Note: Descriptions are shown in the official language in which they were submitted.
CA 02484456 2004-10-12
NON-DESTRUCTIVE TESTING OF IN-SERVICE WOODEN BEAMS
[0001] Background of the Invention
[0002] The invention relates to devices and methods for assessing the
integrity of
structures by measuring the response of the structures to applied excitation.
More
particularly, power transmission line cross-arms, vertical poles, and other
wooden
structures are tested while in service. The testing comprises applying a shock
by
physical impact, and determining the response of the subject structure to the
shock. In
one embodiment, instrumentation is placed or mounted temporarily on the
structure, for
example at one end of an elongated wood cross-arm structure. An impact is
applied at
a remote point, such as the opposite end , using a sensor equipped hammer to
apply the
shock and to trigger timing. The shock is sensed and reported to a recording
device.
Sensing and/or reporting can involve a wireless data signal path or a wireless
excitation/response path, for example using radar. Attributes of the results
correlate
with structural integrity.
[0003] A lateral span of thickness or other dimensional measurement can be
factored into the analysis of the response to the impact. A gamma ray
source/detector
pair can be used to measure opacity to radiation, which with dimensions is a
measure of
material density.
[0004] Preferably, a sharp impact is applied on the excitation side, using a
weight
coupled to an accelerometer. The accelerometer produces timing triggering
signals and
enables assessment of the hardness or softness of the structure at the point
of impact.
[0005] An acoustic wave from the impact propagates through the structure and
is
detected at a remote point, such as the opposite end. The detected wave can be
analyzed for propagation time, amplitude, dispersion and the like. A range of
measurements taken in this manner are useful to distinguish intact cross-arms
from
those having damage from one cause o r another.
Prior Art
[0006] Utility lines such as electric power transmission line conductors are
frequently
routed along roads and other paths, from successive insulated attachment
points on
-1 -
CA 02484456 2004-10-12
cross-arms attached to utility poles. Utility structures of this type may be
more or less
extensive. Frequently, the structures comprise one or more vertical poles a nd
one or
more horizontal cross-arms. The cross-arms provide positions at which
conductors can
be supported and kept at a space from one another, horizontally and/or
vertically.
There are various possible braces and struts, trellis and tee configurations,
insulator
arrangements, grounded guard wire arrangements, parabolic support cables, and
other
particulars. Many such structures are made in substantial part from wood.
Wooden
structures may comprise one species or another and may be treated or clad,
etc. but
wood structures inherently decay after a period of years. It is advisable at
times to
inspect the structures and to replace those that have deteriorated.
[0007] Such inspections are complicated by the nature of the structures. It is
advantageous, for example, if complete and dependable structural assessments
can be
made while electric power lines remain powered and in use. However, exposure
of
inspectors to risks of electric shock, falls and the like is to be avoided
where possible. It
is advantageous if structural assessments can be made quickly and
conveniently, with a
high degree of accuracy.
[0008] A typical cross-arm on a utility pole might comprise Southern yelow
pine,
dimensioned four-by four inches on a side, perhaps ten feet in length, mounted
twenty-
five feet high. Fasteners such as lag bolts may affix the cross-arm to the
mast.
Fasteners likewise support insulators for the conductors. The wood has a
number of
modes of failure, for example caused by wet rot, dry rot, insect damage,
natural
structure (e.g., knots), etc., complicated by mechanical stress from wind and
loading .
[0009] One method of assessing whether a cross-arm is "solid" is to assess the
ease
with which a piercing tool such as a n awl or screwdriver can be shoved
endwise into the
cross-arm material. This assessment technique is destructive to the cross-arm,
particularly if repeated at various positions. The process of making such an
assessment may involve damage to the mast or pole, for example if it is
climbed using
climbing spikes. The technique brings a human inspector into proximity with
the
conductors and carries risk.
-2-
CA 02484456 2004-10-12
[0010] Utility lines are frequently routed along roadsides where they are
accessible,
but also often are routed through right-of ways across cultivated fields,
forests, hillsides,
watercourses and other locations that may be inconvenient to visit, or to
visit with a n
entourage of test equipment.
[0011] It is possible statistically to assess the state of a large number of
poles and/or
cross-arms. A statistically significant sample of such structures is removed
and tested,
for example stressing the sampled structures to measure breaking strength.
From the
results, probabilities can be calcula~d as to the useful life of the full
population of
similar structures. Such information is useful in deciding when to replace all
the
structures in the population. It would be much more efficient, however, if it
became
possible dependably to identify the few structures that are on the verge of
failing, for
example due to rot or other damage, and to replace those damaged ones while
continuing to use other structures that are not damaged.
[0012] Efforts have been made to develop testing procedures for some wood
structures, especially bridges, railway beds and trestles and the like. US
Pat. 5,024,091
- Pellerin et al., for example, discloses a test bed that takes a series of
physical
measurements of a spar or other member that is loaded into the device, applies
certain
stresses, and senses certain responses. Similar tests are outlined in "'In
situ'
Nondestructive Testing of Built in Wooden Members" - Zombori, NDT.net, Vol. 6,
No. 3,
March 2001. Another example is "An Overview of the Wood in Transportation
Program
in the United States" - Duwadi, S. R., et al., Proceedings of the 5th World
Conference
on Timber Engineering, Vol. 1, p 32 (ISBN 2-88074-380-X), 1998.
[0013) Some techniques for statistical testing and/or for identifying specific
structurally defective members in situ, are not applicable to utility tower
poles and cross-
arms. Utility poles and cross-arms may be inaccessible as well as equipped
with
potentially dangerous electrical conductors. A non-contact testing apparatus
comprising
a laser vibrometer has been proposed in US Pat. 6,505,130 - Springer III, et
al. The
idea is to excite a structure in place, by acoustic wave energy, and to
measure its
vibrational response by reflecting laser light from a surface of the
structure. The
Springer technique mounts the sensing device in a helicopter and uses the
thumping
-3-
CA 02484456 2004-10-12
noise of the helicopter blades as the source of acoustic wave excitation. This
solves
inaccessibility problems. It might be expected that a hard wooden member will
produce
a sharper acoustic response than a soft or rotted one. However, even using
neural
network techniques to °learn° the acoustic response of a weak
member, it has been
difficult to obtain sufficient correlation between the vibrational response
and the
structural integrity of utility poles and cross-arms.
[0014] The foregoing prior art demonstrates conventional efforts to address
the
problems of testing and is hereby incorporated insofar as applicable to the
present
invention. What is still needed is a practical and effective way to conduct
nondestructive testing of cross-arms in the field.
Summary of the Invention
[0015] It is an object of the invention to provide a technique and apparatus
for
nondestructive testing of potentially inaccessible structures such as utility
pole cross-
arms, without removing the structures from service.
[0016] It is also an object to facilitate rapid successive testing of one or
more utility
pole structures after another along a path, using an air or ground based
vehicle to move
apparatus from one structure to the next.
[0017] According to one embodiment, testing the structural integrity of
utility pole
structures, especially wooden cross-arms, includes temporarily installing a
sensing
device on the structure, applying a n excitation signal at a position remote
from the
sensing device, and assessing the material between the two devices by
comparing the
excitation to the sensed response.
[0018] According to another embodiment, the integrity of such structures is
determined by applying an excitation signal at one position, such as by
striking the near
end of a cross-arm with a tool coupled to an accelerometer, and remotely
sensing
vibration at a remote position. This is particularly apt using radar
reflection to detect
vibration of a remote bolt or fastener in a utility pole cross-arm.
-4-
CA 02484456 2004-10-12
[0019] In one embodiment, a sensing pod is employed for making concurrent
measurements that permit assessment of attributes of the structure and/or
material,
enabling better interpretation of vibrational signal data and timing as
otherwise
collected. For example, measurements for determining physical and electrical
properties such as dimensions, density, moisture content and the like, can
assist in
assessing measurements and correlating measured values with structural
information.
Brief Description of the Drawings
[0020] These and other aspects and advantages are disclosed in or rendered
obvious by the following detailed description of examples and preferred
embodiments of
the invention, which are to be considered together with the accompanying
drawings
wherein like numbers refer to like parts and further wherein:
[0021] Fig. 1 is a schematic perspective illustration of application of the
invention to
the testing of a utility pole cross-arm.
(0022] Fig. 2 is a block diagram illustration of a n exemplary sensor head
adapted to
be temporarily installed at one position on the structure to be tested, such
as on one
end of a spar or cross-arm, to take certain measurements and/or to detect
certain
signals that are otherwise generated.
(0023] Fig. 3 is a block diagram illustration of an excitation and sensing
unit that can
be applied at another position on the structure, and is useful together with
the sensor
head shown in Fig. 2. For example, the excitation and sensing unit can be
applied at
the opposite end of a cross-arm from the sensor head, to develop an impact
signal that
is to be detected by the sensor head, and also to take further measurements.
[0024] Fig. 4 is an overall block diagram illustration showing the functional
connections of the respective elements.
[0025] Fig. 5 is a graph showing a hammer impact acceleration profile in the
heavy
line trace (at 10% scale) and the resulting signal detected at the remote pod,
the pod
accelerometer being screwed into the beam.
_5_
CA 02484456 2004-10-12
(0026] Fig. 6 is a graph corresponding to Fig. 5, likewise showing the hammer
impact and detected signals, with the pod accelerometer clamped to the beam.
[0027] Fig. 7 is a graph showing the correlation of received impact energy
versus
strength in experimental results.
[0028] Fig. 8 is a graph showing the correlation with strength of a composite
variable, namely the product of density and the square of shock wave velocity.
[0029] Fig. 9 is a photograph showing a housing structured for centered
frictional
attachment to an end of a spar.
[0030] Fig. 10 is a photograph showing the housing as in Fig. 9, attached to
the end
of a spar.
[0031] Fig. 11 is a schematic illustration of an alternative particularly for
airborne
use, wherein the housing comprises a vertical extension such that a helicopter
bringing
the housing into position is displaced vertically relative to the cross-arm.
[0032] Figs. 12 and 13 are schematic illustrations of a radar sensing
technique
according to a further embodiment.
Detailed Description of Preferred Embodiments
[0033] This description of preferred embodiments is to be read in connection
with the
accompanying drawings, together forming the description of the invention and
illustrating certain nonlimiting examples. The drawing figures are not
necessarily to
scale and represent some features in schematic form in the interest of clarity
and
conciseness.
[0034] Referring to Fig. 1, it is desired to assess the structural integrity
of a structure
20 that remains in service, without damaging the structure or substantially
risking
damage. In the arrangement shown, this is to be accomplished by placing a self
contained sensor unit 22 at one point on the structure, especially at one end
of a cross-
arm 21 on a utility pole 25, and applying an exciting signal at a different
point on the
structure such as the opposite end of the cross-arm 21. The exciting signal is
coupled
-6-
CA 02484456 2004-10-12
to the sensor unit 22 throug h the structure 20. Assessment of the structure
20 is made,
based on the signal received at the sensor unit or "pod" 22. Preferably these
measurements include data that relates the exciting signal to the received
signal, and
optionally data that particularly define the structure 20 such as the
dimensions, density
or other aspects of cross-arm 21.
[0035] The invention is apt for testing utility pole cross-arms 21 from a
vehicle such
as a helicopter 30, and is applicable to other structures and access
techniques. Cross-
arms on a series of utility poles 25 often comprise the same sort of material
(e.g., the
same species of wood), are of the same dimensions, and are similarly mounted
and
loaded. This makes it possible to detect cross-arms that are failing from
undue
weathering, rot or the like, by comparing measurements of the cross-arms
against
nominal measurements and statistics. Cross-arms 21 that are in service often
are not
easily accessible for individual assessment measurements, being mounted some
distance above the ground and on poles 25 that are some distance apart along
the
route of the utility service 27. The invention provides a method and apparatus
to
facilitate measurements that can be managed using an airborne vehicle such as
helicopter 30, or alternatively a ground-based vehicle such as a boom truck
(not
shown). The vehicle can be used to mount a sensor unit 22 and to bring to bear
a
means 33 (shown generally in Fig. 1 ) for applying an exciting signal to be
detected by
the sensor unit22, among other measurements.
[0036] In Figs. 1 and 2, the sensor pod 22 can be temporarily installed on the
cross-
arm 21 using vehicle 30, the same vehicle being moved for thereafter applying
the
exciting signal. Two vehicles or other techniques can be employed. In an
alternative
embodiment, discussed herein, the excitation and the measurements can be
effected
from the same end of the cross-arm 21, for example using a radar vibration
sensor
trained on a remote point on the structure 20 to collect data on the signal
passing
through the structure.
[0037) In the embodiment shown in Figs. 1 and 2, helicopter 30 is employed to
bring
a self-contained sensor unit 22 up into engagement with the cross-arm 21. The
sensor
unit 22 is carried in a housing having measurement and attachment mechanisms
such
-7-
CA 02484456 2004-10-12
as shown in Fig. 2. An air operated or spring-biased movable clamp 42 can be
used to
grasp or release the pod 22 on the cross-arm 21 by action of the vehicle 30,
e.g., by
coupling a pneumatic pressure source to open the clamp against spring pressure
when
the sensor unit 22 is placed and removed. Simple spring mounting also can be
used,
but for sensing vibration in the cross-arm 21, it is advantageous to provide a
sensor 43
such as an accelerometer or acoustic input device or the like, that is placed
in contact
with the cross-arm 21 for receiving the impact signal coupled through the
cross-arm 21
or other structure from a remote point. Fig. 3 generally shows application of
an impact
using a mallet or hammer 45 coupled to an excitement-side accelerometer 47
associated with an exciter "head" 48. See also Fig. 4. The exciter head can
involve a
modular device similar to the sensor head 22, or it is possible simply to
employ a
hammer 45 with an accelerometer 47 rigidly attached thereto and wired to a
signal and
data processor 50 as shown in Fig. 4.
[0038] The sensor unit 22 preferably is in short range wireless radio
communication
with a data collection processor unit 50 carried on the vehicle 30, and
contains a
transmitter 51 for this purpose. Alternatively, the processor 50 can be
remotely located
and otherwise in data communication with the excitation and measurement units
48, 22,
such as over a cellular or other wireless communication path.
[0039] Independently or in connection with human intervention, the excitation
head
48 and the sensor pod 22 effect certain excitation and measurements, and
report either
raw or processed data to the data collection processor 50. In one embodiment,
the
sensor pod 22 reports over one or more telemetry communication channels. For
example, a signal from accelerometer sensor 43 (see Fig. 2) can be digitized
and
reported in time division multiplexed or packet communications with the data
collection
processor 50. The sensor pod 22 also can report values for certain additional
parameter measurements that assist in interpreting the impact or vibrational
signal
passed through the cross-arm 21.
[0040] In the embodiment of Fig. 2 and Fig. 3, the sensor unit 22 contains
accelerometer 43 for collecting an impact signal coupled through cross-arm 21
from the
remote impact of hammer 45. Sensor unit 22 in this case also has a paired
gamma
_g_
CA 02484456 2004-10-12
radiation source 53 and detector 55 for measuring gamma opacity of the cross-
arm, and
a thickness measurement caliper 57.
[0041] The sensor unit or pod 22 preferably has an analog to digital converter
associated with the transmitter for providing the transmitted data in digital
form. The
pod can have power supplies, other onboard circuits and a local processor (not
shown),
for collecting and reducing measurements to processed results such as
calculated
averages or the like, to be reported to the data collection processor 50. As a
further
alternative, the sensor unit can contain a processor that stores rather than
contemporaneously transmits the data, and uploads data to the processor later
together
with time base information.
[0042] In an embodiment involving communications for reporting data to a
processor
50 on a nearby vehicle 30, the communications can be one-way, i.e., with the
sensor
unit broadcasting a continuous stream of telemetry in a time division
multiplex manner.
Alternatively, the communications can be two-way, for example with the data
collection
processor 50 controlling the sensor unit such as polling or triggering
transmission of
particular data. For example, the processor 50 can be triggered to commence a
measurement sequence upon sensing of an impact at the hammer 45 by
accelerometer
47 (Fig. 3), and then signal to trigger measurement and transmission of
samples from
sensor accelerometer 43 for a predetermined time interval thereafter as needed
to
sample the resulting signal at the remote end of cross-arm 21.
[0043] Accordingly, certain of the measurements taken by the sensor pod 22 are
measurements that record or assess the result at sensor pod 22 ofan exciting
signal
applied by exciting unit 48 that can be directly coupled to the data
collection processor
50 on the vehicle, whereas the pod 22 is in wireless communications. The
process
therefore is first to mount the self-contained sensor unit 22 to the cross-arm
21 using
the vehicle 30, i.e., on one end of the cross-arm. Next the vehicle is
repositioned to a
remote point, preferably the opposite end of cross-arm 21. The exciting unit
48 is then
applied to such opposite end (or to some other strategic point) and the data
is collected.
Afterwards, the sensor unit 22 is recovered, and presumably moved to a new and
different cross-arm 21 that is to be assessed.
_g_
CA 02484456 2004-10-12
[0044] The impact from hammer 45 imparts a force impulse over a predetermined
area of the cross-arm 21 or other structure and produces a corresponding
electrical
signal in hammer accelerometer 47. This signal from accelerometer 47 defines
the
character and timing of the exciting impact. The signal from hammer
accelerometer 47
indicates when and how hard the hammer was struck to develop the excitation
impact,
and also indicates the reaction of the cross-arm to the exciting blow, i.e.,
the hardness
or softness of the material encountered at the point of impact. The exciting
signal from
accelerometer 47 is useful for several purposes including as a triggering
signal
providing a timing reference, a measurement of material characteristics, to
define the
impact or vibrational input signal that propagates through the cross-arm 21
toward the
sensor pod 22, and is affected by passage through the structure.
(0045] Data respecting other parameters also can be collected at the exciter
head
48, such as density and dimensions (alternatively collected on the sensor pod
end in
Fig. 2). The data collected at the data collection processor preferably
includes at least a
representation of the exciting impact (at least a triggering signal developed
from the
output of hammer accelerometer 47) and the corresponding remote vibrational
result
(such as the output of sensor accelerometer 43).
[0046] Moving vehicle 30 from one utility pole and/or cross-arm to the next,
and
optionally from end to end on the cross-arms, the operators proceed to assess
the
structural integrity of a series of cross-arms, for example collecting
excitation and
response waveform data for all or for a sampling of cross-arms over a given
section of
the route of the utility line. The data for each measurement preferably is
associated
with a specific utility pole and cross-arm, to enable change-outs of specific
cross-arms
found to be in need of replacement. Alternatively, the data can be applied
generally to
assess the structures in connection with planning a more substantial project
involving
assessment and planning for refurbishing or replacing a number of poles and
cross-
arms.
[0047] The self contained sensor pod unit 22, shown schematically in Fig. 2,
can
include pneumatic or mechanically retractable clamps that hold the unit 22 to
an end of
the cross-arm or other beam being evaluated. The sensor unit 22 contains an
-10-
CA 02484456 2004-10-12
accelerometer 43 that should be held in intimate contact with the end of the
cross-arm
21 to provide a robust signal. For this purpose, the pod 22 can be
mechanically
coupled for contact with the end of the cross-arm using a structure that
centers and
optionally draws the sensor pod 22 toward the cross-arm. The accelerometer can
be
spring mounted in the pod 22 to bear against the end of the cross-arm under
some
spring pressure.
[0048] Figs. 5 and 6 compare the excitation and response shock waves detected
by
a mounted sensor pod. The delay between the leading edges of the excitation
and
response represents the shock wave travel time, and with beam length can
determine
the shock wave velocity. The shock wave can also be detected by radar as
discussed
herein. Fig. 7 shows a weak correlation of received energy to beam breaking
strength.
Fig. 8 shows a stronger correlation of density * shock velocity ~2 versus
breaking
strength.
[0049] In Figs. 9 and 10, the housing of a sensor pod 22 is shown and
generally
comprises a cup-shaped assembly sized to fit over the end of cross-arm 21. The
base
62 of the cup shape holds the accelerometer 43 (not shown in Figs. 9 and 10).
The rim
63 of the cup has inwardly inclined leaf springs that admit and grasp a range
of cross-
arm diameters. The accelerometer is spring mounted in base 62 so as to be
displaceable against spring bias for a short distance, e.g., two cm or so,
sufficient to
permit the base to be placed adjacent to the end of cross-arm 21, or to
accommodate
an irregularly cut end on cross-arm 21, while obtaining contact between
accelerometer
43 and the end of the cross-arm 21.
[0050] Fig. 11 shows a preferred mounting technique for placing pod 22 on the
cross-arm 21. In this arrangement, the operator (not shown) can sit in an open
doorway
of helicopter 30 or on a platform mounted to the helicopter skids. The pod 22
has an
associated vertical extension 72 whereby the working level for affixing pod 22
to cross-
arm 21 is displaced downwardly. This allows the helicopter to hover at an
elevation that
keeps the helicopter rotors high, for example well above the top end of a mast
or pole
25 carrying the cross-arm 21. The extension 72 can be affixed permanently to
sensor
pod 22 or can be a removable appliance for manipulating pod 22. The extension
72 can
-11 -
CA 02484456 2004-10-12
have handles at the upper end and optionally can carry controls that are
coupled to pod
22 to assist in placing the pod snugly against axial end of the cross-arm 21.
[0051] In an embodiment as in Fig. 1 having a sensor pod 22 at the end of
cross-arm
21 remote from the point of impact from a hammer, the sensor pod can report
measurements to processor 50 over the telemetry link. As alternative, a sensor
pod 22
can be mounted at the impact side and wired to the processor 50 to report
gamma
opacity and thickness (which can be used as a measure of material size and
density).
As another alternative, shown in Figs. 12 and 13, the entire measurement can
be taken
from the impact side. In this case, the helicopter 30 has a radar dish 82 used
to direct a
beam for reflection by a structure subject to vibration from the impact
signal, such as a
bolt 85 at a remote point on the cross-arm 21. The operator in this case
activates the
radar illumination via dish 82, strikes the end of cross-arm 21 with hammer 45
and
records the vibrational result at bolt 85 as the impact wave propagates
through the
cross-arm 21.
[0052] Referring again to the embodiment of Fig. 2, pod 22 comprises a gamma-
ray
densitometer, namely a source 53 and detector 55 on opposite sides of cross-
arm 21
when the pod is mounted. The detector 55 can count gamma rays from the source
over time, or otherwise determine the gamma-opacity of the beam in a lateral
direction.
Often, the lateral thickness of the cross-arm is a known factor. If the
thickness is not
predetermined, a ruler or caliper device may be closed on the cross-arm 21 to
measure
the width of the beam in the same direction that the densitometer is operating
. The
gamma opacity and dimensions together are factors used to determine material
density.
This data is reported to processor 50.
[0053] A real time sampling telemetry technique was described above for
transmitting information to an operating system on the vehicle. The data
reported to
processor 50 can be transmitted immediately upon sampling or reduced by
processing,
stored, relayed in original or modified form or otherwise handled according to
any of
various specific data collection, communication and processing arrangements.
However
an object is to obtain at least a representation of the impact signal and the
received
response at a remote point on the structure, for comparison and processing to
assess
-12-
CA 02484456 2004-10-12
structural integrity. The collection technique can be varied, such as whether
the data is
collected using a contact sensor, radar or the like. The particulars can
involve no
mounted pods, a pod at the impact end or the remote, or a pod at both ends,
each
being more or less sophisticated as to the measurements taken and the method
by
which the results are communicated.
[0054] Nevertheless, at least a n impact is physically applied to the cross-
arm or
other structure (the "beam"), preferably on an end of the beam if an end is
accessible,
but at least at another position that is spaced from the point at which
resulting vibrations
are to be sensed. The impact propagates through the material of the beam to
the point
where a resulting impact wave is sensed or at least its timing and similar
characteristics
are detected. Preferably, the particulars of the impact are encoded at the
point of
application of the impact as well, but at least the timing of the exciting
impact is noted.
[0055] A simple hammer 45 can be instrumented with accelerometer 47 and swung
manually to apply a n impact to the beam 21. A possibly more repeatable
mechanism
can be used to apply the impact, such as to produce a predetermined motion in
a mass
of a given size and shape, but a simple manual blow has proven to provide good
results
notwithstanding variations between one blow and another. Variations in the
nature of
the material also affect whether the impact from the blow, e.g., whether the
hammer
bounces relatively sharply to "knock" on a solid beam, or dully "thuds:
against one that
is softened by wear, age or rot. These characteristics also affect the
propagation of the
impact through the beam.
[0056] One or more blows or impacts are used to insert one or more successive
impact shocks into each tested beam. The impact applicator is shown
schematically in
Fig. 3 as a hammer, and satisfactory test results have been achieved for beams
of the
size of utility pole cross-arms, using a small sledge-type hammer tool, for
example of
three to five pounds weight, swung comfortably hard against the beam. A simple
hammer with an accelerometer attached to the head can produce the desired
impact
signal when manually struck against the beam under test.
[0057] The processor 50 or other data acquisition unit processes time-
synchronized
signals from the two accelerometers or data-reduced versions of such signals,
namely
-13-
CA 02484456 2004-10-12
from the sensing point and from the exciting point. Associated information can
also be
measured or entered, such as density information from the gamma ray
densitometer,
dimensions as measured or estimated, a location or serial number identifying
the
specific pole or cross-arm, etc.
[0058] The data acquisition and/or data collection processor unit 50 extracts
or
operates upon certain discriminating parameters and factors from the collected
data.
The discriminants preferably include the shock wave speed or the propagation
time of
the shock wave through the beam; an index or measure of the hardness of the
beam on
the struck end; and a ratio of the apparent energy exerted by the hammer blow
and the
energy of the shock signal received by the sensor unit at the far end of the
beam.
Additional parameters and factors also can be sensed, such as measured or
inferred
density.
[0059] The discriminants can be further processed, for example via multiple
regression analysis, or logical analysis such as a selection or evaluation of
disciminant
values, to estimate beam strength or to assist in ranking for distinguishing
weaker
beams from stronger ones. The data can be analyzed for the apparent presence
of
decay or voids in the beam. An alternative logical analysis, for example, can
involve
ranking beams comparatively in order of expected strength, which is applicable
to
identify relatively weaker beams regardless of absolute breaking strength.
[0060) A technical description of a strength prediction system according to an
exemplary embodiment follows below, including a discussion of measurement data
that
has been taken on sample beams, although not in service, for demonstrating the
nature
of the analysis and its capabilities. It should be appreciated that this
specific system
may be enlarged, abbreviated or othervuise modified within the scope of the
invention.
[0061] The system was configured in particular to predict the breaking
strength (and
evidence of decay) of wooden power-pole cross-arms in situ, using a helicopter
as a
testing platform to facilitate moving from one supporting structure to another
during
testing. As stated above, other platforms are possible, such as boom trucks,
etc. In
this embodiment, cross-arms for refativety high voltage transmission lines
were the
planned subject structures under test. Such cross-arms are typically mounted
-14-
CA 02484456 2004-10-12
horizontally on two spaced mast poles. The cross-arms (sometimes basically de-
barked
tree trunks known as spar-type cross-arms), often have some taper from a
larger
diameter end to a smaller diameter end and may be, for example, up to 35 feet
or more
in length. The invention is also applicable to smaller cross-arm structures
such as
rectilinead~trimmed spars and the like, for example on the order of ten feet
in length or
less.
[0062] The estimated breaking-strength of a cross-arm is considered to provide
a
reasonably dependable measure of the likely remaining lifetime of the cross-
arm. With
a further knowledge of other factors such as the potential cost of a failure,
the expense
of replacement and the like, the invention provides a measure by which a
utility can
reasonably assess whether it would be of economic advantage to replace one or
more
cross-arms with new cross-arms or otherwise to plan ahead. Performing the
assessment from a helicopter allows rapid and relatively inexpensive testing
in
comparison to requiring linemen to travel to individual power poles using
ground
transportation and to climb each pole for evaluation of the strength of the
cross-arms.
[0063] Breaking-strength can be defined as that force which must be applied to
a
wooden cross-arm to produce 100 percent failure, i.e., structural breakage.
Destructive
testing of a statistical sample of cross-arms, i.e., loading to the point of
failure, can be
used to determine the breaking-strength of new or aged cross-arms and to
assess the
breaking-strength of cross-arms that may fail if subjected to forces of a
similar degree,
e.g., by ice or wind-loading on the structure and/or on the electrical
conductors it
supports.
[0064] The breaking-strength of cross-arms is determined and the degree of
expected structural loading are quantified. This information is then related
to
measurable aspects of cross-arms that are in service, which aspects are
measurable
nondestructively. The nondestructive test results are used to estimate the
breaking-
strength of a cross-arm that is in service, to ultimately enable a judgment as
to the
probability of failure of that cross-arm and to plan if necessary for its
replacement.
(0065] An exemplary system comprises the helicopter-borne unit 50 for data
collection, processing and storage, the instrument pod unit 22, comprising a
self-
-15-
CA 02484456 2004-10-12
contained and powered set of sensors, and a device 45, 47 for applying an
impact
shock wave that propagates through the subject cross-arm 21 . The helicopter-
borne
unit in this arrangement contains a data-collection and storage system (a
computer and
associated instrumentation), coupled to an instrumented hammer or mallet, an
air-
storage tank, and connecting air hose and cabling for operating a mechanism
that
temporarily affixes the sensor pod 22 to the cross-arm 21.
[0066] The operator manually strikes the end of the spar arm, causing a shock
to
propagate down the arm. Preferably, spar arms under test are struck
consistently with
an axially directed blow to one end of the spar. The propagating shock is
sensed at the
opposite end. In a situation where the end of the spar is not accessible, a
shock can be
applied by a lateral blow. This technique can be used to test the integrity of
vertical
poles.
[0067] The accelerometer coupled to the hammer (specifically to the hammer
head)
senses the shock of deceleration of the hammer head upon contact with the
cross-arm.
This signal provides a timing trigger and also is recorded or digitized over
time as data
that characterizes the force of the blow and the softness or hardness or the
struck end
of the spar. This data is recorded as a function of time, by the data-
collection apparatus
and the computer, and the results are stored. The force of the shock is
variable based
on the speed with which the hammer is swung against the end of the spar arm.
The
sharpness and decay particulars of the shock varies with the character of the
spar.
These variables become determinants in the subsequent data analysis.
[0068] A preferred computer applicable to the system is a n environmentally
hardened laptop PC. The computer or other controlling apparatus associated
with the
system can be used simply to record information, or the same computer can be
used to
control data recordation, to process or analyze the data, and can maintain all
or part of
a database of information that contains additional useful data including, for
example, a
serial number and/or location coordinate set that identifies the cross-arm,
various
atmospheric variables at the time of the test and other factors that may be
useful or may
affect the data analysis.
-16-
CA 02484456 2004-10-12
[0069] The vehicle (e.g., helicopter) can carry a compressor or an air storage
tank,
for example pre-pressurized to 100 pounds as a mission preparation step,
coupled to a
pressure regulator that produces a predetermined working pressure. For
example, a
regulator coupled to the tank can be used to meter or regulate air pressure at
60 psi to
an on-board pneumatic supply hose that can operate the sensor clamp, and
optionally a
pneumatic hammer for producing the excitation impacts.
[0070] The air-hose connects the air-tank to the instrument pod and is used to
open
or close a clamping mechanism that attaches the instrument pod or sensor to
the cross-
arm, such as clamp 42 shown in Fig. 2. The air can be coupled to a cylinder
that opens
clamping jaws against spring pressure whe n coupled to the air supply, and
releases the
jaws to clamp to the spar when released from the air supply.
[0071] The instrument pod may be coupled to the data collection processor by
cable,
radio link or other communication path. The instrument pod can communicate
with the
data collection processor by radio. On the opposite end of the spar, the
hammer and
accelerometer can be wired to the data collection instrumentation and
processor, e.g.,
by a coaxial cable or other noise-insensitive connection (e.g., shielded
twisted pair).
[0072] The data collection system may be operated from the aircraft DC power
supply, typically at 24 volts DC. It is also possible to use battery power.
The sensor or
instrumentation pod preferably is self contained and can be powered from an on-
board
battery.
[0073] In one embodiment, the operator has certain inputs for controlling the
operation of the data collection system. A multi-conductor cable can connect
the data
system to a hand-held control box, to be used by the operator to provide
commands to
the data-collection system, e.g., to commence a data collection cycle
including
application of an impact.
[0074] The instrument pod 22, which is a self contained unit, can be manually
handled for placement on the spar end . The pod also can be carried on an
elongated
handle (Fig. 11 ), for manipulation at a level below the normal working level
of an
operator on a helicopter. The instrument pod can have an air-operated clamp
assembly
for urging the pod 22 against the cross-arm 21, a spring-loaded accelerometer
with a
-17-
CA 02484456 2004-10-12
charge-amplifier, a gamma densitometer with power-supply and pulse-shaper, and
radio-telemetry broadcast capability, with associated amplifier, modulator and
antenna
equipment.
[0075] The instrument pod can have optional external controls and readouts. In
order to make a cross-arm structural assessment, the operator can be required
to
attach and activate the instrument pod, which can involve making one or more
connections or operating switches. In an embodiment wherein the pod is clamped
to
the spar during measurement, the pod is temporarily connected to the pneumatic
supply
by the air supply hose. The primary function of the pneumatic supply to the
pod is to
operate the clamps. The pod can be supported temporarily, for example using an
elastic cord tied to the helicopter or other vehicle, while maneuvering into
position.
[0076] Powering-up the instrument pod can turn on the accelerometer amplifier.
The
scintillator or gamma ray detector is powered, as well as the telemetry
transmitters. The
instrument pod may have an onboard microprocessor or logic network for data
handling
and optionally also for preliminary data processing. The onboard processor (or
another
controller) can control the other components such as the transmitters, the
clamping
devices, etc. The processor can store raw or processed data or can process the
data to
the point of reaching results, potentially even a pass/fail indication.
Preferably,
however, the processor reports over a data communication path to the system
computer on the aircra~ and the analysis is undertaken with the benefit of
other data
including the specific data on the impact shock that was applied.
[0077) In the embodiments having a sensor pod 22 remote from the point of
impact,
after the data has been collected, the helicopter is maneuvered back to the
opposite
end of the spar to retrieve the instrument pod. It is possible to have plural
selectably
operated pods 22 to reduce the inconvenience of retrieval. According to the
embodiments of Figs. 12 and 13, if it is possible to dispense with measurement
of
density and cross-arm dimensions, no pod is necessary. These parameters may
improve the correlation of shock wave propagation data with breaking strength,
but are
not absolutely necessary in every case.
-18-
CA 02484456 2004-10-12
[0078] The mobility of a helicopter platform provides inherent efficiency in
dealing
with utility pole cross-arms, but temporarily mounting the instrument pod 22
on the
wooden beam before changing positions to apply a n impact is relatively
expensive in
time and effort. Accordingly, the technique of Figs. 12 and 13 has been
developed for
non-contact remote sensing of the vibrational results at a remote point along
a cross-
arm, from a shock or impact applied at a nearby point. The technique uses
continuous-
wave (CW) K-band radar illumination of a remote structure, especially a
fastening bolt
85, to produce a reflected signal that is modulated by the vibration of the
bolt 85 due to
the vibrations. "K-band" radar uses a radio wavelength of about 1.5 cm, and
has been
found efficient for sensing shock vibrations but not sensitive to small
variations in the
smoothness of surfaces and the like. Sensitivity to surface variation is a
problem when
attempting laser vibrometry at sub-micrometer wavelengths, particularly when
aiming
from an unstable platform such as a helicopter. K-band radar uses a wavelength
that is
10,000 times larger than the laser wavelength, making material textures
substantially
irrelevant.
[0079] The K-band continuous wave radar is aimed by a dish 82 specifically at
a bolt
85 on the far side of the cross-arm. A CCD camera (not shown) can be mounted
on the
dish 82 to assist in aiming the radar beam at the bolt 85, and optionally to
record and
store the image of the bolt or other target for later review if the data
appears odd. The
shock induced vibration of bolt 85 causes a phase variation in the reflected
signal from
bolt 85, which is received back at the radar dish 82 as the vibration input
signal.
[0080] The shock signal is still initiated by a manually swung hammer. A
transverse
density measurement can be taken at the end of the cross-arm closest to the
helicopter
30, e.g., with the same sort ofgamma-ray conduction densitometer discussed
hereinabove. Using the radar method, there is no need to move the helicopter
from one
end of the cross-arm to the other. However it is generally desirable to have
two
operators, namely one to strike the wood end with the instrumented hammer 45
while
another aims the radar dish at the fixed bolt 85 or other hardware on the far
end of the
cross-arm. The dish can be aimed over the shoulder of the person wielding the
hammer 45, to illuminate the bolt 85 and to collect the reflected phase-
modulated
signal. Any radar reflections off the wooden cross-arm as opposed to bolt 85
are at a
-19-
CA 02484456 2004-10-12
shallow angle, and scatter forward. It is difficult or impossible to detect
vibration at the
remote end of cross-arms that lack a bolt or some other remote reflecting
structure,
However, a cross-arm typically has a vertically mounted bolt holding the
insulator for
supporting a conductor, at the far end, which produces the reflected phase
modulated
continuous wave signal received by the dish 82.
[0081] Transverse density measurements may be possible using other sensing
techniques, perhaps including radar, transmitted microwaves or another
technique in
lieu of gamma ray opacity measurements. Alternatively, the dimensions and
density of
the cross-arms can be assumed and full reliance placed on vibrational
measurements.
[0082] The operator can be provided with the capability to review the test
results or a
summary thereof. The data collection computer can be programmed to signal if
problems were detected such as lack of a signal or data out of expected valid
ranges. If
the operator concludes that a successful measurement has been taken, the
operator
can validate the measurement and associate it with the cross-arm. If the
operator is not
satisfied that the measurement was successful and sufficient, he can repeat
the
measurement as necessary or he may choose just to mark the measurement as
being
rejected or at least suspect.
[0083] In one possible arrangement, five categories of measurements are made
to
assess the structural condition of the spar-arm. These are (1 ) the speed of a
shock-
wave traveling longitudinally down the length of the spar, e.g., from end to
end; (2) the
character of the applied hammer shock, including the level of energy applied;
(3) the
level of shock energy transmitted through the wood to the end of the data
collection
pod; (4) the gamma opacity of the wood; and optionally (5) the dimensions of
the wood
in the zone where the gamma opacity was measured.
[0084] The speed of a shock propagating in wood has been shown to correlate
with
the strength of the wood, its ability to hold a cantilevered force, and the
degree of rot
and pre-rot condition in the wood. This measurement is made using two
accelerometers, mounted with measuring axes coincident with the major axis of
the
spar. One is in a fixed position on the small end of the spar, and one is on
the back of
the hammer which delivers the shock. Speed is calculated by noting the time
between
-20-
CA 02484456 2004-10-12
the leading edges of the hammer-shock and the end-of spar shock. Provided all
the
spars tested are of a standard length, the shock wave propagation time and the
length
of the spar are used to determine speed.
[0085] Often, the spars used to construct a given power-line were made and
installed in a situation whereby the spars are likely to all be substantially
of a nominal
size and to employ the same jigs and fixtures. For example, they were
installed in
accordance with specifications and/or were obtained from the same source, etc.
Generally, even after some aging and variations in caused by differences, the
spars are
close to the same length (e.g., within one inch of being the same length).
This is useful,
so long as it proves true, and permits reliance on certain assumptions (such
as nominal
length or thickness, etc.) when determining shock wave velocity, density or
other
dimensionally related variables. However, care must be exercised when making
such
assumptions. For example, a given spar may have been previously broken and
replaced from a stock other than the original stock, resulting in a
dimensional difference
that needs to be taken into account.
[0086] Similarly, the locations of holes bored for insulator-hanging bolts and
the like
are sometimes determined by using a jig, such that their lateral positions
and/or spacing
from the spar ends often can be assumed to fall within a certain precision.
This
precision may be important to the radar measurement, which for comparing one
spar to
another depends on the insulator bolts being at the same positions on the wood
in each
case. Using radar, the shock speed is measured between the near-end of the
wood
and the far insulator bolt. If the length is invariable, the correlation to
strength applies
equally to the propagation time. If the spars are of variable length, the
operator can
enter a length value, for example using a range finding device to visually
assess length.
[0087] The acceleration profile of the hammer-shock provides a measure of how
sharply or dully the hammer blow falls, and correlates with the hardness or
softness of
the wood and with corresponding strength or weakness. The time-width of the
hammer
shock pulse as determined by deceleration of the hammer head during impact
indicates
the softness of the end of the spar. A long pulse indicates a soft spar-end,
and a soft
spar-end suggests possible end-rot. This measurement provides a measure of the
-21 -
CA 02484456 2004-10-12
general quality of the spar being measured. The respective measurements can
have
acceptability thresholds individually or can be combined with one another to
arrive at a
passlfail or other result.
[0088] The extent of energy transmitted through the wood from the point of the
blow
to the point of detection, is another measure of structural strength. The
integrated area
under the end-shock pulse in a time plot, as collected by the accelerometer
mounted in
the instrument pod, is proportional to the amount of shock-energy transmitted
through
the wood, and an indication of wood quality or strength. A strong spar-arm
transmits a
higher proportion of applied shock energy than a weak spar-arm. The time
integrated
energy of the end-shock pulse is divided by the likewise-integrated energy of
the
hammer-shock pulse, resulting in a data point representing the fraction of
applied shock
energy that was transmitted to the detector.
[0089] The gamma opacity of the wood on the end of the spar provides a measure
of
density if the spars are all of the same thickness. In a preferred arrangement
the
gamma opacity is measured using a radioactive source and a scintillation
detector,
disposed diametrically opposite from one another on the spar.
[0090] The relative number of 0.662 MeV gamma-rays from a fixed quantity of Cs-
137 that are able to penetrate the wood and to be detected by a semiconductor
sensor
over a known counting period is collected as a measure of opacity. This count
can be
correlated with the density of the wood when combined with the diameter
measurement
of the spar-end being tested. Activity of the gamma source is 10.0 to 20.0
micro curies,
20 being preferred. Gamma rays are counted by summing the scintillations or
brief
flashes of energy produced in a 2.0 by 2.0-inch right-cylindrical Thallium-
activated
Sodium-Iodide crystal, collected by a photomultiplier tube.
[0091] The detector crystal and the gamma-ray source of the gamma opacity
detector need not be movable because intervening air has a trivial opacity as
compared
to wood. The source and detector can be located across a diameter from one
another
of the circular small end of the spar, separated, for example by 10 inches or
by another
distance sufficient to admit the largest diameter cross-arm to be tested. When
the pod
is clamped to a spar, the wood occupies the gamma ray path between the source
and
-22-
CA 02484456 2004-10-12
the detector crystal, enabling measurement ofthe degree of opacity. The
density of the
wood is related to the breaking strength of the spar, and when combined with
other
measurements provides a good correlation with the quality of the spar.
[0082] The diameter of the spar-ends encountered may vary substantially, thus
affecting the correlation of the gamma opacity with material density. However,
the
diameter of the spar as an independent variable correlates with spar strength
(i.e., a
thicker spar is likely to be stronger than a thinner one). Therefore, in
general, gamma
opacity correlates with strength, making it possible to rely on opacity and to
ignore
diameter as a pertinent factor.
[0093] If the thickness of the wood sample is to be taken into account, a
diameter
measurement can be derived in various ways. Two diametrically opposed linear
potentiometers, linearly variable differential transformers (LVDTs ) or other
such length
encoders, can be located on opposed clamp plungers and their signals summed.
The
sum of two clamp-depths, regardless of the accuracy of concentric placement of
the
pod, could then indicate the span on that diameter. One diameter measurement
can be
expected at least to approximate the average diameter of the spar-end (with
some
leeway for non-circular growth tendencies of wood).
[0094] The instrument pod may comprise a collection of subsystems relating to
mounting, measurement and communications. There are several such subsystems
including the clamp assembly or other instrument attachment structure, the
accelerometer sensor responsive to the received shock signal, the gamma
densitometer
(comprising a source and a detector), optionally the diameter measuring
caliper device,
and the radio telemetry link for communicating with the data collection
processor on the
aircraft or other vehicle.
[0095] In an example of a practical embodiment, a PCB model 353803
accelerometer was employed as the primary shock sensor in a remotely mounted
pod
22 as in Figs. 1-4. The pod was equipped with springs and pneumatically
operated
paddles (not shown) arranged to grasp the cross-arm 21 and to pull the pod 22
axially
against the remote end of the cross-arm. The accelerometer was fitted to a
three-
quarter-inch aluminum shaft, arranged such that the base of the accelerometer
rests
-23-
CA 02484456 2004-10-12
positively against the end of the cross-arm. The accelerometer was spring-
loaded
against the wood. A two-inch by three-quarter-inch helical spring can be used
for this
purpose, sharing its major axis with the accelerometer shaft. A relatively
large axial
displacement span, such as one or two inches, is advantageous for the
accelerometer,
because the ends of the cross-arms may not be cut square.
[0096] The accelerometer connects, by way of a PCB lov~noise coaxial cable, to
a
PCB Model 480E09 ICP Sensor Power unit. The power unit contains a 24-volt DC
power supply for the on-board charge-amplifier in the accelerometer, an
adjustable-
attenuation signal-conditioner, a cable-integrity and batter~state indicator,
and three 9-
volt batteries. The sensor power unit is bolted to the back of the base of the
instrumentation pod. The power unit is preferably easily accessible for
battery
replacement purposes.
[0097] An exemplary gamma densitometer comprises a 10-micro Curie Cs-137
gamma-ray source and a scintillation-counter gamma-detection system. The
radioactive cesium gamma source is embedded in a 1-inch diameter plastic disk.
The
disk is held in a machined cylinder of nylon, fitted with a 0.125-inch
aluminum cover,
held in place with three screws. The gamma-source is held near the face of the
cylinder, behind 0.0625 inches of nylon, under tension of a 0.750-inch
diameter helical
spring. The source-carrier bolts to the rim of the pod base, and is oriented
to emit
across the diameter of the spar, in the direction of a scintillation-counter
gamma
detector head on the diametrically opposite part of the rim.
[0098] A scintillation counter is preferred as the gamma-ray detector, because
energy-discrimination is needed, with only gamma-peaks above a predetermined
threshold energy being counted so as to provide immunity from background
noise. A
2X2-inch scintillation crystal provides reasonably good discrimination between
gamma
peaks and background noise, although there is some compromise between weight
and
counting efficiency, as well as other considerations such as expense.
[0099] A possible scintillation detection system for use as described includes
a
Bicron 2M2/2 scintillator head, a Bicron P-14 voltage divider module, a built-
in signal-
splitter, an Aware Electronics PMI-30 HV power supply and pulse-shaper, and
-24-
CA 02484456 2004-10-12
connecting cables. The system is calibrated by setting the pulse-height
threshold on
the PM4-30 to just below the leading tail of the high-gamma peak for Cs-137.
The
calibration pulse is generated by an adjustable function generator, which is
set to the
correct voltage using a nuclear multi-channel analyzer, having previously
analyzed a
Cs-137 source to find the voltage level of the peak in question.
[00100] The radio-telemetry system requires a relatively short range
capability to
communicate and report the results of measurements taken, including the
accelerometer signal, the gamma count reading and optionally the spar
diameter. The
accelerometer signal specifically requires time-function signaling. It is
possible to report
the accelerometer data as a function of time in various ways, one way being to
provide
instantaneous signal amplitude samples at the Nyquist sampling rate required
to ensure
sampling at the highest frequency of interest. Depending on the analysis to be
undertaken, at least a triggering signal to report the timing of the arrival
of the leading
edge of the shock wave is required to determine shock wave propagation time.
In more
extensive analyses, the frequency domain character of the received signal can
be
compared to that of the imparted one at the exciting end, for example
permitting an
analysis of the phase dispersion of the packet, amplitude of received shock as
a
function of wavelength, and/or other attributes.
[00101] To qualify the subject measurement techniq ues, a number of ten foot
wood
beams were broken and their breaking strength measured. Salvaged portions of
the
beams were cut to 56 inch lengths and subjected to measurements as described,
using
laboratory mounted devices in lieu of the airborne system by which such
measurements
can be taken in the field.
[00102] Accelerometers having a 7 kHz bandwidth were provided for measuring
the
applied shock and detecting the propagated shock wave. The test beams were
bolted
to a holder with a bolt 10" from each end of the beam (36" between the bolt
centers).
The beams were not loaded in any way. A set of experiments was run to
investigate
gamma ray opacity measurement; to test portable hammer designs; and , to
verify the
expected statistical correlation of strength with impact propagation data.
-25-
CA 02484456 2004-10-12
[00103] Two gamma ray sources were tested in these experiments and proved
useful,
namely a low-rate source (1 micro curie) providing high-energy gamma rays; and
a
high-rate source (cobalt 57) that provided low-energy gamma rays. In each
case, the
source was placed on one side of the beam under test with a detector on the
other side.
Particle detections were accumulated via a multi-channel analyzer, which
counted the
number of particles received in discrete energy level ranges or bins. For the
lov~rate
source the particles were accumulated over a 100-second period. For the high-
rate
source the particles were accumulated over a 10-second period.
[00104] Absorption levels (correlating to gamma opacity) were found to be
somewhat
affected by moisture content. Absorption in air was minimal, so that air gaps
of the size
encountered according to the invention could be ignored. The correlation
coefficient
between material density and gamma absorption was found to be generally better
for
low and medium energy levels, and lower for high energy levels. Thus the gamma
opacity measurement can be optimized by discriminating in favor of counting
lower
energy level particles.
[00105] The two graphs of Figs. 5 and 6 show example signatures recorded by
the
accelerometers fastened respectively to the portable hammer (heavier line
trace), and
the signature of the accelerometer on the remote pod (lighter line trace). In
these
examples the hammer signal is shown at one tenth scale compared to the
received
pulse at the remote pod. In Fig. 5, the pod accelerometer was attached to the
beam
using fasteners and in Fig. 6, the pod accelerometer was clamped to the beam,
achieving substantially similar results.
[00106] The ratio of energy applied to the beam by the hammer, versus the
energy
received at the far end, is one indicator of beam strength. Defects and decay
in the
beam attenuate the energy and reduce the ratio. A wood hardness index can be
defined as the ratio of the amplitude of the hammer pulse to the width of the
pulse, and
is a useful indicator of the quality of the wood at the impact excitement end
of the beam.
A sharp narrow "knock" pulse indicates a bouncing hammer and a higher wood
quality
than a longer dull and penetrating "thump."
-26-
CA 02484456 2004-10-12
[00107] The impact pulse propagates throug h the beam to the remote sensor.
The
impact pulse propagation delay, or time between application of an impact and
the arrival
of the resulting energy pulse at the detector, is another measure of the
strength and
quality of the beam. The acceleration/deceleration pulse that occurs upon
striking the
beam with the hammer has a discrete leading edge, that is acceptable for
calculating
shock delay times (i.e., start time at the hammer and stop time at the
detector).
[00108] The following table shows the strength discriminant data collected in
a run of
43 available beams. The column labeled AvgAbsorb shows a measure of gamma ray
absorption, in particular the average count obtained using a low-rate high-
energy source
over both ends of the 56° beams. The Far Energy is the energy received
at the far
accelerometer.
- 27 -
CA 02484456 2004-10-12
Strength Discriminants
Shock Propagation
# Beam Strength AvgAbsorb ShockTime FarEnergy
1 146 2072 13.7 21.8 554.1
2 020 2800 11.3 25.5 144.0
4 118a 2819 15.9 30.9 49.1
114 2992 13.2 45.1 14.7
6 145 3025 12.5 28.3 112.3
7 140 3220 14.1 29.4 180.6
8 069 3343 18.6 41.0 38.6
9 130 3393 14.1 25.6 264.7
142 3685 14.4 27.8 249.7
11 120 3735 15.3 42.1 26.3
13 031 a 3918 14.2 23.5 162.8
14 268 4219 12.5 22.4 213.6
150 4345 17.5 26.0 340.8
16 213 4417 15.4 24.0 129.2
18 057a 4451 15.6 23.9 210.5
19 055 4465 15.1 24.4 121.4
121 4516 12.1 23.6 112.0
22 129a 4544 13.6 26.0 142.0
23 134 4792 13.1 29.8 53.3
24 148 5035 11.3 30.7 281.7
021 5044 17.7 24.0 344.6
26 128 5136 12.1 29.3 254.7
27 063 5350 15.1 27.4 16.3
31 119a 5397 16.1 22.3 301.5
33 122a 5518 16.8 24.4 125.2
34 269 5529 16.7 23.9 115.5
166 5531 16.8 25.1 178.9
36 101 5641 16.9 25.1 243.1
37 124 5695 12.9 24.4 88.8
38 152 6446 20.3 23.0 244.2
39 054 6529 17.8 24.5 141.9
112 6606 18.0 22.2 293.6
41 164 6841 17.5 21.8 281.6
43 067a 7952 19.0 20.3 411.8
[00109] Figs. 7 and 8 illustrate graphically how the discriminants in this
experiment
related to beam breaking strength. The weakest beam (146), with a measured
strength
of only 2072 Ib, was consistently an outlier on the graphs, especially on the
Far Energy
graph. The results for that beam suggest that the beam may have had relatively
strong
_28_
CA 02484456 2004-10-12
material except for a local defect (such as a knot in tension near the center
of the beam)
that resulted in b reakage at a lower stress than othenNise could have been
predicted.
The Far Energy discriminant shows a trend consistent with strength,
particularly if beam
146 is not taken into account.
[00110] A good individual discriminant of breaking strength was found in the
beam
density or gamma opacity multiplied by the square of the shock velocity. That
discriminant correlates with breaking strength at a level of 0.69, and if the
anomalous
beam (146) is deleted, the correlation with breaking strength was 0.78.
Generally,
favorable correlations (better than two thirds) were found with respect to
average
energy absorption, shock propagation time, and proportion of impact energy
received at
detector. One variable, namely the product of average absorption and the
square of the
pulse speed, had a correlation of 0.74.
[00111] The respective measures can be combined to produce related factors
representing measures of beam strength. A high correlation in sets of
parameters in the
foregoing results is found in the combination of gamma ray opacity multiplied
by shock
wave velocity squared (AvgAbsorb*SpeedSq).
[00112] The most valuable discriminant for strength was the combination of
gamma
ray density multiplied by the shock wave velocity squared (AvgAbsorb*SpeedSq).
Adding the far energy (FarEnergy) slightly improves the correlation with
strength (0.69
to 0.74) and slightly improves the statistical validity of the model (0.46 to
0.51 ).
[00113] The Far Energy discriminant may be improved when normalized by the
hammer energy. The wood hardness index, which reflects the quality of the wood
at the
point where the hammer strikes it, is another useful indicator of beam
strength. Using
aii of the available discriminants, with accurate measurements and a multiple
regression
analysis, a correlation of about 0.8 with the actual beam strength is believed
to be
possible.
[00114] From these results it can be hypothesized that gamma ray opacity is
useful to
obtain an estimate of material density and a measure of strength. However the
technique works best if a specific gamma particle energy band is selected. The
opacity
-29-
CA 02484456 2004-10-12
and density are affected by moisture content (a 22% increase has been observed
in a
beam soaked in water as compared to the beam when dry).
[00115] The results further suggest that longitudinal shock wave travel time
is
unaffected by lateral stress in the beam and is a measure of strength. An
estimate of
beam strength with a correlation coefficient of 0.7 or above can be obtained
by
combining discriminants such as:
a. Density * ShockTime Squared.
b. Energy Ratio between the hammer pulse and the far accelerometer.
c. Hardness Index of the beam at the hammered location.
(00116] A range of practical embodiments are possible in accordance with the
foregoing details. As stated above, the invention can be embodied with
diameter
measuring devices that can accommodate off-center mounting on the spar end.
The
measurement of diameter can in some instances be ignored in favor of gamma
measurement alone, under the theory that within certain limits, a larger
diameter spar
and a smaller diameter o ne are likely to be of equal strength if their gamma
opacities
are equal, indicating that a similar amount of structural material is present.
Provided
that the sensing accelerometer is engaged against the spar end, it is possible
to employ
specific mounti ng structures that employ pneumatic clamps, or instead to rely
on
springs. Figs. 9-13 illustrate a practical embodiment for the sensor housing
or pod, that
has been developed taking into account the robustness and durability that is
needed in
a real world setting, and the minimum requirements of taking reasonably
dependable
measurements of actual cross-arms or spars. The spars may be out of round,
weathered, cut unevenly or at an angle other than perpendicular to the spar
axis,
provided with attached structures near the ends, etc.
(00117] The sensing accelerometer needs to be pressed into contact with the
spar
end, which can be accomplished using an appropriate mounting to present a
protruding
part the accelerometer for contact with the spar end. Preferably such contact
is made
at or near the axial center of the spar, but whether the accelerometer is
centered is not
considered critical because the determining factors substantially only involve
the
longitudinal propagation of the impact shock wave.
-30-
CA 02484456 2004-10-12
[00118] In contrast, the gamma opacity measurement is substantially a
diametrical
matter. The measurement can be affected, for example in a spar having a non-
round
cross section or some irregular shape, by the relative angular positioning of
the line
between the gamma source and detector. In an effort to improve the
repeatability and
consistency of gamma-density measurements, a mounting was devised for
improving
the centering of the instrument pod relative to the spar, while holding the
instrument pod
securely axially against the spar end, yet permitting installation and removal
with
reasonable facility.
[00119] Among the objectives are to provide a measurement that produces valid
measurement information that is relatively insensitive to the cross sectional
shape of the
spar, e.g., whether it is round or non-round, oval or ovoid, continuous or
gapped, etc.
The spar end is assumed to be of arbitrary diameter, within maximum and
minimum
limits, and of arbitrary cross-sectional shape, within limits. These limits in
the radial or
diametric directions are determined by the radial displacement span of the
structures
that grasp or engage the spar adjacent to the spar end. The maximum and
minimum
spar diameters must be consistent with the maximum and minimum radial span of
the
grasping structures, e.g., springs, clamps, or the like, so that (1 ) the pod
can be fit onto
the spar when the grasping structures are at their maximum relative distance
and (2)
the pod grasps even a small spar with sufficient force when the grasping
structures are
at their minimum relative distance to securely fix the pod and seat the
accelerometer
hard against the spar.
[00120] The end of the spar is typically cut perpendicular to the elongation
of the spar,
but may have an arbitrary angle that is not precisely perpendicular to the
axis. Although
the end of the spar could be inclined or skewed somewhat, the accelerometer
must be
held close against the spar in order to bear against the spar end. Assuming
that the
pod housing is to be aligned perpendicular to the longitudinal axis of the
spar, the limits
to which the spar end can be skewed and still operatively receive the sensing
pod, are
determined in part by the extent to which the accelerometer protrudes axially
toward the
spar from the sensor pod housing. Also, in order to accommodate a large skew
angle
and still grasp the spar, the grasping structures must be spaced farther from
the pod
housing along the axial length from the end of the spar. It is possible to
align the pod to
-31 -
CA 02484456 2004-10-12
the end face of the spar, which may be skewed or cut at an angle. A skewed cut
at the
end of a cylinder produces an end face that is elliptical. Therefore, the
matter of out-of-
round spars and round (cylindrical) spars with skew-cut end can be considered
as
related problems. The dimensions of the spar, the permitted tolerances of end
flatness
and alignment, texture (e.g., voids) and similar considerations may need to be
considered in order to obtain repeatable measurements for different particular
spars
and/or to compare the results obtained for different spars.
[00121] The invention has been disclosed with respect to the foregoing
examples, but
it should be appreciated that certain variations and routine extensions of the
subject
matter within the scope of the invention, will now become apparent to persons
skilled in
the art and aware of this information.
-32-
