. The MFCs are custom made by Smart Material Corp. and have an active area of 200 200 mm2 , but their
total dimensions are 207 218 mm2 due to the peripheral areas housing the electrode rails and leads between the polyimide
films. Therefore, the laminate has 5.5 mm of overhang on each edge with a total bonded area of 207 207 mm2 . The 3M
DP-460 epoxy adhesive is used for bonding and cured for 24 h at room temperature. After the cure cycle, the laminate is
qualitatively confirmed to be bistable once it is taken out of vacuum and the power to both MFCs are shut off. Figure 1.1a
identifies the two stable states of the manufactured Œ0MFC =90MFC T laminate. In this paper, the major curvature axis of state I
is along the piezoceramic fiber direction of MFC 1 and that of state II corresponds to MFC 2 in the same manner.
1.2.2 Experimental Setup
Characterizing the Œ0MFC =90MFC T laminate’s dynamic regimes and power output under vibrational excitation is achieved
with the experimental setup shown in Fig. 1.1b. The harmonic input to the laminate is provided by the APS 113 seismic shaker
and powered by the APS 125 amplifier while the laminate is fixed at the center to a stinger that is attached to the shaker. To
avoid invasive attachment methods such as drilling a hole and potentially damaging the MFCs, a small neodymium magnet
is bonded to the stinger and another identical magnet is then used to mount the laminate onto the stinger. The acceleration at
the laminate center is measured by a PCB Piezoelectronics 352C67 accelerometer and the velocity response is measured at a
single point by a Polytec OFV-534 laser vibrometer with the Polytec OFV-5000 controller. Reflective tape is adhered to the
laminate to improve the signal return of the laser vibrometer. The acceleration, voltage, and velocity responses are recorded
with the NI USB-6211 DAQ connected to LabVIEW at a sample rate of 2500 Hz while this combination simultaneously
provides the shaker signal based on the acceleration measurements. The voltage responses of each MFC are measured
separately and voltage divider circuits are used to reduce the signals to adhere to the DAQ input limit of ˙10 V during
frequency sweeps. All signals are low-pass filtered and processed in MATLAB. When measuring the harvested power with
resistor sweeps, a Tektronix TDS2004C oscilloscope is used to record the root mean square (RMS) voltage output of each
MFC without the voltage divider circuits to allow for simple variation of resistive loads, which is quickly implemented with
an IET ohmSOURCE resistance box.
1.3 Results and Discussion
1.3.1 Vibration Modes
The dominant modes of the harvester for both stable states are identified with low amplitude forward frequency sweeps and
corresponding velocity to base acceleration frequency response functions (FRF) are generated. Determining the resonant
frequencies allow subsequent sweep range and locations to be appropriately chosen for characterizing nonlinear dynamic
regimes at high excitation amplitudes, since they appear around these modes. Specifically, sweeps are conducted for short
and open circuit electrical boundary conditions from 10 to 50 Hz at 0.05 g to minimize nonlinear effects. The velocity
measurement locations are at the edges of the active area in the middle of the laminate along the major curvature axis of
each state, which allow just their dominant modes to be recorded. Figure 1.2 shows the FRFs of each stable state at low
frequencies for the open circuit condition, which are found to be nearly identical to the closed circuit condition.
The first modes of both stable states are identical at 16.58 Hz and exhibit out-of-plane rigid body rotation about the stinger
that is inherent to the shaker setup, and is associated with much less elastic deformation than the second mode. The second
modes of states I and II are respectively at 27.26 and 26.45 Hz and they are traditional plate bending modes. The slight
difference in the second mode indicate that the potential wells of the bistable laminate are not symmetric and reflect the
shape discrepancies between the two configurations where state I has higher out-of-plane displacements than state II. This
may be due to manufacturing errors involving imperfect layup alignment, a non-uniform epoxy bond line, MFC performance
degradation from long actuation times during bonding, and the variation of piezoelectric constants and elastic properties
between the two MFCs [24]. These geometric and material imperfections are difficult to control during the manufacturing
process, and the resulting state I over state II bias must be considered in the experimental analysis. However, the separation
of modes may be advantageous for energy harvesting applications since it extends the total cross-well bandwidth associated
with large amplitude oscillations.
1.3.2 High Amplitude Frequency Sweeps
Based on where the linear vibration modes are, forward and backward frequency sweeps at higher amplitudes are conducted
in order to identify the nonlinear dynamic regimes of the bistable harvester. The sweeps range from 10 to 34 Hz, which
encompass the first two observed modes of each state and the time histories of the base acceleration, corner velocity, MFC
101
State I
State II
Velocity FRF ((m/s)/ g)
100
10-1
10-2
10 15 20 25 30
Frequency (Hz)
Fig. 1.2 Experimental velocity to base acceleration frequency response function for both stable states
Fig. 1.3 Peak to peak amplitudes in initial state I for (a) MFC 1 open circuit voltage, (b) MFC 2 open circuit voltage, (c) corner velocity, and in
initial state II for (d) MFC 1 open circuit voltage, (e) MFC 2 open circuit voltage, and (f) corner velocity during forward (blue) and backward (red)
frequency sweeps at 3 g excitation
1 and MFC 2 open circuit voltages are recorded every 0.5 Hz. This procedure is done for input accelerations of 0.5 to 4 g
in steps of 0.5 g for both initial states. The corner location is chosen for velocity measurements because it is the only point
on the laminate to have significant out-of-plane displacements for both stable configurations, which makes it appropriate for
measuring cross-well vibrations from a single location. Peak to peak amplitudes of the recorded time histories are obtained
with stroboscopic sampling at excitation frequencies over multiple forcing periods and the resulting figure for 3 g input is
shown in Fig. 1.3. This figure is a combination of an FRF and a bifurcation diagram with the excitation frequency being
the bifurcation parameter under constant acceleration amplitude [13]. For linear regimes, the sampled amplitude for a given
excitation frequency appears as a single point while nonlinear responses are given by multiple points indicating the variation
of amplitudes over several consecutive periods. The 3 g acceleration is the lowest excitation level where all dynamic regimes
are present for both states, and therefore a suitable input parameter for harvester response and performance characterization.
Note that Fig. 1.3 does not include all coexisting solutions per sweep, but does show some of them due to the nonlinearities
inherent to the harvester structure and their dependency on initial conditions. These arise from the softening stiffness effect
causing the reduction of resonant frequencies, sweep direction dependent hysteresis forcing the boundaries between single
and cross-well regimes to shift, and the asymmetry of the bistable laminate’s potential wells causing static state dependent
responses and bandwidths [31]. As expected, the strength of these nonlinear effects increase with higher excitation levels. For
both states, the softening effect pushes the boundaries between the single and cross-well regions below their second linear
modal frequencies regardless of the sweep direction. Hysteretic regions separating the jumps in amplitude are seen from 12.5
to 18 Hz for state I and 11.5 to 17.5 Hz for state II where the backward sweeps induce high amplitude limit cycle oscillations
while the forward sweeps remain linear over the same bandwidths, albeit subharmonic resonance can be seen at half the
resonant frequency for the bending mode. In contrast, the forward sweeps extend the cross-well regions from 24 to 24.5 Hz
for state I and 24.5 to 26.5 Hz for state II over the opposite path. These differences are a consequence of different solutions
gaining and losing stability at various frequencies, and are not necessarily the only steady-state solutions in existence. The
transition into limit cycle oscillations during backward sweeps may be due to the coalescence of vibrational energy carried
by the cross-well motion and the first rigid body mode at 16.58 Hz. Once the backwards sweep passes through the first mode,
there is a drop off in amplitude for limit cycle oscillations until returning back to the linear regime. It does not exist in the
forward sweep due to the harvester being in low energy orbit when approaching the first mode.
When comparing initial conditions, state II produces a larger 15 Hz cross-well bandwidth over the 12 Hz bandwidth of
state I with 2.5 Hz more range for limit cycle oscillations desirable for energy harvesting. In addition, the corner velocities
and corresponding output voltages of both MFCs are overall higher for state II. This is due to the out-of-plane displacements
being lower for state II, and thus the vibrational energy required to snap through the harvester is also lower. There is a clear
correlation between the corner velocity and voltages of the harvester, with cross-well vibrations producing larger amplitudes
than their single-well counterparts. During single-well vibrations for state I, MFC 1’s voltage output is much higher than
MFC 2 due to the induced strains on the former being dominant in the piezoceramic fiber direction associated with the P1
type MFC’s primary d33 piezoelectric effect, while the latter is prominently strained in the electrode direction normal to
the fibers. These dominant strain directions and the ratio of voltage outputs are flipped between MFCs for state II. Once
the harvester exhibits cross-well oscillations, the voltage differences are considerably reduced since both MFCs are now
operating in the d33 mode.
1.3.3 Characterization of Dynamic Responses
Aside from limit cycle oscillations, other nonlinear regimes are associated with lower response amplitudes and include
intermittencies, chaotic, and subharmonic oscillations. In Fig. 1.3 over a broad range of amplitudes, several points for a given
frequency indicate the presence of harmonics while a dense cluster of points suggest chaotic behavior. However, it provides
limited detail and identification of dynamic regime per frequency is difficult, especially for intermittencies consisting of both
periodic and aperiodic characteristics. Therefore, the time histories over multiple forcing periods and their corresponding
Fourier spectra, phase portraits, and Poincaré maps of every observed regime are analyzed for distinguishing characteristics.
This section evaluates example cases for each type of response found in the 3 g frequency sweeps. Figure 1.4 presents the
aforementioned plots for high amplitude limit cycle oscillation with input parameters of 14.5 Hz and 3 g during the state
I backward sweep. The corner displacements in Fig. 1.4f and subsequent figures are estimated by integrating the corner
velocity signal, applying high pass filters to prevent drift, and aligning their peaks with the static corner displacement of state
I, which is measured to be 10.42 mm out of plane. The frequency content is obtained with fast Fourier transform (FFT) of
time histories recorded during the sweep.
Cross-well limit cycle oscillations are characterized by continuous snap through events which allow a periodic high-
energy orbit be sustained between the two stable states. It is associated with the largest velocity and voltage amplitudes out
of all dynamic regimes due to the periodic attractor motion having the most penetration into both potential wells, as seen
in Fig. 1.4f. The slight asymmetry between states still exist due to the state I initial condition and the state I over II bias
Fig. 1.4 MFC 1 (red) and 2 (blue) open circuit voltages in the forms of (a) time histories, (b) FFTs, (c) phase portraits and Poincaré maps, and
corner velocity in the form of (d) time history, (e) FFT, (f) phase portrait and Poincaré map for high amplitude limit cycle oscillations at 14.5 Hz
and 3 g during state I backward sweep
discussed previously, but this difference decreases with increasing excitation levels. The electromechanical phase portrait
in Fig. 1.4c is enabled by the 90ı phase difference between the open circuit voltages and corner velocity. It differs from
Fig. 1.4f, where the voltage output is sharply cut when the MFCs are not operating in the d33 mode. This effect alternates
between MFCs according to their 180ı phase difference as also seen in Fig. 1.4a, and is a result of the continuous snap
throughs. Consequences of asymmetry and bias towards the initial state are higher output voltages for MFC 1 over 2, and
this is correlated by the slightly larger corner velocities and displacements while the harvester is in state I. The Poincaré maps
in Fig. 1.4c, f indicate a single-periodic response and the Fourier spectra in Fig. 1.4b, e reveals the associated main harmonic
(i.e. !) of 14.5 Hz to be dominant with 2! and 3! superharmonic components contributing much less energy. In contrast
to the periodic nature of limit cycle oscillations, Fig. 1.5 presents cross-well chaos under various representations at 18.5 Hz
and 3 g in initial state I. To generate enough data for the phase portraits, the measurements are taken separately for 15 min
instead of processing the sweep data.
Under chaotic response, the harvester exhibits a strange attractor sensitive to initial conditions [32] as shown in the
Poincaré maps in Fig. 1.5c, f, i. The time histories in Fig. 1.5a, d, g reveal that the harvester motion involving snap through
events never quite repeats itself, though there are passages of nearly recurrent behavior, and therefore remains completely
aperiodic. The FFTs in Fig. 1.5b, e, and h shows a much more broadband spectrum with the primary component being
the excitation frequency of 18.5 Hz, but energy is present over a wide range of frequencies. Although the phase portrait
trajectories escape the potential well of state I and displays cross-well behavior, the majority of vibration is centered around
state I with much less penetration into state II when compared to limit cycle oscillations. Besides limit cycle oscillations,
all other cross-well regimes display this same behavior to varying degrees and is attenuated as the acceleration input rises.
This growth in asymmetry in combination with infrequent snap through events, lower velocities, and output voltages suggest
the chaotic response to be less favorable for harvesting power than the high energy orbits of Fig. 1.4. Figure 1.6 presents
the rest of the harvester’s dynamic responses in the initial state I 3 g backward sweeps and include subharmonic and linear
oscillations at 22.5 and 34 Hz, intermittency between limit cycle and chaotic oscillations at 18 Hz, and intermittency between
subharmonic and chaotic oscillations at 20.5 Hz.
Fig. 1.5 MFC 1 (red) open circuit voltage in the form of (a) time history, (b) FFT, (c) phase portrait and Poincaré map, MFC 2 (blue) open circuit
voltage in the form of (d) time history, (e) FFT, (f) phase portrait and Poincaré map, and corner velocity in the form of (g) time history, (h) FFT,
(i) phase portrait and Poincaré map for chaotic oscillations at 18.5 Hz and 3 g in initial state I
Fig. 1.6 Intermittency between limit cycle and chaos at 18 Hz for MFC 1 (red) and 2 (blue) open circuit voltages in the forms of (a) time histories,
(b) FFTs, (c) phase portraits and Poincaré maps, subharmonic oscillations at 22.5 Hz for MFC 1 (red) and 2 (blue) in the forms of (d) time histories,
(e) FFTs, (f) phase portraits and Poincaré maps, intermittency between subharmonic and chaotic oscillations at 20.5 Hz for MFC 1 (red) and 2
(blue) in the forms of (g) time histories, (h) FFTs, (i) phase portraits and Poincaré maps, and linear oscillations at 34 Hz for MFC 1 (red) and 2
(blue) in the forms of (j) time histories, (k) FFTs, (l) phase portraits and Poincaré maps during 3 g state I backward sweep
Intermittency is attributed by the sporadic switching between two qualitatively different behaviors while the excitation
input is kept constant [33], and these behaviors are cross-well chaos and limit cycle oscillations in the case of Fig. 1.6a–c.
The voltage time histories in Fig. 1.6a display the nearly periodic motion interrupted by occasionally irregular bursts that
are statistically distributed, causing the respective spread and loss of periodicity in the phase portraits and Poincaré maps
in Fig. 1.6c. The FFTs in Fig. 1.6b show that the main excitation frequency and its superharmonic components are still
dominant like in Fig. 1.4b, e, but the presence of chaos spreads the energy contribution across the spectrum. The behavior of
this regime suggest that its energy harvesting capability is between those of limit cycle and chaotic oscillations. Figure 1.6d–f
exhibits cross-well subharmonic oscillations, or specifically a period-3 oscillation that takes 3 forcing periods to complete
a full cycle [34]. The Poincaré maps in Fig. 1.6f assess the periodicity of the response by sampling it once per period, and
reveal three distinct locations within the phase portrait of each MFC. Its frequency content in Fig. 1.6e shows the presence
of order-1/3 harmonic components (!=3, 2!=3, 3!=3, etc) with fairly even distribution, and responsible for inducing the
period-3 response. It should be noted that a range of subharmonic oscillations (period-2 through 8) is observed during
the frequency sweeps across all excitation levels in both single and cross-well motion, but they are not presented here.
Intermittency is also found between subharmonic and chaotic regimes as seen in Fig. 1.6g–i. At 20.5 Hz specifically, irregular
alternation of period-2 and chaotic oscillations can be seen in the corresponding time histories and the Poincaré maps reveal
the spreading of two distinct clusters for both MFCs, indicating the aperiodic nature within the subharmonic response.
FFTs show the dominance of order-1/2 harmonic components (!=2, 2!=2, 3!=2, etc) with contributions by in-between
frequencies resulting from the intermittent chaos. The harvester’s final dynamic regime is single-well linear oscillations as
seen in Fig. 1.6j–l at the non-resonant frequency of 34 Hz. As expected, the two MFC voltage signals are periodic and in
phase with each other, MFC 1 outputs much larger values, and the only frequency component is the excitation signal.
Fig. 1.7 Observed dynamic responses at multiple acceleration levels during forward frequency sweeps with initial (a) state I, (b) state II, and
during backward frequency sweeps with initial (c) state I, and (d) state II
With all of the observable dynamic regimes characterized, the rest of the frequency sweep data at other acceleration inputs
are evaluated to identify how the harvester responds under a range of excitation levels. A summary of this evaluation is shown
in Fig. 1.7, but note that not all coexisting solutions are present. At 0.5 g, nonlinear effects such as softening and hysteresis
discussed previously are minimal and resonance is seen very close to the plate bending mode at 27.26 and 26.45 Hz for states
I and II. With higher accelerations, the strength of nonlinearities rise where the bending mode frequency is pushed backwards,
hysteretic regions and subharmonic resonance grow, and single-well subharmonic oscillations are observed near the resonant
frequency. Cross-well vibration in the form of chaos first appears at 1.5 g forward sweep for state II at the reduced resonant
frequency of 22 Hz, and the rest of the sweeps follow suit with the addition of cross-well subharmonic oscillations at 2 g. As
the excitation level continues to increase, the cross-well bandwidth is extended primarily in the direction that the sweep is in.
High amplitude limit cycle oscillations first appear during the 2.5 g backward sweep for state I at 16.5 Hz and its bandwidth
grows with higher acceleration levels at the expense of other cross-well regimes. As expected, intermittencies are observed
between the regimes they are respectively alternating from. At 4 g, the total cross-well bandwidth is 14 Hz for state I and
20.5 Hz for state II due to the asymmetry of potential wells.
1.3.4 Energy Harvesting Capability
Excluding limit cycle oscillations, Fig. 1.3 shows that the open circuit voltage amplitudes of cross-well regimes are similarly
ranged, and this holds true for other acceleration levels. Even after accounting for the steady rise in maximum amplitudes
towards the onset of limit cycle oscillations, it is not clear how each regime’s average power output will compare relative to
each other due to their intermixing in this bandwidth, and warrants further investigation. To evaluate the energy harvesting
capability of the harvester, resistor sweeps are conducted across 1, 1.78, 3.16, 5.62, 10, 15.8, 25.1, 39.8, 63.1, 100, 178,
316, 562 k, and 1 M in order to match the impedance of each MFC to the load resistance, which maximizes their power
output. With each MFC, the sweeps are run at 3 g for the observed regimes outlined in the previous section, and the results
are presented in Figs. 1.8 and 1.9 for states I and II. The excitation frequencies are where the maximum power is generated
for each regime with the exception of cross-well subharmonic oscillations, which are chosen to be period-3 for both states to
allow fair comparison. As observed in [7], the range of load resistances in the sweeps do not suppress any of the nonlinear
phenomena (i.e. shunt damping effect) exhibited by the harvester, and so it is able to maintain the cross-well responses
(a) 102 (b) 102 (c) 102 MFC 1

MFC 2
Power (mW)
Power (mW)
Power (mW)
100 101 101
10-2 100 100
10-4 10-1 10-1

103 104 105 106 103 104 105 106 103 104 105 106
Load Resistance ( Ω) Load Resistance ( Ω) Load Resistance ( Ω)
(d) 102 (e) 102 (f) 102

Power (mW)
Power (mW)
Power (mW)
101
101 101
100
10-1 100 100

103 104 105 106 103 104 105 106 103 104 105 106
Fig. 1.8 Power output of MFC 1 (red) and 2 (blue) in initial state I during resistor sweep at 3 g for (a) linear (24.5 Hz), (b) subharmonic period-3
(22.5 Hz), (c) intermittency subharmonic-chaotic (20.5 Hz), (d) chaotic (18.5 Hz), (e) intermittency limit cycle-chaotic (18 Hz), and (f) limit cycle
oscillations (17 Hz)
(a) 102 (b) 102 (c) 102 MFC 1

MFC 2
101
Power (mW)
Power (mW)
Power (mW)
101 101
0
10
100 100
10-1
10-2 10-1 10-1

103 104 105 106 103 10 4 105 106 103 104 105 106
(d) 102 (e) 102 (f) 102

Power (mW)
Power (mW)
Power (mW)
101
101 101
100
10-1 100 100

103 104 105 106 103 104 105 106 103 104 105 106
Fig. 1.9 Power output of MFC 1 (red) and 2 (blue) in initial state II during resistor sweep at 3 g for (a) linear (26.5 Hz), (b) subharmonic period-3
(23 Hz), (c) intermittency subharmonic-chaotic (21 Hz), (d) chaotic (19 Hz), (e) intermittency limit cycle-chaotic (18.5 Hz), and (f) limit cycle
oscillations (17.5 Hz)
resulting from each excitation frequency. This is certainly favorable for examining the harvesting performance of each
dynamic regime, and is likely due to the mechanical nonlinearity inherent to bistability being dominant over the piezoelectric
coupling of the MFCs.
Figures 1.8 and 1.9 show that the optimal resistor load is 39.8 k for every regime except for the linear response in state
I, which is 25.1 k. High amplitude limit cycle oscillations output the most power at 110.3 and 130.7 mW for states I and
II, with the least amount of difference between the MFCs due to the minimal asymmetry that exist for the two potential
wells. This is the most favorable regime for energy harvesting as the single-periodic cross-well orbit allow snap through
Table 1.1 Resistor sweep results and peak power output of each dynamic regime at 3 g excitation
Optimal load MFC 1 MFC 2 Total Normalized
Initial Frequency resistance power power power power density
Dynamic regime state (Hz) (k) (mW) (mW) (mW) (kgs m3 )
Linear single-well I 24.5 25.1 29.41 0.09 29.50 2.43
II 26.5 39.8 0.18 24.57 24.75 2.04
Subharmonic cross-well I 22.5 39.8 15.82 5.54 21.37 1.76
II 23 39.8 7.50 15.45 22.95 1.89
Intermittency subharmonic-chaos I 20.5 39.8 20.49 9.26 29.76 2.45
II 21 39.8 10.37 22.51 32.88 2.71
Chaos I 18.5 39.8 25.26 11.61 36.87 3.04
II 19 39.8 12.39 26.25 38.64 3.18
Intermittency limit cycle-chaos I 18 39.8 35.90 16.93 52.83 4.35
II 18.5 39.8 18.81 38.07 56.88 4.69
Limit cycle I 17 39.8 65.35 44.99 110.34 9.09
II 17.5 39.8 52.87 77.79 130.66 10.76
events to most frequently and consistently occur, and the large associated velocities translate to deeper well penetration. The
other cross-well responses show larger power gaps between the MFCs as asymmetry grows, and this is the maximum for
linear oscillations due to their orbit about a single state allowing just one MFC to be operating in the d33 mode. After limit
cycle oscillations, the power generation of the regimes in descending order are intermittency between limit cycle and chaotic
oscillations, chaos, single-well linear oscillations, intermittency between subharmonic and chaotic oscillations, and cross-
well subharmonic oscillations for state I. The peak linear response outperforms cross-well subharmonic behavior since it is at
the boundary between single and cross-well regimes which is near the reduced resonant frequency. Although linear maximum
amplitudes may be lower as seen in Figs. 1.3 and 1.6d, g reveal that snap through occurs more infrequently for subharmonic
oscillations and signifies that their peak amplitudes are hit less often when compared to other cross-well responses. It is
interesting to see that the resonant single-well response of one MFC can outperform the cross-well power output of both
MFCs in some cases. For state II, the performance order is the same as state I except for linear oscillations now being placed
between intermittency and cross-well subharmonic oscillations due to the peak linear response being further away from
the resonance. Periodic regimes retain smooth power vs. load resistance peaks due to their consistent output while those
with aperiodic behavior are more disjointed even when the measurements are averaged over significant durations. When
comparing initial states, state II consistently has better cross-well performance with less separation between MFCs because
of its lower critical out-of-plane displacements for initiating snap through. A summary of the resistor sweep results and the
power output of each dynamic regime are presented in Table 1.1.
Since the optimal resistor load of 39.8 k is found to be common between almost all dynamic regimes across various
frequencies, backward frequency sweeps at 3 g with this load resistance is conducted to evaluate the harvester’s power output
across its first two modes. The results are shown in Fig. 1.10 for both initial states. The responses per forcing frequency
follows the trend seen in Fig. 1.3 amplitudes, where peak power is generated during limit cycle oscillations near the first
rigid body mode and tapers off away from this mode. Close to resonant conditions at 24 to 25 Hz, single-well vibrations
outperform the primarily subharmonic cross-well responses for MFC 1 in state I and MFC 2 in state II. Even with the large
jump in contribution from the other MFC during cross-well motion, the single-well near-resonance still generates more total
power, albeit over a smaller portion of the cross-well bandwidth. As the frequencies decrease and the harvester displays
more chaotic behavior, the power output quickly overtakes those of single-well regimes. Like the resistor sweep results, both
MFCs generates more power during cross-well oscillations in state II and exhibit lower power gaps across the spectrum due
to the previously discussed asymmetry. For the same reason, state II has larger bandwidth for both limit cycle oscillations
and other cross-well regimes.
1.3.5 Performance Comparison of Bistable Composite Harvesters
Given the wide array of physical characteristics and excitation levels for bistable composite harvesters, it is necessary to
utilize performance metrics that take these factors into account if fair evaluation is to be made. Size constraints of systems
involved in energy harvesting applications for small electronic components mean that minimizing the harvester dimensions
and excitation levels while generating sufficient power are of utmost importance. Differences in the acceleration input
(a) 120 (b) 140 MFC 1

MFC 2
Total
120 Single-well
100
Cross-well: Limit Cycle
Cross-well: Intermittency, Chaos, & Subharmonic
100
80
Power (mW)
Power (mW)
80
60
60
40
40
20 20
0 0
10 15 20 25 30 35 10 15 20 25 30 35
Frequency (Hz) Frequency (Hz)
Fig. 1.10 Average power output of harvester during backward frequency sweep at 3 g excitation with 39.8 k optimal resistor load for initial
(a) state I and (b) state II
between various harvester configurations must especially be considered, as this has a significant effect on their broadband
power generation. The normalized power density (NPD) [29] is a suitable single frequency metric that calculates the amount
of power per unit volume and per unit input energy as defined by Eq. (1.1).
Prms
NPD D (1.1)
Va2rms
Where Prms and arms are the RMS power output and base acceleration in m/s2 , and V is the total volume of the harvester.
Table 1.1 shows the NPD of each regime where the total harvester volume is 28.03 cm3 for an input of 3 g. A limitation to
this metric is that only one frequency can be evaluated at a time and fails to account for the broadband nature of nonlinear
harvesters. Ambient excitation signals are often distributed over a wide spectrum and so frequency sweeps are utilized
in vibration based harvesting applications. Since bistable harvesters respond with cross-well motion over input dependent
bandwidths, multi-frequency metrics are required to assess their performance and versatility. A non-dimensional performance
index [4] is shown in Eqs. (1.2) and (1.3) which accounts for the excitation frequency f and the equivalent mass of the
harvester m made up of its oscillating portion and any attached proof mass.
Prms f
ID (1.2)
ma2rms
Since the excitation is over a frequency range, the average value of I over the operating bandwidth from a to b are calculated
with Eq. (1.3).
Z
1 b
Iab D Idf (1.3)
ba a
In addition to Iab being the mean of I, the performance variability of the harvester within its bandwidth of operation is given
by the coefficient of variation (CV) [4] in Eq. (1.4).
s
Z
1 1 b
2
CV D I 2 df Iab (1.4)
Iab ba a
When considering the performance metrics described above, the ideal harvester has high valued NPD and Iab with relatively
low CV, thus implying consistently large power output over the frequency range. For the piezoelectrically generated bistable
laminate, its power output over the cross-well bandwidth at 3 g in Fig. 1.10 is evaluated. With the measured mass of the
laminate being 120 g (grams), the mean performance index Iab is 0.0185 and the CV is 0.458 from 12.5 to 23.5 Hz in state I.
Table 1.2 Energy harvesting performance comparison of bistable composite harvesters

Peak Normalized Cross-well
Boundary Piezoelectric Volume Proof Frequency Acceleration power power density bandwidth
Year Authors condition element (cm3 ) mass (g) (Hz) (g) (mW) (kgs m3 ) (Hz)
2010 Arrieta et al. [13] Fixed center QP16N 21.74 62.4 9.8 2 34 8.12 4.7
2013 Arrieta et al. [14] Cantilevered QP16N 8.94 None 20.5 0.25 5 185.92 4.5
2013 Betts et al. [15] Fixed center MFC M8585-P2 20.22 48 18 2.3 3.2 0.62 Unknown
2014 Betts et al. [16] Fixed center MFC M8585-P2 22.17 None 54 10 244 2.29 19.4
2014 Betts et al. [17] Fixed center MFC M2814-P2 5.51 None 26 9 0.18 0.01 8
2015 Li et al. [18] Cantilevered PZT-5H 3.20 15 59 3.1 0.98 0.66 0
2015 Harris et al. [19] Cantilevered MFC M8528-P2 12.11 None 16 6 7.2 0.34 7
2017 Harris et al. [20] Cantilevered MFC M8528-P2 10.29 Unknown Unknown 3 4.5 1.01 5.6
2017 Pan et al. [21] Cantilevered PZT-5H 2.66 None 38 5 36.2 11.31 8.4
2017 Udani et al. [22] Cantilevered QP16 P1 2.26 3.08 26.5 1.07 4.5 35.88 9.2
2017 Udani et al. [22] Cantilevered QP16 P1 1.76 3.08 21.7 0.64 1.2 34.91 6.7
Current work Fixed center MFC P1 28.03 None 17.5 3 130.7 10.76 15
For state II, Iab is 0.0192 and the CV is 0.579 from 11.5 to 24 Hz, indicating higher power output, but with more variability
when compared to state I. Although these metrics are a step towards fairer assessment of harvesters, limitations still exist such
as not accounting for the size of cross-well bandwidths, the effects of hysteresis between forward and backward frequency
sweeps, and the effects of coupled external circuits.
A performance comparison of bistable composite harvesters available in literature are presented in Table 1.2. Due to the
large variation of provided data in each work, some factors and metrics have to be estimated, while others are completely
unavailable. The frequency, acceleration, and cross-well bandwidth values are those associated with the reported peak power
of each harvester, and do not necessarily reflect the only parameters tested in the corresponding works. It is unfortunate that
multi-frequency metrics cannot be calculated due to insufficient Prms data over cross-well bandwidths, and the only work
that provides these results are most recently by Udani et al. [22]. Nevertheless, the comparison of NPD provides a first
glance at the relative performance of each harvester. Under comparison, the piezoelectrically generated bistable laminate
has favorable peak power output and cross-well bandwidth for a reasonable excitation level, but its NPD shows moderate
performance primarily due to its volume and acceleration input. The squared factor on the RMS acceleration in Eq. (1.1)
imply that minimizing this parameter leads to the most favorable NPD values, as seen in [14]. Despite the limitations of
utilizing P1 type MFCs, the proposed laminate shows viable broadband energy harvesting performance without the need for
inertial masses or additional piezoelectric elements. This enables the laminate to be multi-functional by letting it retain its
original snap through capability.
1.4 Conclusions
This paper presents an experimental investigation into the energy harvesting performance of a bistable laminate manufactured
by bonding two actuated MFCs in a Œ0MFC =90MFC T layup and shutting off the power afterwards. Its inherent structural
nonlinearities are exploited to induce high amplitude broadband cross-well vibrations to maximize power generation from
the MFCs. Frequency sweeps are conducted at various excitation levels to capture all linear and nonlinear responses of
the harvester and evaluate the effects of hysteresis, softening, and asymmetry of potential wells on the voltage output of
each MFC. Characteristics of the observed regimes are found through time histories, spectrum analysis, phase portraits, and
Poincaré maps of select data, which are then used to determine the response of all other sweep results. The power output of
each regime are measured through resistor sweeps and high amplitude limit cycle oscillations are found to be the optimal
dynamic response for energy harvesting. A performance comparison with other bistable composite harvesters reveal that the
laminate has viable harvesting capability, which in combination with morphing allows it to be multi-functional. Future work
include applying the harvested power from external vibrations back towards actuating the MFCs for snap through between
either stable states.
Acknowledgements This work was supported by the US Air Force Office of Scientific Research (AFOSR) under grant number FA9550-16-1-
0087, titled “Avian-Inspired Multifunctional Morphing Vehicles” monitored by Dr. B.L. Lee.
References
1. Harne, R.L., Wang, K.W.: A review of the recent research on vibration energy harvesting via bistable systems. Smart Mater. Struct. 22(2),
023001 (2013)
2. Emam, S.A., Inman, D.J.: A review on bistable composite laminates for morphing and energy harvesting. Appl. Mech. Rev. 67(6), 060803
(2015)
3. Hu, N., Burgueño, R.: Buckling-induced smart applications: recent advances and trends. Smart Mater. Struct. 24(6), 063001 (2015)
4. Pellegrini, S.P., Tolou, N., Schenk, M., Herder, J.L.: Bistable vibration energy harvesters: a review. J. Intell. Mater. Syst. Struct. 24(11),
1303–1312 (2012)
5. Tang, L., Yang, Y., Soh, C.K.: Toward broadband vibration-based energy harvesting. J. Intell. Mater. Syst. Struct. 21(18), 1867–1897 (2010)
6. Erturk, A., Hoffmann, J., Inman, D.J.: A piezomagnetoelastic structure for broadband vibration energy harvesting. Appl. Phys. Lett. 941(10),
254102–254103 (2009)
7. Erturk, A., Inman, D.: Broadband piezoelectric power generation on high-energy orbits of the bistable duffing oscillator with electromechanical
coupling. J. Sound Vib. 330(10), 2339–2353 (2011)
8. Tang, L., Yang, Y., Soh, C.K.: Improving functionality of vibration energy harvesters using magnets. J. Intell. Mater. Syst. Struct. 23(13),
1433–1449 (2012)
9. Lin, J.T., Alphenaar, B.: Enhancement of energy harvested from a random vibration source by magnetic coupling of a piezoelectric cantilever.
J. Intell. Mater. Syst. Struct. 21(13), 1337–1341 (2010)
10. Karami, M.A., Farmer, J.R., Inman, D.J.: Parametrically excited nonlinear piezoelectric compact wind turbine. Renew. Energy 50, 977–987
(2013)
11. Cottone, F., Gammaitoni, L., Vocca, H., Ferrari, M., Ferrari, V.: Piezoelectric buckled beams for random vibration energy harvesting. Smart
Mater. Struct. 21(3), 035021 (2012)
12. Masana, R., Daqaq, M.F.: Electromechanical modeling and nonlinear analysis of axially loaded energy harvesters. J. Vib. Acoust. 133(1),
011007 (2011)
13. Arrieta, A.F., Hagedorn, P., Erturk, A., Inman, D.J.: A piezoelectric bistable plate for nonlinear broadband energy harvesting. Appl. Phys. Lett.
97(10), 104102 (2010)
14. Arrieta, A.F., Delpero, T., Bergamini, A.E., Ermanni, P.: Broadband vibration energy harvesting based on cantilevered piezoelectric bi-stable
composites. Appl. Phys. Lett. 102(17), 173904 (2013)
15. Betts, D., Bowen, C., Kim, H., Gathercole, N., Clarke, C., Inman, D.: Nonlinear dynamics of a bistable piezoelectric-composite energy
harvester for broadband application. Eur. Phys. J. Spec. Top. 222(7), 1553–1562 (2013)
16. Betts, D.N., Guyer, R.A., Le Bas, P.Y., Bowen, C.R., Inman, D., Kim, H.A.: Modelling the dynamic response of bistable composite plates for
piezoelectric energy harvesting. In: 55th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference (2014)
17. Betts, D.N., Bowen, C.R., Inman, D.J., Weaver, P.M., Kim, H.A.: Investigation of geometries of bistable piezoelectric-laminate plates for
vibration-based energy harvesting. In: SPIE Active and Passive Smart Structures and Integrated Systems (2014)
18. Li, H., Dai, F., Du, S.: Broadband energy harvesting by exploiting nonlinear oscillations around the second vibration mode of a rectangular
piezoelectric bistable laminate. Smart Mater. Struct. 24(4), 045024 (2015)
19. Harris, P., Skinner, W., Bowen, C.R., Kim, H.A.: Manufacture and characterisation of piezoelectric broadband energy harvesters based on
asymmetric bistable cantilever laminates. Ferroelectrics 480, 67–76 (2015)
20. Harris, P., Arafa, M., Litak, G., Bowen, C.R., Iwaniec, J.: Output response identification in a multistable system for piezoelectric energy
harvesting. Eur. Phys. J. B 90, 1–11 (2017)
21. Pan, D., Ma, B., Dai, F.: Experimental investigation of broadband energy harvesting of a bi-stable composite piezoelectric plate. Smart Mater.
Struct. 26(3), 035045 (2017)
22. Udani, J.P., Wrigley, C., Arrieta, A.F.: Performance metric comparison study for non-magnetic bi-stable energy harvesters. In: SPIE Active
and Passive Smart Structures and Integrated Systems (2017)
23. Lee, A.J., Moosavian, A., Inman, D.J.: A piezoelectrically generated bistable laminate for morphing. Mater. Lett. 190, 123–126 (2017)
24. Lee, A.J., Moosavian, A., Inman, D.J.: Control and characterization of a bistable laminate generated with piezoelectricity. Smart Mater. Struct.
26, 085007 (2017)
25. Hyer, M.W.: Some observations on the cured shape of thin unsymmetric laminates. J. Compos. Mater. 15(2), 175–194 (1981)
26. Betts, D.N., Kim, H.A., Bowen, C.R., Inman, D.J.: Optimal configurations of bistable piezo-composites for energy harvesting. Appl. Phys.
Lett. 100(95), 114104–114117 (2012)
27. Sodano, H.A.: An experimental comparison between several active composite actuators for power generation. Smart Mater. Struct. 15, 1211–
1216 (2006)
28. Choi, Y.T., Wereley, N.M., Purekar, A.S.: Energy harvesting devices using macro-fiber composite materials. J. Intell. Mater. Syst. Struct. 21(6),
647–658 (2010)
29. Beeby, S.P., Torah, R.N., Tudor, M.J., Glynne-Jones, P., O’Donnell, T., Saha, C.R., Roy, S.: A micro electromagnetic generator for vibration
energy harvesting. J. Micromech. Microeng. 17(7), 1257–1265 (2007)
30. Gigliotti, M., Wisnom, M.R., Potter, K.D.: Loss of bifurcation and multiple shapes of thin [0/90] unsymmetric composite plates subject to
thermal stress. Compos. Sci. Tech. 64(1), 109–128 (2004)
31. Moon, F.C.: Chaotic and Fractal Dynamics: an Introduction for Applied Scientists and Engineers. Wiley, New York (1992)
32. Strogatz, S.H.: Nonlinear Dynamics and Chaos: With Applications to Physics, Biology, Chemistry, and Engineering. Westview Press,
Cambridge (1994)
33. Pomeau, Y., Manneville, P.: Intermittent transition to turbulence in dissipative dynamical systems. Commun. Math. Phys. 74, 189–197 (1980)
34. Virgin, L.N.: Introduction to experimental nonlinear dynamics: a case study in mechanical vibration. Cambridge University Press, Cambridge
(2000)
Chapter 2
Performance Assessment of Several Low-Cost Consumer-Grade
Analog-to-Digital Conversion Devices
G. R. Wetherington Jr
Abstract The Oak Ridge National Laboratory (ORNL) has pioneered an approach where low-cost consumer-grade
electronics can be used as the basis of a highly reliable data acquisition architecture. One twenty-channel system based on this
approach has been operating almost 4 years at ORNL. The architecture allows a mix-and-match level of configurability so
the designer can select devices that best match the desired performance trade-offs. The most important device to be selected
is the analog-to-digital converter. This paper explores the performance of several candidate audio recording devices that can
be used as high-speed analog-to-digital converters for such measurement systems. Various performance metrics are explored
including frequency response, noise floor, and synchronous sampling. Sample rates as high as 192 KHz are supported with
16 and 24-bit resolution. While most of the devices are dual-channel, there are multi-channel devices now available that will
allow modal type of synchronous sampling. One such device was tested and is discussed is this paper.
2.1 Introduction
Accelerometers can be used for a wide range of applications. One broad class of usage is when an accelerometer is used to
understand the nature of vibrations in a structure or rotating machine. For situations where the operating state of the measured
system is stationary (i.e., the operating state is constant and does not change), FFTs are typically used to understand the
spectral makeup of the signal. In this case, the FFT converts the time domain signal to frequency domain, which decomposes
the signal into an ordered set of sine waves where the level of the sine wave at each frequency is related to the energy at
that frequency. The process of using FFTs to understand the spectral makeup of a signal is significant in another way. By
making the assumption the time domain signal is a combination of individual sine waves we allow ourselves the option
to process that signal with any instrumentation or electronics designed to handle sine waves. The most common form of
instrumentation for sine waves is audio equipment. If the upper frequencies of interest are below 50–100 KHz, an entire class
of extremely affordable devices become candidates for conditioning and converting the signal to digital form. Specifically,
low-cost consumer grade audio devices become valid components in systems that can be used to make these measurements
[1]. ORNL has developed several systems using this approach and found it to be a reliable and cost-effective way to process
accelerometer signals [2]. The challenge for the user is these devices are not robustly specified by the manufacturers for this
type of application. At best, general audio performance characteristics may be offered. Nevertheless, if the user is able to
implement some basic performance validation steps, these devices can be used as the basis for accurate and reliable vibration
measurements at a fraction of the cost other approaches would require.
Besides its lower cost, consumer grade audio equipment is plentiful and easily accessed. It is available online from many
sources as well as from retail music equipment supply stores. This paper discussed some basic testing that was done at ORNL
to qualify a few such devices. This testing was not exhaustive but it was sufficient to categorize the devices so the user would
have knowledge about its response range and drop off characteristics.
Notice: This manuscript has been authored by UT-Battelle, LLC under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy.
The United States Government retains and the publisher, by accepting the article for publication, acknowledges that the United States Government
retains a non-exclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this manuscript, or allow others
to do so, for United States Government purposes. The Department of Energy will provide public access to these results of federally sponsored
research in accordance with the DOE Public Access Plan (http://energy.gov/downloads/doe-public-access-plan).
G. R. Wetherington Jr ()
Oak Ridge National Laboratory, Oak Ridge, TN, USA
e-mail: randallw@ornl.gov
© The Society for Experimental Mechanics, Inc. 2019 15

E. Wee Sit et al. (eds.), Sensors and Instrumentation, Aircraft/Aerospace and Energy Harvesting, Volume 8,
Conference Proceedings of the Society for Experimental Mechanics Series, https://doi.org/10.1007/978-3-319-74642-5_2
16 G. R. Wetherington Jr
Table 2.1 Audio adapters tested Device Channels

OMAPBeagle TPS65950 2
White, generic USB 2
Asus Xonar U7 2
Asus Xonar U3 2
SIIG HD digital 7.1 USB audio box 2
Presonus 1818VSL 8
Diamond extreme sound 2
2.2 Background
In 2014, ORNL deployed a novel high-speed data acquisition system that was based on low-cost consumer electronic devices.
This system provides 20 channels of 16-bit ADC sampling at 48 KHz on each channel. The ADC devices are audio adapters,
designed for stereo audio applications. Two of the stereo pairs are implemented using the audio input jacks on the two main
boards that comprise the heart of the system. The main boards are called Beagleboard-XMs. An additional 16 channels are
provided by USB Audio Adapters, which again are configured as stereo pairs and interface to the Beagleboards via USB
connections. The Beagleboards operate under the control of the Angstrom Linux operating system and they stream data
continuously to a desktop PC configured as a data receiver. The desktop PC runs Centos Linux.
The 20-channel system was deployed in March 2014 on a chilled water system at ORNL and it continues to operate almost
4 years later. Data is streamed from the Beagleboards to the desktop PC continuously. Using a custom Python application, a
5 min ring buffer is maintained in the PC desktop’s memory at all times for all channels while data archive files are produced
periodically. Three types of data files are provided for each channel; a monitor file that contains 1 s of data which is updated
each second, a trend file that is updated with 4 s of new data each hour along with a time stamp, and a 300 s long archive
file that is saved every 12 h, again for each channel. About once a month the trend and archive files are downloaded to an
analysis computer where MATLAB scripts are used to make spectral plots of the data.
Extensive testing of the Beagleboard XM was previously done in support of another DOE program. In that case, the
performance of the Beagleboard XM compared favorably to a system based on National Instruments hardware and their
LabVIEW software.
2.3 Candidate Devices
Seven audio adapters were tested as part of this work. These are listed in Table 2.1. Sample rates as high as 192 KHz are
supported with 16 and 24-bit resolution. While most of the devices are dual-channel, there are multi-channel devices now
available that will allow modal type of synchronous sampling. One such device was tested.
The testing showed that running the adapters at higher resolution than 16-bits was not justified in some cases. As a result,
the testing reported in this paper is for operating all the devices in 16-bit mode except for the Presonus which was tested at
24-bit resolution by operating in 32-bit mode.
2.4 Test Equipment
The test setup is shown in Fig. 2.1. A Tektronix AFG3122B arbitrary waveform generator was used to source the test signals.
The function generator included an Ethernet port and was VISA compatible, which allowed for the automation of some of
the tests using Python scripts that executed on the server.
The audio adapters need a host computer system on which to operate. For this testing, a series of small Linux development
boards were used for this purpose. In most cases, the data was recorded on the Linux development board and then uploaded
to the server after the test was complete.
2 Performance Assessment of Several Low-Cost Consumer-Grade Analog-to-Digital Conversion Devices 17
Fig. 2.1 Test setup
2.5 Test Results
As mentioned previously, the Beagleboard was used as the foundation for implementing a small high-speed monitoring
system at ORNL. The Beagleboard XM is unique in that was the only Linux board that was tested to incorporate its own
audio CODEC, which is based on a TPS65950 device. This device is referred to as “omap3beagle” in this paper. The other
ADC used in the monitoring system was a generic USB audio adapter that was purchased from TMART at a cost of about
$17. These were interfaced to the Beagleboard XM using its USB interfaces. The generic USB audio adapter is referred to
as “WhiteADC” in this report. Several additional devices were also acquired and tested. These are listed in Table 2.1 and
referred to in test results as Asus U3, Diamond, Presonus, SIIG, and Xonar7. All of these devices are stereo adapters except
for the Presonus which is an eight-channel device.
2.5.1 Full Spectrum Response
One of the most important tests is to evaluate the full spectrum response of the devices and their response to a known pure
tone as input. Figure 2.2 shows the baseband response for all of the test devices. A 300 Hz sine wave was used as the
excitation. The response for several of the devices is limited since the highest supported sample rate was 48,000 Hz. The
Presonus and Xonar7 both supported higher sample rates and their response extends farther as shown in the figure. What is
notable to point out is the noise floor for all the devices was quite low. From this Figure, the noise floor is shown to range
from about 130 dB to as low as 150 dB (referenced to 1 volt zero-to-peak). Also of note is the noise floor response is not
uniform across all frequencies. In particular, the Presonus and Xonar7 both exhibited humps in their noise floor response in
the upper frequency range.
Fig. 2.2 Normalized system response to 300 Hz sine wave
Figure 2.3 shows the results of the test around the 300 Hz range. All of the adapters accurately reported the test tone
energy at 300 Hz with very symmetric responses. This implies the base clock for the sampling by the adapters is fairly
accurate.
Figure 2.4 shows the 0–3000 Hz baseband responses of all the devices. From this figure, the 300 Hz test tone response
is seen as well as its harmonics. Also shown are the 60 Hz interference tone and its numerous harmonics. One important
conclusion that can be made from Fig. 2.4 is there are no sampling errors in the response. Since small development Linux
boards were used as the host devices for the audio adapters an important consideration is whether the boards can handle the
streaming data load from the audio adapters. Test files of 300 s were acquired and a very large FFT was then performed on
the data. A sampling error would effectively cause an impulse response in the spectrum which would result in a broadening
of the 300 Hz tone shape. This is usually observed as an “Eifel Tower” shaped tone response. For these tests, no sampling
errors were observed for the Raspberry Pi, Cubox, or Odroid development boards. For some audio adapters, the Beagleboard
XM did exhibit sampling errors and as a consequence those test results are not included in this paper.
2.5.2 Low-Frequency Response
Another important aspect to evaluate is low-frequency response. Tests showed that the low frequency roll-off response
varied significantly between devices as shown in Figs. 2.5 and 2.6. The worst roll-off response was for the Beagleboard
XM’s onboard CODEC and the generic “White ADC”. Both of these devices started to roll off below 50–60 Hz. On the
other hand, devices like the Presonus and Xonar7 showed very good low frequency response. Since all of the audio adapters
are AC coupled, none of them were able to go down all the way to zero Hertz. There are now some audio adapters that can
go down to zero Hertz but those were not available for this evaluation.
Fig. 2.3 Test tone response
Fig. 2.4 Test tone response

Fig. 2.5 0–250 Hz response
Fig. 2.6 0–20 Hz response

2.5.3 High-Frequency Response
The high frequency response of an audio adapter is also important to evaluate, although it can be tricky to implement due to
how the higher frequencies “fold over” onto lower frequency locations. In fact, it is effectively impossible to say for sure if an
FFT bin response is for the frequency it is assigned or a higher order component that has folded down due to aliasing unless
aliasing is prevented. A manual slew mode of the function generator frequency was used for this test. Basically, the function
generator was run up to the point it was near the Nyquist sampling frequency and then manually slewed slowly up to see
how the response tone behaved in the spectra using a real-time monitor of the spectrum data. In all cases a virtual brick-wall
filter response was provided by the audio adapters. It is believed this is due to the Sigma-Delta technology on which the
audio adapters are based which inherently provides this type of response. The benefit of the Sigma-Delta technology is no
anti-aliasing filter is needed to block frequencies above the Nyquist point. The only challenge is in the manner in which the
adapter is operated. For this testing, the audio adapters were tested at are their maximum sampling frequency. However, most
of the devices would support lower sample rates. The audio libraries used under Linux typically allow almost any sample
rate to be used with the proper driver parameters. Some earlier Linux boards, such as the Beagleboard XM, will shift the
brick wall filter with the sample rate for the integrated CODEC on the board. However, when used with external USB audio
adapters the brick wall filter stayed at the design point for the maximum sample rate which means those devices would alias
for lower sampling rates. This issue was not noticed with the other Linux development boards. It is highly recommended
users test their specific setup to verify performance and roll-off characteristics on both the high and low end of the response.
2.5.4 Synchronous Sampling
All but one of the audio adapters were stereo units which means they provided two channels each. Since the adapter designs
are based on Sigma-Delta technology both of the channels are synchronously sampled because they used the same clock.
Synchronous sampling between multiple devices is not possible because their clocks are internally generated and cannot
be externally sourced. The exception to this is the Presonus. The Presonus is an eight-channel professional audio recording
device. It too is based on Sigma Delta technology but the design allows for both internal and externally generated clocks.
All of the channels within one Presonus use the same clock thus providing synchronous sampling at the unit level. Both high
frequency sine and square waveforms were input to all eight channels of the Presonus and then the data was analyzed for
equivalence both for the primary raw data and its derivative. No observable time delay or sampling distortion was detected.
2.6 Conclusions
Seven consumer grade audio adapters were evaluated for potential application in high-speed data acquisition such as typically
used for monitoring accelerometers. The audio adapters were operating in conjunction with a small development Linux
computer board. A laboratory grade waveform generator was used to provide test signals to all the devices and the resulting
data analyzed for baseband full spectrum response, high and low frequency response, and noise floor. One eight-channel
adapter was evaluated for synchronous sampling. The adapters ranged in cost from $17 for two channels to $400 for eight
channels. The average per channel cost is close to $50. All of the devices performed well with the low-frequency response
being the main differentiator. It is important to evaluate any such device with the host computer and operating system to
verify performance characteristics since they can vary by adapter, the type of Linux board that serves as the host to the
audio adapter, and the variant of Linux used. As a result of these test, low-cost audio adapter devices are recommended for
consideration in many high-speed vibration applications where stable continuous operation of the test article is expected.
Since most audio adapters are AC coupled, DC response should not be expected.
Acknowledgment The author greatly appreciates the assistance and support that was provided by Daniel Lizama Molina, a National GEM
Consortium Fellowship student and candidate for a Ph.D. in Mechanical Engineering from the University of Puerto Rico, Mayaguez Campus.
Funding for this work was provided by the Department of Energy’s National Nuclear Security Administration’s NA-MB0504000 Defense
Program, Stockpile and Stewardship.
Another random document with
no related content on Scribd:
sprinkling of persons of a higher grade, who formed the small
Opposition. Though the opinions of these were of the most moderate
shade of what would now be called “liberalism,” the slightest
expression of them was attended with positive risk. Spies were
employed to watch such of them as had any social position; in several
cases there were trials for sedition, with sentences of transportation;
and only the impossibility of finding grounds for indictment
prevented more. The negative punishment of exclusion from office,
and from every favour of Government and its supporters, was the
least; and it was universally applied. Burns nearly lost his
excisemanship for too free speaking; and a letter is extant, addressed
by him to one of the commissioners of the Scottish Board of Excise,
in which, without denying his Liberalism, he protests that it is within
the bounds of devout attachment to the Constitution, and implores
the commissioner, as “a husband and a father” himself, not to be
instrumental in turning him, with his wife and his little ones, “into
the world, degraded and disgraced.” Part of the poet’s crime seems to
have been his having subscribed to an Edinburgh Liberal paper
which had been started by one Captain Johnstone. This Johnstone
was imprisoned after the publication of a few numbers; and the very
printer of the paper, though himself a Tory, was nearly ruined by his
connection with it. No subsequent attempt was made during the
Dundas reign to establish an Opposition newspaper. From 1795 to as
late as 1820, according to Lord Cockburn, not a single public meeting
on the Opposition side of politics was, or could be, held in
Edinburgh. Elections of members of Parliament, whether for burghs
or for counties, in Scotland, were a farce: they were transacted
quietly, by those whose business it was, in town-halls or in the
private rooms of hotels; and the people knew of the matter only by
the ringing of a bell, or by some other casual method of
announcement. Abject Toryism, or submission to Dundas and the
existing order of things, pervaded every department and every corner
of established or official life in Scotland,—the Church, the Bench, the
Bar, the Colleges and Schools; and so powerfully were any elements
of possible opposition that did exist kept down by the pressure of
organised self-interest, and by the fear of pains and penalties, that
the appearance at last from the Solway to Caithness was that of
imperturbable political stagnation.
Once, indeed, a crisis occurred which put the Scottish people
nearly out in their calculations. This was in March 1801, when Pitt
resigned office, and Dundas along with him, and a new ministry was
formed under Pitt’s temporary substitute, Mr. Addington, afterwards
Lord Sidmouth. Dundas out of power was a conception totally new to
the Scottish mind,—an association, or rather a dissociation, of ideas
utterly paralysing. “For a while,” says Lord Brougham, “all was
uncertainty and consternation; all were seen fluttering about like
birds in an eclipse or a thunderstorm; no man could tell whom he
might trust; nay, worse still, no man could tell of whom he might ask
anything.” Dundasism, which had hitherto meant participation in
place and patronage, now seemed in danger of losing that meaning;
and the bulk of the Scottish population feared that they might have
to choose between the name and the thing. They were faithful to
Dundas, however; and they were rewarded. The Addington ministry,
which had come into power principally to conclude peace with
France by the Treaty of Amiens, came to an end after that Treaty had
been rendered nugatory by the recommencement of the war; and in
May 1804 Pitt returned to the helm. Dundas, who had in the interim
been raised to the peerage by the title of Viscount Melville, then
resumed his place in his friend’s cabinet, to yield his Parliamentary
service thenceforward in the Upper House, and official service
mainly in the First Lordship of the Admiralty. Scotland then rolled
herself up comfortably once more for her accustomed slumber,—the
only difference being that her bedside guardian had to be thought of
no longer as her Harry Dundas, but less familiarly now as her Lord
Melville. So for another year; but then what a reawakening! It was in
April 1805 that, in consequence of the report of a Committee of the
House of Commons that had been appointed for the investigation of
alleged abuses in the naval service, the Whigs, through Mr.
Whitbread as their spokesman, opened an attack on Lord Melville on
charges of malversation of office, and misappropriation of public
moneys, during his former Treasurership of the Navy, either directly,
or by collusion with his principal financial subordinate. The attack
grew fiercer and fiercer, as well as more extensive in its scope; and,
although it was evidently inspired mainly by the political
vindictiveness of a party made furious by long exclusion from office,
it became more formidable from the fact that some of Pitt’s own
friends either abetted it fully or thought that the irregularities in
account-keeping which had been disclosed ought not to pass without
Parliamentary censure. Pitt reeled under such a blow at once to his
private feelings and his administration; and, after doing his best to
resist, he had to consent that Lord Melville should quit office, and
that Lord Melville’s name should be struck off from the list of His
Majesty’s Privy Council till the charges against him were formally
and publicly tried. The trial was to be in the shape of an
impeachment before the House of Lords. Before it could come on Pitt
was dead. He died on the 23d of January 1806; and the longexcluded
Whigs had then their turn of power for somewhat more than a year
in what is remembered as the Fox and Grenville ministry,—a name
accurate only till the 13th of September 1806, when Fox followed his
great rival to the grave, and Lord Grenville became Premier singly. It
was in April and May 1806, when this Fox and Grenville ministry
was new in office, that the great trial of Lord Melville in Westminster
Hall was begun and concluded. The charges against him had been
formulated into ten articles; and he was acquitted upon all the ten,—
unanimously on the only one which vitally impeached his personal
integrity, by overwhelming majorities on five of the others, and by
smaller but still decisive majorities on the remaining four. On the
whole, it was a triumphant acquittal; and it was received as such
throughout Scotland,—where, at one of the dinners held in honour of
the event by the jubilant Scottish Tories, there was sung a famous
song beginning with this stanza:—
“Since here we are set in array round the table,
Five hundred good fellows well met in a hall,
Come listen, brave boys, and I’ll sing as I’m able
How innocence triumphed and pride got a fall.
But push round the claret,—
Come, stewards, don’t spare it;
With rapture you’ll drink to the toast that I give:
Here, boys,
Off with it merrily:
‘Melville for ever, and long may he live!’”
Melville did live for some time longer, restored to his place in the
Privy Council, and rehabilitated in honour, but never again in office,
hardly caring to concern himself further with politics, and spending
his last years mainly in Scotland. He died on the 27th of May 1811, in
the seventieth year of his age.
That system of the government of Scotland by proconsulship of
which he had been so conspicuously the representative did not by
any means die with him. It was continued, with variations and
modifications, through those successive ministries of the later part of
the reign of George III. and the whole of the reign of George IV.
which fill up the interval between the death of Pitt and the eve of the
Reform Bill; nay, not only so continued, but continued with the
accompanying phenomenon that it was still a Dundas that exercised,
occasionally at least, what did remain of the proconsulship. Robert
Dundas, 2d Viscount Melville, who died as late as 1851, was a
member of most of the successive administrations mentioned, from
Perceval’s of 1809–12, through Liverpool’s of 1812–27, to Canning’s
and the Duke of Wellington’s of 1827–30, holding one or other of his
father’s old posts in these administrations, and so or otherwise
maintaining the hereditary Dundas influence in Scottish affairs while
Toryism kept the field. But, while this prolongation of the Dundas
influence in the second Lord Melville is not to be forgotten, it is the
father, Henry Dundas, 1st Lord Melville, that has left the name of
Dundas most strikingly impressed upon the history of Scotland, and
it is the stretch of two-and-twenty years between 1783 and 1806,
during which this greatest of the Dundases exercised the
proconsulship, that has to be remembered especially and
distinctively in Scottish annals as the time of the Dundas Despotism.
“The Dundas Despotism!” O phrase of fear, unpleasing to a

modern ear! What a Scotland that must have been which this phrase
describes! A country without political life, without public meetings,
without newspapers, without a hustings: could any endurable
existence be led in such a set of conditions,—could any good come
out of it?
Incredible as it may seem, there is evidence that the Scottish
people did contrive, in some way or other, to lead not only an
endurable but a very substantial and jolly existence through the
Dundas Despotism, and that not only a great deal of good, but much
of what Scotland must now regard as her best and most
characteristic produce, had its genesis in that time, though the
exodus has been later. The various liberties of the human subject
may be classified and arranged according to their degrees of
importance; and a great many of them may exist where the liberty of
voting for members of parliament and of openly talking politics is
absent. So it was in Scotland through the reign of Henry Dundas and
his Toryism. The million and a half of human beings who then
composed Scotland, and were scattered over its surface, in their
various parishes, agricultural or pastoral, and in their towns and
villages, went through their daily life with a great deal of energy and
enjoyment, notwithstanding that Dundas, and the lairds and the
provosts and bailies as his agents, elected the members of parliament
and transacted all the political business of the country; nay, out of
the lairds and the bailies themselves, and all the business of
electioneering, they extracted a good deal of fun. What mattered it to
them that now and then some long-tongued fellow who had started a
newspaper was stowed away in jail, or that an Edinburgh lawyer like
Muir was transported for being incontinent in his politics? Could not
people let well alone, obey the authorities, earn their oatmeal, and
drink their whisky in peace? Few of Scott’s novels come down so far
as to this period of Scottish life, and it has not been much described
in our other literature of fiction; but till lately there were many alive
who remembered it, and delighted in recalling its savageries and its
humours. O the old Scottish times of the lairds, the “moderate”
ministers, the provosts and the bailies!—the lairds speaking broad
Scotch, farming their own lands, carousing together, seeing their
daughters married, and writing to London for appointments for their
sons; the “moderate” ministers making interest for their sons,
preaching “Blair and cauld morality” on Sundays, and jogging to
christenings or to Presbytery dinners through the week; the provosts
and bailies in their shops in the forenoon, or meeting in the morning
at their “deid-chack” after a man was hanged! Every considerable
town then had its hangman, who was frequently a well-to-do person
that sold fish or some such commodity. And then, all through society,
the flirtations, the friendships, and the long winter evenings at the
fireside, with the cracks between the “gudeman” and his neighbours,
and the alternative of a hand at cards or a well-thumbed book for the
young folks! What stalwart old fellows, both of the douce and of the
humorous type, oracular and respected in their day, and whose
physiognomies and maxims are still preserved in local memory, lived
and died in those days and made them serve their turn! Nay, of the
Scotsmen who have been eminent in the intellectual world, what a
number belong by their birth to the reign of Dundas, and were
nurtured amid its torpid influences! Burns closed his life in the midst
of it; Dugald Stewart and James Watt lived through it; Scott, Jeffrey,
Chalmers, Wilson, Hamilton, and Carlyle are all, more or less,
specimens of what it could send forth. Vixere fortes ante
Agamemnona: there was pith in Scotland before there was
Parliamentary Reform.
Naturally it was in Edinburgh that the various elements of
Scottish life at this time were seen in their closest contact and their
most intimate union or antagonism. It was here that Dundas lived
when he was in Scotland; and here were the central threads of that
official network by which, through Dundas, Scotland was connected
with the English Government. Edinburgh was then still the chief city
of Scotland, even in population; for, though now Glasgow has far
outstripped it in that particular, then the two cities were happy in
numbering little more than 80,000 each. At least, in the census of
1801 Edinburgh stands for 82,000, or almost exactly neck to neck
with Glasgow, which stands for 83,000. Dundee, which came next,
reckoned but 29,000; Aberdeen, 27,000; and Leith and Paisley each
about 20,000. Few other Scottish towns had a population of more
than 10,000.
Was there ever another such city to live in as Edinburgh?
“And I forgot the clouded Forth,
The gloom that saddens heaven and earth,
The bitter east, the misty summer,
And gray metropolis of the North.”
One regrets that this is all that our noble Laureate’s experience of
Edinburgh enabled him to say. The east winds do bite there fearfully
now and then, and blow a dust of unparalleled pungency in your eyes
as you cross the North Bridge; but, with that exception, what a city!
Gray! why, it is gray, or gray and gold, or gray and gold and blue, or
gray and gold and blue and green, or gray and gold and blue and
green and purple, according as the heaven pleases, and you choose
your ground! But, take it when it is most sombrely gray, where is
another such gray city? The irregular ridge of the Old Town, with its
main street of lofty antique houses rising gradually from Holyrood
up to the craggy Castle; the chasm between the Old Town and the
New, showing grassy slopes by day, and glittering supernaturally
with lamps at night; the New Town itself, like a second city spilt out
of the Old, fairly built of stone, and stretching downwards over new
heights and hollows, with gardens intermixed, till it reaches the flats
of the Forth! Then Calton Hill in the midst, confronted by the
precipitous curve of the Salisbury Crags; Arthur Seat looking over all
like a lion grimly keeping guard; the wooded Corstorphines lying soft
away to the west, and the larger Pentlands looming quiet in the
southern distance! Let the sky be as gray and heavy as the absence of
the sun can make it, and where have natural situation and the hand
of man combined to exhibit such a mass of the city picturesque? And
only let the sun strike out, and lo! a burst of new glories in and
around. The sky is then blue as sapphire overhead; the waters of the
Forth are clear to the broad sea; the hills and the fields of Fife are
distinctly visible from every northern street and window; still more
distant peaks are discernible on either horizon; and, as day goes
down, the gables and pinnacles of the old houses blaze and glance
with the radiance of the sunset. It is such a city that no one, however
familiar with it, can walk out in its streets for but five minutes at any
hour of the day or of the night, or in any state of the weather, without
a new pleasure through the eye alone. Add to this the historical
associations. Remember that this is the city of ancient Scottish
royalty; that there is not a close or alley in the Old Town, and hardly
a street in the New, that has not memories of the great or the quaint
attached to it; that the many generations of old Scottish life that have
passed through it have left every stone of it, as it were, rich with
legend. To an English poet all this might be indifferent; but hear the
Scottish poets:—
“Edina! Scotia’s darling seat!
All hail thy palaces and towers!”
was the salutation of Burns, when first brought from his native
Ayrshire to behold the Scottish capital. “Mine own romantic town,”
was the outburst of Scott, in that famous passage where, after
describing Edinburgh as seen from the Braids, he makes even an
English stranger beside himself with rapture at the sight:—
“Fitz-Eustace’ heart felt closely pent;
As if to give his rapture vent,
The spur he to his charger lent,
And raised his bridle hand,
And, making demi-volte in air,
Cried, ‘Where’s the coward that would not dare
To fight for such a land?’”
Here, though it is an Englishman that is supposed to speak, it is a

Scotsman that supplies the words; but there can be no such objection
in the case of the following lines from a sonnet, entitled “Written in
Edinburgh,” by Tennyson’s friend, Arthur Hallam:—
“Even thus, methinks, a city reared should be,
Yea, an imperial city, that might hold
Five times a hundred noble towns in fee ...
Thus should her towers be raised; with vicinage
Of clear bold hills, that curve her very streets,
As if to indicate, ’mid choicest seats
Of Art, abiding Nature’s majesty,—
And the broad sea beyond, in calm or rage
Chainless alike, and teaching liberty.”
At the time with which we are concerned this city had the
advantage of containing, as has been said, only about eighty
thousand people. For comfortable social purposes, that is about the
extreme size to which a city should go. The size of London is
overwhelming and paralysing. There can be no intimacy, no unity of
interest, in such a vast concourse. Ezekiel might be preaching in
Smithfield, Camberwell might be swallowed up by an earthquake,
and the people of St. John’s Wood would know nothing of either fact
till they saw it announced in the newspapers next morning. Hardly
since the days of the Gordon Riots has London ever been all agitated
simultaneously. In Ancient Athens, on the other hand, we have an
illustration of what a town of moderate size could be and produce.
That such a cluster of men as Pericles, Socrates, Æschylus,
Sophocles, Euripides, Aristophanes, Plato, Phidias, Alcibiades,
Xenophon, and others,—men of an order that we only expect to see
now far distributed through space and time, nantes rari in gurgite
vasto,—should have been swimming contemporaneously or nearly so
in such a small pond as Athens was, and that this affluence in
greatness should have been kept up by so small a population for
several ages, seems miraculous. The peculiar fineness of the Hellenic
nerve may have had something to do with the miracle; but the
compactness of the place,—the aggregation of so many finely and
variously endowed human beings precisely in such numbers as to
keep up among them a daily sense of mutual companionship,—must
also have had its effect. In “Modern Athens” the conditions of its
ancient namesake are not all reproduced. To say nothing of any
difference that there may be in respect of original brain-and-nerve
equipment between the modern and the ancient Athenian, “Modern
Athens” is, unfortunately, not a separate body-politic, with separate
interests and a separate power of legislation. There are no walls now
round the Edinburgh territory; nor have the Edinburgh people the
privilege of making wars and concluding treaties with even the
nearest portions of the rest of Great Britain. They cannot meet
periodically on the Castle Esplanade to pass laws for themselves in
popular assembly, and hear consummate speeches beginning “O men
of Edinburgh.” But, with many such differences, there are some
similarities. Everybody in Edinburgh knows, or may know,
everybody else, at least by sight; everybody meets everybody else in
the street at least once every day or two; the whole town is within
such convenient compass that, even to go from one extremity of it to
the other extremity, there is no need to take a cab unless it rains. It is
a city capable of being simultaneously and similarly affected in all its
parts. An idea administered to one knot of the citizens is as good as
administered to the whole community; a joke made on the Mound at
noon will ripple gradually to the suburbs, and into the surrounding
country, before the evening. If such is the case even now, when the
population is over 260,000, must it not have been still better when
the population was only 80,000, and that population was more shut
in within itself by the absence as yet of telegraphs and railroads?
Moreover, the eighty thousand people who were in Edinburgh
when Henry Dundas ruled Scotland were people of a rather peculiar,
and yet rather superior, mixture of sorts. There never has been any
very large amount of trade or of manufacture in Edinburgh, nor
much of the wealth or bustle that arises from trade and manufacture.
For the roar of mills and factories, and for a society ranging
correspondingly from the great millionaire uppermost to crowds of
operatives below, all toiling in the pursuit of wealth, one must go to
Glasgow. In Edinburgh the standard of the highest income is much
lower, and the standard of the lowest is perhaps higher, than in
Glasgow; nor is wealth of so much relative importance in the social
estimate. Roughly classified, the society of Edinburgh in the days to
which we are now looking back consisted, as the society of
Edinburgh still consists, of an upper stratum of lawyers and resident
gentry, college officials, and clergy, reposing on, but by no means
separated from, a community of shopkeepers and artisans sufficient
for the wants of the place. Let us glance at these components of the
society of Old Edinburgh in succession:—
First, The Lawyers and Resident Gentry.—These two classes
may be taken together, as to a certain extent identical. From the time
of the Union, such of the old nobility of Scotland as had till then
remained in their native country, occupying for a part of the year the
homely but picturesque residences of their ancestors in the Old Town
of Edinburgh, had gradually migrated southwards, leaving but a few
families of their order to keep up their memory in the ancient capital
of Holyrood and St. Giles. In the room of this ancient nobility, and,
indeed, absorbing such families of the old nobility as had remained,
there had sprung up,—as might have been expected from the fact
that Edinburgh, though it had parted with its Court and Legislature,
was still the seat of the supreme Scottish law-courts,—a new
aristocracy of lawyers. The lawyers,—consisting, first, of the judges
as the topmost persons, with their incomes of several thousands a
year, and then of the barristers, older and younger, in practice or out
of practice, but including also the numerous body of the “Writers to
the Signet” and other law-agents,—are now, and for the last century
or two have been, the dominant class in the Edinburgh population.
From the expense attending education for the legal profession, the
members of it, till within a time comparatively recent, were generally
scions of Scottish families of some rank and substance; and, indeed,
it was not unusual for Scottish lairds or their sons to become
nominally members of the Scottish bar, even when they did not
intend to practise. The fact of the substitution of the legal profession
for the old Scottish aristocracy in the dominant place in Edinburgh
society is typified by the circumstance that the so-called “Parliament
House,”—retaining that name because it enshrines the hall where the
Estates of the Scottish Kingdom held their meetings during the last
eighty years of the time when Scotland had no Parliaments but her
own,—is now the seat of the supreme Scottish law-courts, and the
daily resort of the interpreters of the laws in these courts. Any day
yet, while the courts are in session, the Parliament House, with its
long oaken ante-room, where scores of barristers in their wigs and
gowns, accompanied by writers in plainer costume, are incessantly
pacing up and down, and its smaller inner chambers, where the
judges on the bench, in their crimson robes, are trying cases, is the
most characteristic sight in Edinburgh. Even now the general hour of
breakfast in Edinburgh is determined by the time when the courts
open in the morning; and, dispersed through their homes, or at
dinner-parties, in the evening, it is the members of the legal
profession that lead the social talk. In the old Dundas days it was the
same, with the addition that then the lawyers were perhaps more
numerous in proportion to the rest of the community than they are
now, and were more closely inter-connected by birth and marriage
with the Scottish nobility and lairds.
Of hardly less importance socially was the Academical Element.
As Edinburgh possesses a University, as its University has long been
in high repute, and as, by reason of the comparative cheapness of
board and education in Edinburgh, many families, after a residence
in England or the Colonies, have been attracted thither for the sake
of the education of their sons, or, without going thither themselves,
have sent their sons thither, the business of education has always
been prominent, if not paramount, among the industries of the city.
The teachers of the public and of other schools have always formed a
considerable class numerically, as well as in rank; while to the
University professors, partly from the higher nature of their
teaching-duties, partly from the traditional dignity conferred on
them by the great reputation of some of their body in past times, and
partly from some superiority in their emoluments, there has alway
been accorded a degree of social consideration not attached to the
same function anywhere out of Scotland. The reputation of the
Medical School of Edinburgh, in particular, has always invested the
professors in the Medical Faculty of the University with special
distinction; and, as these professors have been generally also at the
head of the medical practice of the city, the Medical element, and
with it the Scientific element, in Edinburgh society have from times
long past been, to a considerable extent, in union with the
professorial.
In all Scottish cities The Clergy have, from time immemorial,
exercised an amount of social influence not willingly allowed to any
other class of persons. This arises partly from the same causes which
give the clergy influence in other parts of Britain, but partly from the
peculiar affection of the Scottish people for the national theology
with which they have been saturated through so many centuries of
clerical teaching. In Edinburgh, in consequence of the perpetuation
there of relics of that old Scottish aristocracy which never was
completely brought into subjection to Presbytery, and in
consequence of the presence in society of a distinct intellectual
element in the lawyers, the clergy have not perhaps had, relatively,
the same weight as in other towns. Still they were powerful even in
the old Edinburgh of the Dundas rule. At the very least, a negative
respect was paid to them by the preservation throughout the place of
an external Presbyterian decorum and strictness; and in all houses
“the minister” was treated with distinction. Add to this that there
generally were among the Edinburgh clergy men possessing claims
to respect in addition to those belonging to their profession. Some,
even in that age of “Moderatism,” were remarkable for their
eloquence and zeal as preachers and as pastors; others had literary
pretensions; and others were professors in the University as well as
parish clergymen. More, indeed, than now, the professorial and the
clerical elements were then intermixed in Edinburgh. Perhaps,
however, that which gave the greatest dignity to the clerical or
ecclesiastical element in Edinburgh was the annual meeting in that
city, every May, of the General Assembly of the Church of Scotland.
In the history of Scottish society since the Union there is, perhaps, no
one fact of greater importance than the regular and uninterrupted
succession of those annual “General Assemblies” in Edinburgh for
the discussion of the affairs of the National Church. Let an
Englishman fancy that during the last two centuries there had been
no Parliament in England, no meetings of the House of Lords or of
the House of Commons, but that regularly during that period there
had been annual convocations of representatives of the whole body
of the English Clergy, together with such leading members of the
laity as churchwardens or the like from all the English parishes, and
that these convocations had sat ten days in every year, discussing all
public matters in any way bearing on the Church, and making laws
affecting the entire ecclesiastical organisation of England, and he will
have an idea of the extent to which the national history of Scotland
since her union with England is bound up in the records of her
“General Assemblies.” The General Assembly, in fact, from the year
1707 to the Disruption of the Scottish Church in 1843, was, to some
extent, a veritable Parliament, in which, though the secular
Parliament had been abolished, the united people of Scotland still
saw their nationality preserved and represented. All through the year
the clergy individually, in the thousand parishes or so into which
Scotland was divided, managed their own parochial affairs with the
assistance of select laymen called elders; these clergymen, again,
with some of their elders, held frequent district meetings, called
“presbyteries,” in order to regulate by deliberation and voting the
church affairs of their districts; there were still larger meetings,
periodically held, called “provincial synods”; but the grand
rendezvous of all, the supreme court of appeal and ecclesiastical
legislation, was the annual General Assembly in Edinburgh. The time
of its meeting was one of bustle and excitement. Black coats
swarmed in the streets; the Assembly was opened with military
pomp and circumstance by a Lord High Commissioner representing
the Crown; this Commissioner sat on a throne during the meetings,
and held levees and dinner-parties in Holyrood Palace all through
the ten days; the clergy, with lay representatives, some of whom were
usually noblemen or baronets, deliberated and debated during those
ten days, under a president of their own choosing called the
“Moderator”; the proceedings were in parliamentary form, and the
decisions by a majority of votes; and in many cases,—as in trials of
clergymen for moral or ecclesiastical misdemeanour,—barristers
were called in to plead professionally, as they did in the secular law-
courts. As was natural in a deliberative assembly almost all the
members of which were of the speaking class, the speaking was of a
very high order,—far higher, indeed, than has ever been heard in
these later days in the British Parliament; while at the same time
there was ample opportunity for the exercise of business talent and
of all the tact and skill of party-leadership. Much of the general
politics of Scotland took necessarily the form of church politics; and,
indeed, the connections between church politics and state politics
were pretty close. The vast majority of the clergy were adherents of
Dundas in general politics, and bent on swaying church polity in the
same direction; while the small minority of “Evangelicals” or “High-
Fliers,” as they were called, corresponded to the proscribed
“Liberals” in secular politics. The leading clergymen of both parties
were to be found in or near Edinburgh.
Respecting the Mercantile and Artisan classes it need only be
repeated that they were by no means separated by any social
demarcation from the fore-mentioned classes, but were intertwined
with these by family-relationships, and often also by the sympathies
belonging to superior natural intelligence and superior education.
Booksellers and printers were more numerous in Edinburgh
proportionally than in any other British town.
In a population of such dimensions, composed as has been
described, there was necessarily a good deal of leisure; and leisure
leads to sociability. Edinburgh in those days was one of the most
sociable towns in the world. By that time “society,” in the
conventional sense, had, with a few lingering exceptions, shifted
itself out of the Old Town into the New, or into the suburbs; and,
with this change, there had been a considerable change of manners.
Much of the formality, and at the same time much of the coarseness,
of an older stage of Scottish life had been civilised away,—the absurd
etiquette of the old dancing assemblies, for example, and the more
monstrous excesses of hard drinking. But the convivial spirit, and
many of the old convivial forms, remained. Dinner parties were
frequent; and the old custom of “toasts” and “sentiments” by the
hosts and the guests over their wine was still in fashion. Lord
Cockburn’s description of those dinner parties of his youth is one of
the best passages in his book. But it is on the supper parties that he
dwells with most evident affection. There were various kinds of
supper parties: the oyster supper at taverns, the bachelor supper in
lodgings, and the real domestic supper, to which both sexes were
invited; which last Lord Cockburn vaunts as a delightful institution
of Edinburgh, which the advancing lateness of the dinner-hour had
unhappily superseded. In short, in every form and way, from the set
dinner party, with its immense consumption of claret, in the houses
of the more wealthy, to the homely tea parties of gentlewomen of
moderate means, living in the suburbs of the Old Town, or in flats in
the New Town, and the roystering suppers of young men, where
culinary deficiencies were compensated by good humour and the
whisky punch, people were in the habit of incessantly meeting to
spend the evenings together. Lord Cockburn mentions, as illustrative
of the continuance of those sociable habits of the Edinburgh folks to
a somewhat later period than that with which we are immediately
concerned, the fact that for a great many years after his marriage,
which was in 1811, he had not spent above one evening in every
month, on the average, in solitude, i.e. without either being out as a
guest, or having friends with him at home. Even Sydney Smith,
though not native and to the manner born, and, with his English
tastes, more fastidious in his ideas of conviviality, retained to the last
a pleasant recollection of those Edinburgh hospitalities, as
experienced by him during his stay in Edinburgh from 1797 to 1802.
“When shall I see Scotland again?” he says in one of his letters:
“never shall I forget the happy days passed there, amidst odious
smells, barbarous sounds, bad suppers, excellent hearts, and the
most enlightened and cultivated understandings.”
Sydney Smith’s allusion to “the enlightened and cultivated
understandings” he encountered amid such roughish surroundings,
suggests the mention of what was, all in all, the most characteristic
feature of Edinburgh society at the close of the eighteenth century
and the beginning of the nineteenth—its intellectualism. In a
community composed in so large a measure of practitioners of the
learned professions, it was inevitable that there should be more of
interest in matters intellectual than is common, more of a habit of
reasoning and discussion, more play and variety in the choice of
topics for conversation. What mattered it that many of the most
intellectual men and women gave expression to their ideas in broad
Scotch? Ideas may be expressed in broad Scotch, and yet be the ideas
of cultivated minds; at all events, it was so then in Edinburgh, where
many excellent lawyers, University professors, and medical men kept
up the broad Scotch in their ordinary conversation, though the
majority had gone over to the English in all save accent, and some
were sedulous in trying to Anglicise themselves even in that. But,
whether the dialect was English or Scotch, there was a great deal of
very pleasant and very substantial talk. True, in Sydney Smith’s
recollection of the conversation of the Edinburgh people at the time
he moved among them, two great faults are specified. It ran too
much, he records, to that species of jocosity, perfectly torturing to an
Englishman, which the Scotch themselves called wut; and it also ran
too much, he records, into disputation and dialectics. “Their only
idea of wit,” he says, speaking of the Scotch generally, but of the
Edinburgh people in particular, “or rather of that inferior variety of
the electric talent which prevails occasionally in the North, and
which, under the name of wut, is so infinitely distressing to people of
good taste, is laughing immoderately at stated intervals.” And again
—“They are so imbued with metaphysics that they even make love
metaphysically: I overheard a young lady of my acquaintance, at a
dance in Edinburgh, exclaim, in a sudden pause of the music, ‘What
you say, my lord, is very true of love in the abstract, but——,’ here
the fiddlers began fiddling furiously, and the rest was lost.” This is
somewhat unfair. Wut, in its place, is as good as wit, and may be a
great deal heartier. As practised in the north, it corresponds more
with what is properly humour. It consists in a general openness to
the ludicrous view of things, a general disposition to call each other
Tam and Sandy, a general readiness to tell and to hear Scottish
stories the fun of which lies in the whole series of conceptions (often
too local) that they call up, rather than in any sudden flash or quip at
the close. At all events, the Scotch like their wut, and find it
satisfying. As for the dialectics, there is, perhaps, too much of that.
The excess in this direction is due, doubtless, in part to the
omnipresence of the lawyers. But wut and dialectics make a very
good mixture; and, dashed as this mixture is and always has been in
Edinburgh with finer and higher ingredients, there has been no town
in Britain for the last century and a half of greater deipnosophistic
capabilities, all things considered.
One element which Englishmen who do not know Edinburgh
always imagine as necessarily wanting in it never has been wanting.
Whether from the influence of the lawyers, and of the relics of the old
Scottish baronage and baronetage, acting conjointly as a
counterpoise to the influence of the clergy, or from other less obvious
causes, there has always been in Edinburgh a freer undercurrent of
speculative opinion, a tougher traditional scepticism, a greater
latitude of jest at things clerical and Presbyterian, than in other
Scottish towns. From the early part of the eighteenth century, when
Allan Ramsay, Dr. Archibald Pitcairn, and others, did battle with the
clergy in behalf of theatrical entertainments and other forms of the
festive, there has never been wanting a strong anti-clerical and even
free-thinking clique in Edinburgh society; and towards the end of the
century, when David Hume and Hugo Arnot were alive or
remembered, no city in Britain sheltered such a quantity of cosy
infidelity. Of hundreds of stories illustrative of this, take one of the
mildest:—Pitcairn, going about the streets one Sunday, was obliged
by a sudden pelt of rain to take refuge in a place he was not often in,
—a church. The audience was scanty; and he sat down in a pew
where there was only another sitter besides,—a quiet, grave-looking
countryman, listening to the sermon with a face of the utmost
composure. The preacher was very pathetic; so much so that at one
passage he began to shed tears copiously, and to use his pocket-
handkerchief. Interested in this as a physiological phenomenon for
which the cause was not apparent, Pitcairn turned to the
countryman, and asked in a whisper, “What the deevil gars the man
greet?” “Faith,” said the man, slowly turning round, “ye wad maybe
greet yoursel’ if ye was up there and had as little to say.” Pitcairn was
the type of the avowed Edinburgh infidel; of which class there were
not a few whose esoteric talk when they met together was of an out-
and-out kind; but the countryman was the type of a still more
numerous class, who kept up exterior conformity, but tested all
shrewdly enough by a pretty tough interior instinct. Indeed, long
after Pitcairn’s time, a kind of sturdy scepticism, quite distinct from
what would be called “infidelity,” was common among the educated
classes in Edinburgh. Old gentlemen who went duly to church, who
kept their families in great awe, and who preserved much etiquette in
their habits towards each other, were by no means strait-laced in
their beliefs; and it was not till a considerably later period, when a
more fervid religious spirit had taken possession of the Scottish
clergy themselves, and flamed forth in more zealous expositions of
peculiar Calvinistic doctrine from the pulpit than had been
customary in the days of Robertson and Blair, that evangelical
orthodoxy obtained in Edinburgh its visible and intimate alliance
with social respectability. Moreover, even those who were then
indubitably orthodox and devout by the older standard were devout
after a freer fashion, and with a far greater liberty both of conduct
and of rhetoric, than would now be allowable in consistency with the
same reputation. There is no point on which Lord Cockburn lays
more stress than on this. “There is no contrast,” he says, “between
those old days and the present that strikes me so strongly as that
suggested by the differences in religious observances, not so much by
the world in general, as by deeply religious people. I knew the habits
of the religious very well, partly through the piety of my mother and
her friends, the strict religious education of her children, and our
connection with some of the most distinguished of our devout
clergymen. I could mention many practices of our old pious which
would horrify modern zealots. The principles and feelings of the
persons commonly called evangelical were the same then that they
are now; the external acts by which these principles and feelings
were formerly expressed were materially different.”
Among the differences, Lord Cockburn notes in particular the
much laxer style, as it would now be called, in which Sunday was
observed by pious people and even by the most pious among the
clergy. There seems also to have been more freedom of speech, in the
direction of what would now be called profane allusion, among the
admittedly pious. One of the gems of Lord Cockburn’s book is his
portrait of one venerable old lady, a clergyman’s widow, sitting
neatly dressed in her high-backed leather chair, with her
grandchildren round her, the very model of silver-haired serenity, till
one of her granddaughters, in reading the newspaper to her,
stumbled on a paragraph which told how the reputation of a certain
fair one at the court of the Prince Regent had suffered from some
indiscreet talk of his about his own relations with her, but then
starting up, and exclaiming, with an indignant shake of her
shrivelled fist,—“The dawmed villain! does he kiss and tell?” There
were not a few old ladies of this stamp in Edinburgh in Lord
Cockburn’s boyhood and youth; some of whom survived far into the
present century, too old to part with their peculiarities, even to
please the clergy. “Ye speak, sir, as if the Bible had just come oot,”
said one such old lady, who lingered long in Edinburgh, to a young
clergyman who was instructing her on some point of Christian
practice on which she was disposed to differ from him. The
continuation in the society of Edinburgh of a considerable sprinkling
of such free-speaking gentlewomen of the old Scottish school,
intermingled with as many of the other sex using a still rougher
rhetoric, imparted, we are told, a flavour of originality to the
convivial conversation of the place for which there is now no exact
equivalent.
Presided over by such seniors, the young educated men of the
time did not stint themselves in the choice or the range of their
convivial topics. They discussed everything under the sun and down
to the centre. Who has not heard of the Speculative Society of
Edinburgh, founded in 1764 in connection with the University, and
kept up from that time to this by successive generations of students;
of which Lord Cockburn says that it “has trained more young men to
public spirit, talent, and liberal thought, than all the other private
institutions in Scotland”? Between 1780 and 1800 this society was in
all its glory, discussing, week after week, as its minutes inform us,
such topics as these:—“Ought any permanent support to be provided
for the poor?” “Ought there to be an established religion?” “Was the
execution of Charles I. justifiable?” “Should the slave trade be
abolished?” “Has the belief in a future state been of advantage to
mankind, or is it ever likely to be so?” “Is it for the interest of Britain
to maintain what is called the balance of Europe?” Here surely was
scepticism enough to keep thought alive; and that such questions,
discussed not only in the Speculative Society, but also in minor
associations of the same kind, and carried doubtless also, with other
more scientific topics, into private assemblages, should have been
ventilated in Edinburgh at that day, shows that, even under the
Dundas Despotism, there was no lack of intellectual freedom.
It is but a continuation of what we have been saying to add that
the old Edinburgh of those defunct decades had already an
established reputation as a literary metropolis. The rise of the
literary reputation of Edinburgh may date, for all purposes except
such as shallow present scholarship would call merely antiquarian,
from the time when Allan Ramsay set up his circulating library in the
High Street, and supplied the lieges furtively with novels, plays, and
song-books, including his own poems. This was about the year 1725,
when his countryman, Thomson, was publishing in London the first
portion of his Seasons. Thomson himself, and his contemporaries or
immediate successors, Mallet, Smollett, Armstrong, Meikle,
Macpherson, and Falconer, all rank in the list of literary Scots; but
they were Scoti extra Scotiam agentes, and had, most of them, but
an incidental connection with Edinburgh. The poets Robert Blair and
James Beattie, the philosopher Reid, and the theologian and critic
Dr. George Campbell, were not only literary Scots, but literary Scots
whose lives were spent on their own side of the Tweed; but, with the
exception of Blair, none of them were natives of Edinburgh, and even
Blair did not live there. After Ramsay, in short, the early literary
fame of Edinburgh is associated with the names of a cluster of men
who, born in different parts of Scotland, had, from various chances,
taken up their abode in Edinburgh, and who resided there, more or
less permanently, during the latter half of the eighteenth century.
The most prominent men of this cluster were these:—David Hume
(1711–1776), known as a philosophical writer since the year 1738, and
who, though he spent a good many years of his literary life in
England and in France, was for the last twenty years of it, and these
the most busy, a resident in Edinburgh; Hume’s senior and survivor,
Henry Home, Lord Kames (1696–1782), one of the judges of the
Court of Session, still remembered for the contrast between the
coarse Scottish facetiousness of his manners and the studied fineness
of his writings; the learned and eccentric Burnet, Lord Monboddo
(1714–1799), also a judge of Session, at whose Attic suppers in the
Old Town all the talent and beauty of Edinburgh were for many years
regularly assembled; the pompous but sensible Dr. Hugh Blair
(1718–1799), Professor of Belles Lettres in the University, and one of
the clergymen of the city; his more celebrated colleague, Dr.
Robertson the historian (1722–1793), Principal of the University, and
likewise one of the city clergymen; the minor historical writers and
antiquarians, Tytler of Woodhouselee (1711–1792), Dr. Henry (1718–
1790), Lord Hailes (1726–1792), Dr. Adam Ferguson (1724–1816),
and Dr. Gilbert Stuart (1742–1786); the poet John Home, author of
the tragedy of Douglas (1722–1808), once the Rev. Mr. Home, but
long bereft of that title, and known since 1779 as a retired man of
letters in Edinburgh; the illustrious Adam Smith (1723–1790),
settled in Edinburgh during the last twenty years of his life in the
post of commissioner of customs; the hardly less illustrious Dugald
Stewart (1753–1828), elected Professor of Mathematics in the
University as early as 1774, but thence transferred in 1785 to the
chair of Moral Philosophy, where he completed his fame; and, lastly,
not to overburden the list, the novelist and essayist Henry Mackenzie
(1745–1831), an acknowledged literary celebrity ever since 1771,
when he had published his Man of Feeling. In a class by himself,
unless we choose to associate him with the Creeches, Smellies, and
other “wuts” of a lower grade, whose acquaintance Burns made in his
leisure hours during his first visit to Edinburgh in 1786, we may
mention Burns’s immediate predecessor in the poetry of the Scottish

Download textbook Sensors And Instrumentation Aircraft Aerospace And Energy Harvesting Volume 8 Proceedings Of The 36Th Imac A Conference And Exposition On Structural Dynamics 2018 Evro Wee Sit ebook all chapter pdf

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Download textbook Sensors And Instrumentation Aircraft Aerospace And Energy Harvesting Volume 8 Proceedings Of The 36Th Imac A Conference And Exposition On Structural Dynamics 2018 Evro Wee Sit ebook all chapter pdf

Uploaded by

Copyright:

Available Formats

Sensors and Instrumentation,

Aircraft/Aerospace and Energy

Sensors and Instrumentation, Aircraft/Aerospace, Energy

Sensors and Instrumentation Aircraft Aerospace Energy

Special Topics in Structural Dynamics, Volume 5:

Biota Grow 2C gather 2C cook Loucas

Dynamics of Civil Structures, Volume 2: Proceedings of

Topics in Modal Analysis & Testing, Volume 8:

Special Topics in Structural Dynamics & Experimental

Topics in Modal Analysis Testing Volume 8 Proceedings

Evro Wee Sit · Chad Walber · Patrick Walter

Sensors and Instrumentation,

Sensors and Instrumentation,

Patrick Walter Alfred Wicks

ISSN 2191-5644 ISSN 2191-5652 (electronic)

Library of Congress Control Number: 2018941837

© The Society for Experimental Mechanics, Inc. 2019

Printed on acid-free paper

Hermosa Beach, CA, USA Evro Wee Sit

1 Broadband Energy Harvesting Performance of a Piezoelectrically Generated Bistable Laminate . . . . . . . . . . . . 1

Andrew J. Lee and Daniel J. Inman

A. J. Lee () · D. J. Inman

© The Society for Experimental Mechanics, Inc. 2019 1

1.2 Experimental Methodology

1.2.1 Design and Manufacturing

1.2.2 Experimental Setup

1.3 Results and Discussion

1.3.1 Vibration Modes

1.3.2 High Amplitude Frequency Sweeps

1.3.3 Characterization of Dynamic Responses

1.3.4 Energy Harvesting Capability

(a) 102 (b) 102 (c) 102 MFC 1

10-2 100 100

10-4 10-1 10-1

(d) 102 (e) 102 (f) 102

10-1 100 100

(a) 102 (b) 102 (c) 102 MFC 1

10-2 10-1 10-1

(d) 102 (e) 102 (f) 102

10-1 100 100

1.3.5 Performance Comparison of Bistable Composite Harvesters

(a) 120 (b) 140 MFC 1

Table 1.2 Energy harvesting performance comparison of bistable composite harvesters

© The Society for Experimental Mechanics, Inc. 2019 15

Table 2.1 Audio adapters tested Device Channels

2.3 Candidate Devices

2.4 Test Equipment

Fig. 2.1 Test setup

2.5 Test Results

2.5.1 Full Spectrum Response

Fig. 2.2 Normalized system response to 300 Hz sine wave

2.5.2 Low-Frequency Response

Fig. 2.3 Test tone response

Fig. 2.4 Test tone response

Fig. 2.5 0–250 Hz response

Fig. 2.6 0–20 Hz response

2.5.3 High-Frequency Response

2.5.4 Synchronous Sampling

“The Dundas Despotism!” O phrase of fear, unpleasing to a

Here, though it is an Englishman that is supposed to speak, it is a

You might also like