Document text
NBSIR 76-840
FM-CW ELECTROMAGNETIC TECHNIQUE OF
MEASURING COAL LAYER THICKNESS
Doyle A. Ellerbruch
Donald R. Belsher
Electromagnetics Division
Institute for Basic Standards
National Bureau of Standards
Boulder, Colorado 80302
May 1976
Prepared for
U.S. Bureau of Mines
United States Department of the Interior
Pittsburgh, Pennsylvania 15222
NBSIR 76-840
FM-CW ELECTROMAGNETIC TECHNIQUE OF
MEASURING COAL LAYER THICKNESS
Doyle A. Ellerbruch
Donald R. Belsher
Electromagnetics Division
Institute for Basic Standards
National Bureau of Standards
Boulder, Colorado 80302
May 1976
Prepared for
U.S. Bureau of Mines
United States Department of the Interior
Pittsburgh, Pennsylvania 15222
U.S. DEPARTMENT OF COMMERCE, Elliot L. Richardson, Secretary
Edward 0. Vetter, Under Secretary
Dr. Betsy Ancker-Johnson, Assistant Secretary for Science and Technology
NATIONAL BUREAU OF STANDARDS, Ernest Ambler, Acting Director
FOREWORD
This report was prepared by the National Bureau of Standards, Boulder,
Colorado, under USBM Contract No. J0155124 . It was administered under
the technical direction of the Pittsburgh Mining and Safety Research Center
with Mr. M. Pazuchanics acting as the technical project officer.
CONTENTS
Page
1 . INTRODUCTION 1
2. MEASURING SYSTEM 1
3. SYSTEM PARAMETERS 4
4 . ANTENNA IMPROVEMENT 5
5. EXPERIMENTAL RESULTS 6
5.1 Laboratory Samples of Shale and Coal 6
5.2 Bruceton Mine 7
5.3 Loveridge Mine 8
5.3.1 Roof Data 8
5.3.2 Floor Data 11
5.4 Hillsboro Mine 11
6 . CONCLUSIONS 12
7. RECOMMENDATIONS 13
8. ACKNOWLEDGMENTS ■ 15
9 . REFERENCES 15
APPENDIX 16
iv
LIST OF ILLUSTRATIONS
Page
Table I. Laboratory Sample Data 17
Table II. Bruceton Mine 18
Table III. Bruceton Mine 18
Table IV. Loveridge Mine 19
Figure 1. Block diagram of the FM-CW electromagnetic system 20
Figure 2. Sample of raw FM-CW system data taken in the Loveridge
Mine 21
Figure 3. H-Plane field strength pattern 22
Figure 4. E-Plane field strength pattern 23
Figure 5 . Antennas in the Bruceton Mine 24
Figure 6 . Test areas in the Bruceton Mine 25
Figure 7. Measurement setup in the Bruceton Mine 2 6
Figure 8. Coal layer depths in the Loveridge Mine 27
Figure 9. Relative amplitude of coal surface response in the
Loveridge Mine ■ 28
Figure 10. Raw FM-CW data from station 10 in the Loveridge mine 29
Figure 11. Raw FM-CW data taken at station 15 in the Loveridge mine-- 3 0
Figure 12. FM-CW floor response data in the Loveridge mine 31
Figure 13. Coal layer depths in the Hillsboro miine 32
V
FM-CW ELECTROMAGNETIC TECHNIQUE OF MEASURING COAL
LAYER THICKNESS
Doyle A. Ellerbruch and Donald R. Belsher
ABSTRACT
An FM-CW microwave system was investigated for measuring coal
layer thickness. Measurements were made in three different mines
near Pittsburgh, Pennsylvania, near Fairview, West Virginia, and
near Coffeen, Illinois. Microwave frequencies in the range 1-2 GHz
were used to measure samples up to 55 cm thick. All samples were
backed with a naturally occurring shale. Measurements were also
made on coal and shale samples compounded in the laboratory at
the Bureau of Mines Pittsburgh Mining and Safety Research Center
near Bruceton, Pennsylvania.
The results indicate that layer thickness can be determined
in most cases, although large anomalies may, in some cases, pro-
duce misleading results. Many anomalies that were detected with
the FM-CW system were verified visually by drilling into the coal
layer .
The dielectric constant of coal apparently varies signifi-
cantly within a coal seam.
The form of the output signals from the FM-CW system seem to
simplify the data interpretation and analysis process as compared
to the manually swept microwave system used previously. It appears
that this technique has the potential of measuring changes in the
dielectric constant of a coal seam and providing an output that
can be used for real-time corrections in layer thickness measurement.
Key words: Automation; coal; coal mine safety; dielectric con-
stant; energy; microwave measurement; nondestructive testing;
thickness of coal layer.
1 . INTRODUCTION
The possibility of using a microwave system to measure coal layer
thickness in a mine was investigated by August 1974, and the results were
reported in NBSIR 74-387, Microwave Measurement of Coal Layer Thickness,
dated September 1974. The basic measurement capability was demonstrated in
that microwaves penetrated the coal and were reflected by a shale background.
The primary objectives of this phase of the proposed effort were to
develop improved methods for microwave measurement procedures, to determine
the thickness of undisturbed coal layers in situ, and to demonstrate these
procedures in actual mines.
2. MEASURING SYSTEM
The microwave hardware used in these experiments was basically the same
as that used previously [1] ; however, the antenna design was modified to
provide better directivity and to decrease the coupling between transmit and
receive antennas over the frequency bandwidth. Details of these modifica-
tions will be presented later in this report.
Appropriate microwave plug-in units were acquired to minimize the
quantity of electronic equipment needed for in-mine experiments, and the
entire system was powered from a 12 Vdc automobile battery. Data recording
was done with a magnetic tape recorder and, in some cases, with an X-Y
plotter .
The fundamental difference between the system used previously and the
one used for this work is that a greater degree of data processing was ac-
complished in real time prior to recording. The microwave system was operated
as a frequency modulated-continuous wave (FM-CW) radar.
The microwave signal generator in figure 1 is swept in frequency over
its bandwidth so that a linear frequency vs. time output signal is produced.
The microwave signal travels from the generator to the mixer via more
than one path. If the electrical lengths of the paths are identical, the
reference and test signals going into the mixer would arrive at the same
time and the instantaneous rf frequencies would be identical at all times.
But the test signal will arrive at a later time because that signal
goes through the antennas via the coal sample. A portion of the test signal
is reflected at the air-coal interface and another portion is reflected at
the coal-shale interface. These reflected signals will arrive at the mixer
at times t^ and t^ respectively as shown on figure 1.
Because all inputs to the mixer arrive at different times, the instan-
taneous microwave frequencies differ. The mixer is a product demodulator
and has an output that is a function of the product of the inputs. Only the
lowest frequency components in the output are preserved; the higher fre-
quencies are filtered out.
The lower frequency components are displayed on a spectrum analyzer.
The location of the first peak is a function of the distance from the antennas
to the coal surface. The location of the next peak is a function of that
distance plus the coal layer depth and its dielectric constant.
Assume for the moment that the microwave signal is incident upon an
infinitely thick layer of coal. Let the reference signal arriving at the
mixer be presented as
e.^ = cos o)j^^ t, (1)*
and the test signal reflected from the coal surface as,
where 2R/c is the time required for the signal to travel a distance R from
the transmitting antenna to the coal surface and back to the receiving antenna.
It is assumed that both antennas are equidistant from the coal surface and that
the angle of incidence is zero degrees.
*See appendix for definition of terms.
Going through the product demodulation process and neglecting the har-
monic microwave frequency component, the mixer output signal is,
Letting,
-1^1^2 2R
^0 " —2 -(t) —
2R
'^(t) = "(t) — '
(3)
(4)
the frequency of the signal out of the mixer is,
d;
f = ^
0 2r dt
ty
(5)
d-r
f = 2R (t)
0 c dt
(6)
If the frequency of the microwave signal is changed at a constant rate,
the frequency of the signal out of the mixer is proportional to
the distance from the antennas to the coal surface.
Extending the analysis to include a finite layer of coal comprised of
the air-coal surface and the coal-shale surface, the total received test
sicrnal [2] is.
-2vd
2R.
1" 2
-2-. d
e J
(7)
Recognizing that coal is a lossy microwave material and utilizing typical
dielectric constants of coal ( e; = 5 ) and shale {z^ = 36) , the amplitude of
the denominator in eq. (7) is found to be approximately unity. Thus ne-
glecting the multiple reflection terms contributed by the higher order
reflection term in the denominator, eq. (7) can be written as [2]
'1^2
^ , 2R
+ r2E^-{cos
(t)
c ^ c
J
-2:xd
(8)
Mixing this received signal with eq . (1) will provide a mixer output
signal
'1^2
' 2R
•/(t) c J
'2^1^3
r
cos to
when the harmonic microwave signal is neglected.
'2R _^ 2d/E
C -C
-2ad
(9)
3
The frequency of the first component is given by eq. (6) and the fre-
quency of the second component is
f = f2R. , 2d/e
02 c c
df
dt
. (10)
The difference between eqs. (10) and (6) is a frequency that is pro-
portional to the thickness and the electrical properties of the coal layer.
2dvT ^^(t) .
"^0 c dt •
The procedure has been to measure the dielectric constant, e, on site
and reduce the working equation to one unknown, d, the coal layer thickness,
AfQ. (12)
2/e df ^^j/dt
It has been found that more than one finite layer of coal may exist at
a given point in a mine. The analysis given here for one layer may be ex-
tended for stratified media.
3. SYSTEM PARAMETERS
Equation (12) is the working equation for the FM-CW system. Based
upon the measurements made last year [1] and anticipating that up to 40 cm
of coal will be measured again, the microwave frequency band 1-2 GHz was
used. That bandwidth was swept in 7.48 milliseconds thus eq. (12) becomes
Af
d - 0.1122 — - , (cm) (13)
This equation contains two unknowns (d,e) both of which will be
determined by measurement.
A measured value for the dielectric constant (e) can be obtained for
the coal by any of the three techniques discussed in [1] . The difference
frequency (^fg) out of the mixer is also a measured value; thus, the layer
depth (d) is calculated from eq. (13).
A sample of raw data obtained with the system at one of the test points
is shown in figure 2. The high amplitude response at 8.05 kHz is the reflec-
tion from the air-coal interface. The low amplitude ripple preceding that
response was caused by imperfections in the microwave system, such as resi-
dual mismatch reflections in the components. The response at 8.4 kHz was
caused by reflections from a sulphur ball (pyrite concretion) within the
coal seam at that particular measurement point.
4
The response at 8.9 kHz was caused by reflections from the coal-slate
interface. The measured dielectric constant for the coal at that test point
was 7.6. Using this with the difference frequency from figure 2 in eq.
(13), the depth of the coal seam at that test point is 34.6 cm. The physical
thickness measured at that test point was found to be 36.8 cm.
The physical distance between the sulphur ball and the shale at that
test point is 20 cm. In discussions with some of the miners in that par-
ticular mine it was learned that a thin layer (up to 0.6 cm thick) of impure
coal exists about 20-25 cm below the coal-shale interface throughout the
Pittsburgh seam. The impure coal layer is referred to as the "top binder,"
or the "soot layer." It was noted that many dielectric discontinuities,
such as sulphur balls and shale lenses, occur at the top binder level.
■ 4. ANTENNA IMPROVEMENT
The theories developed for all microwave measurements in coal mines
assumed ray paths for the electromagnetic energy radiated from the antennas.
It was realized that the antennas used were not directive enough to fully
justify use of a single ray model; however, in the initial measurements last
year the antennas were selectively positioned in the mine to reduce undesired
direct coupling.
A portion of the effort this year was expended in improving the direc-
tivity of both antennas and providing better isolation between them. Improve-
ments like this decrease the susceptibility of the system to spurious reflec-
tions from surrounding obstacles.
The antennas are broadband rectangular aperture horns which utilize
double ridged waveguide techniques [3]. Initially two sides of the horns
were made from a printed circuit board. These were removed and replaced
with solid aluminum to reduce the H plane beamwidth. All the mechanical
junctions in the horn assembly were taped with conducting tape to provide
continuous current paths inside the horn and to minimize current leakage and
resulting spurious radiation. Both antennas were covered externally with rf
absorbing material to minimize the effect of spurious radiation caused by
current flowing on the outside of the antennas.
Finally, because in an operational mine the horns will be pointed up,
it was realized that some type of dust cover would be necessary to prevent
filling the horns with coal dust. A lucite lens was designed and evaluated
for satisfying two purposes. First, it would cover the entire aperture and
would serve as a dust cover, and second, it would serve to focus the radiated
energy and more closely validate the ray theory assumption. Two different
lens designs were evaluated. One has a radius of 7.5 cm and the other had a
radius of 10 cm.
5
The performance characteristics of the antennas were determined by
measuring radiation patterns at 1 and 2 GHz in an anechoic chamber. The
intent was to make comparative field strength measurements. Results of the
antenna measurements are shown in figures 3 and 4 . The narrowest beamv/idth
and the lowest level sidelobes occur when the 10 cm radius lens was flush
mounted on the aperture. Thus a 10 cm radius lens was fabricated for each
antenna.
Figure 5 is a photograph of the antenna assembly positioned under a
test area in the Bruceton mine.
5. EXPERIMENTAL RESULTS
5 . 1 Laboratory Samples of Shale and Coal
The Bureau of Mines had prepared samples of a coal layer on a shale
layer. The shale layer was comprised of 92 percent (by volume) shale dust
and 8 percent cement. These were mixed with water and poured into a form.
Similarly a 92 percent coal dust and 8 percent cement mixture was prepared
and poured on top of the shale.
Several different laboratory samples were measured by placing the FM-CW
system antennas above the samples. A summary of sample size and measured
results are given in table I.
Sample I was the only one considered to have sufficient surface area
for measurements with the FM-CW system. Nevertheless, all of the measured
results obtained from those samples are given on Table I. With samples 2,
3, and 4, the FM-CW system illuminated not only the desired coal surface but
also the floor and, in some cases, other samples adjacent to the one being
measured. Thus, undesirable responses complicated much of the data collected.
Most of the undesired responses were related to their sources at the time of
the measurement with the use of a small metallic reflector.
The dielectric constants for samples 1, 2 and 4 were measured using
return loss techniques. Those values, along with the FM-CW frequency output,
were used in eq. (13) to arrive at the tabulated layer thickness.
No return loss measurement was made on sample 3 . The measured dielec-
tric constant values for sample 3 were determined by using the physical
depths of the coal and shale along with the measured difference frequencies
in eq. (13) .
Moisture contents (by weight) for these samples were measured by the
USBM [4]. Two specimens were chipped from each sample for laboratory moisture
analysis. Average results are shown in parenthesis on table I.
6
5.2 Bruceton Mine
As was done last year [1] , an existing mine face was undercut from the
mine floor up toward the shale to prepare three test areas having different
roof coal depths. The physical arrangement is shown in figure 6. Figure 7
is a photograph of the total area, with the FM-CW system in the foreground
between the miners. The undercut extended approximately 1.2 5 meters into
the coal face so that the antennas could easily be placed under each test
area. Each test area was at least 1.25 meters wide. The coal surfaces
were dry.
The dielectric constant was measured in the three test areas by using
the measured difference frequency and the physical depth in eq . (13). A
summary of that dielectric constant data is given in table II.
As was done last year, the dielectric constant was also measured by
utilizing known step changes in coal depth, when the absolute coal depths
are not known. Let a difference frequency measurement be made in one of
the test areas of figure 5. Equation (13) can be rearranged into the form,
If^ = 0.1122 d^.^. (14)
Now let a difference frequency measurement be made in another of the
test areas of figure 6. The equation for that measurement becomes,
L±^ = 0.1122 d^/e. (15)
Taking the difference between eqs . (14) and (15) and solving for the
dielectric constant results in
79.4 (Af--Af ) 2
£ = (16)
(d2-d^)2
After the dielectric constant has been determined with eq . (16), it can
be used in eqs. (14) and (15) to calculate the depths of the head coal at
those test points.
This measurement and data processing procedure was utilized in the
Bruceton mine. A summary of the results are given on table III.
The results given on tables II and III are calculated from data col-
lected at various points along approximately 3 meters of the face. This
year, as was observed last year, there was some variation in the dielec-
tric constant data. It is believed that the variation is real and is caused
by variations in the constituency and homogeneity of the coal along the seam
For example, the coal in test area 3 appeared to have a 17.7 cm lower layer
of friable coal, as observed on site. The upper 22.9 cm of coal was harder.
Such stratification was noted last year also in the Bruceton mine [1] , and i
consistent with the location of the "top binder" in the Pittsburgh seam.
It was also noted that although the coal surfaces and the floor under the
undercut were dry, some water was dripping from the roof 1.25-1.5 meters
back from the coal face. Variable amounts of moisture within the coal at
the various test points can result in different dielectric constants.
The head coal depth is related to the dielectric constant as given in
eq. (13). Variations in the dielectric constant will produce errors in the
measured depth unless those variations are taken into account. The relative
magnitude of the error can be determined by differentiating eq. (13) with
respect to /e, and utilizing the change in dielectric constant.
lAdL= 1 M_ |A/I| . (17)
d d 9/e
After going through the differentiation, eq. (17) becomes
iMi = lA/Ii. (18)
d /e
Using an average value for /e and the variation from tables II and III
in eq. (18), the error in the depth calculation can be as high as approximately
4 percent.
Much larger variations were noted in subsequent measurements in another
mine; thus, a method has been devised for the FM-CW system to continuously
monitor the dielectric constant. That method will be discussed in the next
section .
5 . 3 Loveridge Mine
5.3.1 Roof Data
Measurements were made at a site where the coal had been mined at least
six years ago. The actual site was also used by Foster-Miller Associates,
Inc., for roof coal measurements with a pulse radar (USBM Contract No.
H0357000). That site was approximately 17 meters (55 feet) long. Measure-
ments of the roof coal depth were made with the FM-CW system every 1.5
meters (5 feet) along the test site.
A physical analysis of the test site was conducted jointly with
Mr. Gregg Riley of Foster-Miller Associates, Inc., immediately after the
electromagnetic measurements were done. This analysis consisted of drilling
4 cm (1 5/8 inch) diameter holes into the coal layer at approximately
1.5 meter intervals along the test site to visually and physically note the
constituency, stratification, and thickness of the coal layer. Core samples
of 10 cm diameter were taken at stations 20 and 30 for moisture analysis.
Mr. Riley has recently completed a survey of the test site by using a
4 cm diameter probe drill at approximately 3 0 cm intervals to determine
stratification and thickness of the coal layer. He has provided us with a
copy of those results and it has been decided that those should be used as
the reference data for the FM-CW measurement.
The results obtained in the Loveridge experiments are shown in figure
8. The radar data have been corrected for apparent changes in the dielectric
constant of the coal along the test site. Changes in dielectric constant
were measured as a function of changes in the amplitude of the signal re-
flected at the air-coal interface. Because no automatic gain control cir-
cuits were used in the system, a condition external to the measurement
system must cause the amplitude of the coal surface response to vary. As
shown previously [1] , the amplitude of that signal is a function of the
dielectric constant of the coal. Thus the changes in the amplitude of the
coal surface reflected signal correspond to changes in the coal dielectric
constant used in (13). A tabulation of delectric constants measured over
the test sight is given in table IV.
The received signal amplitude for the coal surface response is shown on
figure 9. The plots for the two experimental days run essentially parallel
at all except station 40. Parallel responses indicate that the receiver
responded to the same change in coal dielectric constant both days but the
receiver gain was set at different, but constant, levels.
Moisture analysis of samples taken from stations 20 and 30 show the
coal to be relatively dry [5], which is consistent with the lower dielectric
constants measured at those stations.
Near the end of the first experimental day the roof area of station 40
was soaked with water in conjunction with an experiment being conducted by
Foster-Miller Associates, Inc. The water was sprayed on the roof after the
FM-CW amplitude responses on figure 9 were obtained.
It was noted the following morning that the roof area was still wet.
A higher received signal amplitude on figure 9 corresponds to a higher
dielectric constant. Addition of water to a coal sample would increase the
dielectric constant; thus, the received signal amplitude data seems to
correlate well with the sequence of experimental events.
Not shown on figure 8 are other microwave responses, some of which
correlate with the "top binder" and other discontinuity locations. Other of
the microwave system responses probably were caused by discontinuities that
were undetected during the physical analysis.
9
The sample of raw data shown in figure 2 was one of the most straight-
forward to analyze. Many of the data were that straightforward; however,
some were more complicated. Consider the data shown in figures 10 and 11
for example. For figure 10, the closest physical analysis drill hole for
station 10 was at station 11. That analysis showed that a thick sulphur
ball was located 17.1 cm up from the coal surface and was 13.4 cm thick. An
additional 12.7 cm layer of coal existed above the sulphur ball.
In figure 10, the 8.15 kHz response corresponds to the reflection at
the coal surface. The response at 8.5 kHz, along with the measured dielec-
tric constant at station 10 (e = 5.2) indicated that a reflection occurred
17.2 cm in from the coal surface. That distance correlates very well with
the noted location of the lower edge of the sulphur ball; thus, it is con-
cluded that 8.5 kHz response was indeed caused by the sulphur ball. It
probably lies on the plane of the "top binder."
Because the sulphur ball was noted to be 13.4 cm thick at the drill
hole point, it is quite likely that the response at 8.9 kHz was caused by a
reflection at the top of the sulphur ball. Using a dielectric constant
measured for shale in the mine (e - 24.2) along with the difference fre-
quency of 0.4 kHz (8.9 8.5 kHz) results in a FM-CW measured thickness of
9.3 cm.
Finally, the frequency response at 9.2 kHz was probably caused by
reflection at the coal-shale interface. The thickness of the coal above the
sulphur ball was computed to be 13.4 cm, giving a total depth from the coal
surface to the shale at 39.9 cm at station 10.
For figure 11, the closest physical analysis for station 15 was done at
station 14, where the total physical thickness of the coal was 52.1 cm. The
response at 8.45 kHz corresponds to a reflection at a depth of 17.2 cm.
Based upon all the data analyzed and considering the relatively large ampli-
tude of that response, it is assumed that reflection was caused by the top
binder .
The weak response at 8.7 kHz corresponds to a reflection at 30.4 cm
within the coal layer. No discontinuity was noted at that depth in the
physical analysis at station 14. .
The large amplitude response at 9.1 kHz is interpreted as the reflection
at the coal-shale interface, thus the total coal thickness is found to be
52.6 cm at station 15.
In both of these examples the response from the coal-shale interface is
quite definite; however, there can be some problems in interpreting all of
the responses without a complete physical analysis. This measuring system
10
can be readily adapted to continuous recording and then some of the anomalies
such as sulphur balls become obvious because they are small in extent.
5.3.2 Floor Data
Figure 12 is data taken with the FM-CW looking into the floor. There
was no coal on the floor at any point, except for some dust that was mixed
with rock and shale dust at station 40.
The data indicates that some floor layering is present. A very limited
physical analysis revealed the following.
No coal was on the floor at station 50. The first 16 cm of slate was
easily fractured and very easy to remove. At a depth of approximately 16 cm,
the slate become very hard and was essentially impenetrable with a hand
pickax. Most of the excess water that was sprayed on the roof the previous
day collected on the floor at station 50. Water penetrated the shale approxi
mately 2.5 cm.
Approximately 1 cm of coal, shale, and rock dust was on the floor at
station 40. This was the only point where floor coal in any form existed.
The first 2.5 cm of slate fractured very easily and was removed. The
hard slate layer came next. No water collected in this test area.
At station 30, 6.5 cm of shale and rock dust was on the surface.
A 5.0 cm layer of easily fractured shale came next followed by the hard
shale. The entire area was dry.
At station 10, the upper 7 cm of slate was easily fractured. This
was followed by the hard slate. The entire area was dry.
5 . 4 Hillsboro Mine
Measurements were made at a site where the coal had been mined 4-6
week previously. This site was also used by Foster-Miller Associates, Inc.,
for measurements with their pulse radar. This site was approximately
31 meters (100 feet) long. Measurements were made every 1.5 meters
(5 feet) along the test site.
A physical analysis of the test site was conducted jointly with
Mr. Riley of Foster-Miller Associates Inc. This analysis consisted of
drilling 4 cm diameter holes into the coal layer at approximately 30 cm
intervals along the test site to visually and physically note the con-
sistency, stratification, and physical thickness of the coal layer. Samples
were taken for moisture content analysis in the laboratory [5].
The results obtained from these experiments are shown in figure 13.
In these experiments also the amplitude of the signal reflected at the
air-coal interface was noticed to vary from test point to test point; how-
ever, the range of variation at Hillsboro was much greater than that at
Loveridge. Perhaps the wider range of variation at Hillsboro is attributed
to the fact that the coal layer thickness ranged from essentially zero to
approximately 20 cm, thus the dielectric constant at the surface ranged from
values of slate to values of coal. In any event, the wide range of amplitude
variation drove the spectrum analyzer response off scale at times. This
large change in received signal strength was counteracted on site by ad-
justing the system gain. Changes in system gain were not recorded, thus the
measured results shown on figure 13 are not corrected for changes in dielec-
tric constant. A dielectric constant of 4.6 was used for all the results
given in figure 13. ,
Another problem that was present with these experiments (not realized
until the data was being processed) was that the system had not been set up
for very thin coal layers. The rf sweeper used had the capability to be
swept over the bandwidth 2-4 GHz. For coal depths less than 10 cm the rf
bandwidth should have been increased to increase the system resolution.
Thus, measured data was not obtained at all of the thinner test points.
6. CONCLUSIONS
6.1 The FM-CW system can be used to measure the thickness of a layer of coal
under most conditions, as well as detect other anomalies within that
layer. The thickest layer measured with this system was 55 cm. How-
ever, that does not appear to be the upper limit.
6.2 The FM-CW system provides an output signal in a format that is amenable
to data interpretation and analysis processes. Information other than
coal thickness is inherently present in the data output. The distance
from the antennas to the coal surface, for example, can be extracted.
Anomaly presence is also indicated.
6.3 The dielectric constant of coal apparently varies significantly within
a coal seam. The dielectric constant in the Bruceton mine was lower
this year than last year. Last year the coal surfaces were wet be-
cause of the high humidity within the mine. This year the coal surfaces
were dry; thus moisture was evidentally a significant factor in those
measurements .
6.4 This FM-CW measurement technique has the potential of continuously mea-
suring changes in the dielectric constant of a coal seam and providing
an output signal that can be used for real-time measurement of layer
thickness. It should be pointed out, however, that this measurement
is related to the dielectric constant of the coal at the surface. It
is believed that for measurements near the face in operational mines a
measure of the dielectric constant near the surface of a layer of coal
should be representative of that for the entire layer at that point.
6.5 The realization of the existence of the "top binder" in the Pittsburgh
seam immediately explained some of the "spurious" responses last year
in the initial experiments as well as this year. The fact that the top
binder varies in thickness from essentially zero up to 0.6 cm indicates
the ability of the FM-CW system to respond to thin layers whose physical
and electrical differences from the surrounding coal are very slight.
6.6 The dielectric constant and thickness of several laboratory prepared
samples of coal were measured; however, only one sample was considered
to have sufficient surface area for measurements with the radar system.
6.7 The directivity of both antennas and the isolation between them was
improved by the use of lenses and by taping the mechanical junctions
to provide continuous current paths .
6.8 The presence of sulphur balls, shale lenses, and other dielectric dis-
continuities within a coal layer can complicate the data output from
the system and consequently its analysis. Most of those discon-
tinuities are small in extent and will only temporarily appear in the
data output. In an operational system the coal removed may extend
above the top binder; thus many of the discontinuities will be removed
prior to measurement.
6.9 Floor data were collected at several points in the Loveridge mine. The
data indicates floor layering; however, a detailed physical analysis
of the floor was not done. No coal was on the floor where these mea-
surements were made .
6.10 Penetration of the shale layer with the microwave signal was accomplished
last year. The results obtained this year also indicate penetration
of the shale layer; however, a physical analysis of that layer was
not done. Shale layer data may be useful to delineate the layering
within the shale and for guidance during roof bolting operations.
6.11 The FM-CW technique is much more amenable to automation than the tech-
nique used last year.
7 . RECOMMENDATIONS
A program could profitably be undertaken to develop and apply this
microwave measurement technique. The following specific efforts are recom-
mended for advancing the technology toward developing a practical, reliable
coal interface detector (CID) to determine coal-shale interface distance into
the roof and the floor for anomaly detection and for distance measurements.
13
7.1 Instrumentation effort. This effort would be aimed at obtaining a
fully automated, permissible system. The starting point would be the
FM-CW system as it now exists; however, some modifications should be
done in attempts to simplify this existing design. For example, the
system should be modified for one antenna experiments that would, if
successful, simplify the amount of hardware required for an opera-
tional system. This will involve use of state-of-the-art directional
couplers. Solid state microwave sources in the frequency range 1-4 GHz
should be selected for incorporation into a permissible system.
Appropriate solid state circuitry ' must be developed to process the
microwave system output and to display the coal layer thickness on
a continuous basis.
7.2 Material properties measurement effort. The objective is to measure
the permittivity of coal, shale, sulphur balls, etc., in mine environ-
ments and with laboratory samples from operational mines to determine
the ranges of dielectric constant values encountered in situ. In-
cluded here is a determination of the magnitude of changes of dielec-
tric constant caused by localized conditions in a mine particularly
near the front face where actual conditions occur in an operational
mine . ■
7.3 The penetrability of electromagnetic signals into the front face of
a stratified coal seam should be studied to determine the feasibility
of detecting sulphur balls, shale lenses, and other anomalies up to
1 meter ahead of the front face. Knowledge of the electrical properties
of sulphur balls and shale lenses with respect to the surrounding coal
is of importance in this effort. Those materials must continue to be
studied in situ; however, it may be well worthwhile to take samples of
them into a laboratory for sustained evaluation in less hostile environ-
ments. Some possible findings here may include the change of elec-
trical properties of shale for example, when exposed to the atmosphere.
It may be possible to relate microwave response to strength of the
material. The magnitude of changes in the dielectric constant as a
function of moisture content, and atmospheric exposure is of importance.
7.4 In situ mine measurements effort. The study of the electrical charac-
teristics of materials in a mine must continue. Additional measure-
ments need to be made in operational mines to acquire data to establish
feasibility of measuring floor material thicknesses when the floor is
either wet or dry, for probing the front face to locate discontinuities
in the undisturbed coal seam, and to measure distances from a reference
point to the roof, to the floor, and to the front face.
14
7.5 Modeling effort. Some mathematical modeling was done the last two years
to predict responses and in some cases, to verify some of the experi-
mental data. Mathematical modeling efforts should be continued. In
addition, some experimental modeling should be done to determine what
effects the presence of a metallic miner might have on the performance
of the radar system.
8 . ACKNOWLEDGMENTS
None of this investigation would have been possible without the complete
cooperation and excellent assistance of the following: John Burr of Lee
Engineering, A Division of Consolidation Coal Company; Paul Carter, Darrel
Auch, Hershel Moats and Walter Gull at Loveridge mine; Emil Teisa and
Mike Caldwell at Hillsboro mine; Mike Pazuchanics, Bert Nagy and others at
the Bruceton Mine of the USBM Mining and Safety Research Center; Gregg Riley
of Foster-Miller Associates, Inc.
Jocelyn Spencer provided the drafting service, and Sharon Foote provided
the typing service .
9 . REFERENCES
[1] Ellerbruch, Doyle A. and Adams, John W. , "Microwave Measurement of Coal
Layer Thickness," NBSIR 74-387 (Sept. 1974).
[2] Javid, Mansour and Brown, Philip Marshal, Field Analysis and Electro-
magnetics (McGraw-Hill Book Company, Inc., New York, N.Y,, 1963).
[3] Kerr, John L. , "Short Axial Length Broadband Horns," Proceedings of
the 22nd Annual Symposium on USAF Antenna Research and Development,
University of Illinois (Oct. 11-13, 1972).
[4] United States Department of the Interior, Bureau of Mines Analytical
Report on Coal and Shale Samples, Laboratory No. K57384-K57399
(Oct. 28, 1975) .
[5] United States Department of the Interior, Bureau of Mines Analytical
Report on Coal, Laboratory No. K60552-K60568 (Jan. 30, 1976).
15
APPENDIX
ABBREVIATIONS AND SYMBOLS
E^fE^fE^ Electric Intensities in paths 1, 2, and 3, V/m.
R Distance from the antennas to the coal surface, m.
c Velocity of propagation in free space, m/sec .
d ' Coal layer thickness, measured with the FM-CW system, m.
^l'^2 Instantaneous electric intensity in paths 1 and 2, V/m.
Instantaneous electric intensity out of the receiver antenna, V/m.
fg- Frequency of the signal out of the mixer, Hz.
Af^ Difference between two signal frequencies out of the mixer, Hz.
t Time, sec.
a Attenuation constant, neper/m.
^l'^2 Reflection coefficients.
Y Complex propagation constant,
e Relative dielectric constant,
w Angular velocity, radian/sec.
w^^j Instantaneous angular velocity of the microwave signal out of the
generator, radian/sec.
16
Table I. Laboratory Sample Data.
triiy s xodx
Dielectric
Layer Thickness
From FM-CW
1*16 a. o U J. 6 1116 n X. o
Surface
(cm)
Thickness
Constant
(cm)
Sample
iNumDer
(cm)
v^oax ofiaxe
Coal
Shale
uoai onaie
1
63.5 X 137
12.7 12.7
4.8
(6.3%)
51.8
** (16.3%)
12.6
2
33 X 107
30.5 10.2
4.8*
24.7*
30.7
3
33 X 107
19.1 11.4
9.0
(2.8%)
36.4
(1.8%)
4
33 X 107
20.3
12.7
16.6
(3.9%)
*No moisture analysis done.
**Values in parentheses are moisture content.
17
Table II.
Bruce ton Mine.
Test Area
Physical
Thickness of Layer
Measured
Dielectric Constant
1
6.4
2 *
27.9
cm
5.6
2*
27.9
cm
6.3
3
40.6
cm
5.6
*Dif f erent
points in that
test area.
Table III. Bruceton Mine.
Test Area
Measured
Dielectric Constant
Coal Layer Thickness From
FM-CW Measurement
(cm)
1 and 2
5.5
19 .2
28.1
2 and 3
5.8
27.4
40.1
18
Table IV. Loveridge Mine.
Station
Measured
Dielectric Constant
-5
7.6
0
5.3
5
4.9
10
5.2
15
4.6
20
4.2
25
3.2
30
2.5
35
3.7
40
*
45
4.5
50
6.8
*4.9 on the first test day (before water spray)
6.8 the second test day (after water spray)
19
(draw slate)
/ (coal) .
(air)
MICROWAVE
SrGNAL
GENERATOR
Transmit
Horn
FREQUENCY
Reference
Signal
Input
►
Ml
SPECTRUM 1
ANALYZER 1
Test Signal Input and
Receive
Horn
Mixer Output
AMPLITUDE
Figure 1. Block diagram of the FM-CW
electromagnetic system
20
3SN0dS3d dO 3anindi/^v
21
o
O O IT' O in O ITTT
I — — cvj c\J to ro
''I'll
22
9>
3
<
CI
u
d)
4J
4-)
CO
a.
0)
m
a;
CO
iH
I
W
(U
!-i
3
00
•H
gp * dpni!|dujv aAj|D|9bl
23
24
26
O O O O O O-r
a> lO ^ ro CVJ — '
(UUO) 4id9a J9Ad-| |D03
27
1 1
1
1
o
•
IT)
to
1
o
1
ro
ro
1
X
•
O
O _
o
•
X
•
x» —
o
ro
•
—
O
X
•
cvJ
•
X
•
O
•
O
1 1
1
o
0>
E
z
c
o
CM O 00 CC ^ CM T
(4U9UJd3D|ds!p mo) jez^iDuy oinjpeds uo dpn^iidaiv |Du6!S
28
6.5 7.0 7.5 8.0 8.5 9.0 9.5
FREQUENCY, kHz
Figure 10. Raw FM-CW data from station 10 in the Loveridge Mine
29
30
ZH>1 ' Aou9nb9Jj indino J9><!IN
31
NBS-lUA (REV. 7-73)
U.S. DEPT. OF COMM.
RIRLIOCRAPHIC DATA
SHEET
1. PUBLICATION OR REPORT NO.
NBSIR 76-840
2, Gov't Accession
INO.
3. Recipient's Accession No.
4. TITLE AND SUBTITLE
FM-CW Electromagnetic Technique of Measuring Coal
Layer Thickness
5. Publication Date
May 1976
6. Performing Organization Code
276.07
7 AUTHOR(S)
Doyle A. Ellerbruch and Donald R. Belsher
8. Performing Organ. Report No,
9. PERFORMING ORGANIZATION NAME AND ADDRESS
NATIONAL BUREAU OF STANDARDS
DEPARTMENT OF COMMERCE
WASHINGTON, D.C. 20234
10. Project/Task/Work Unit No.
2767389
11. Contract/Grant No.
J0155124
12. Sponsoring Organization Name and Complete Address (Street, City, State, ZIP)
U. S. Bureau of Mines
Pittsburgh Mining and Safety Research Center
4800 Forbes Avenue
Pittsburgh, Pennsylvania
13. Type of Report & Period
Covered
8/75-2/76
14. Sponsoring Agency Code
15. SUPPLEMENTARY NOTES
16. ABSTRACT (A 200-word or less (actual summary of most significant information. If document includes a significant
bibliography or literature survey, mention it here.)
An FM-CW microwave system was investigated for measuring ooal layer thickness.
Measurements were made in three different mines near Pittsburgh, Pa. , near Fairview,
W. Va. , and near Cof feen. 111. Microwave frequencies in the range 1-2 GHz were used
to measure samples up to 55 cm thick. All samples were backed with a naturally
occurring shale. Measurements were also made on coal and shale samples compounded in
the laboratory at the Bureau of Mines Pittsburgh Mining and Safety Research Center
near Bruceton, Pa.
The results indicate that layer thickness can be determined in most cases,
although large anomalies may, in some cases, produce misleading results. Many
anomalies that were detected with the FM-CW system were verified visually by
drilling into the coal layer.
The dielectric constant of coal apparently varies significantly within a coal
seam.
The form of the output signals from the FM-CW system seem to simplify the data
interpretation and analysis process as compared to the manually swept microwave
system used previously. It appears that this technique has the potential of measuring
changes in the dielectric constant of a coal seam and providing an output that can
be used for real-time corrections in layer thickness measurement.
i. 17. KEY WORDS (six to twelve entries; alphabetical order; capitalize only the first letter of the first key word unless a proper
j name; separated by semicolons )
Automation; coal; coal mine safety; dielectric constant; energy; microwave measure-
I ment; nondestructive testing; thickness of coal layer.
18. AVAILABILITY [xl Unlimited
1 For Official Distribution. Do Not Release to NTIS
1 ' Order From Sup. of Doc, U.S. Government Printing Office
Washington. D.C. 20402. SD Cat. No. C13
1 y^tOrder From National Technical Information Service (NTIS)
Springfield, Virginia 22151
19. SECURITY CLASS
(THIS REPORT)
UNCL ASSIFIED
21. NO. OF PAGES
38
20. SECURITY CLASS
(THIS PAGE)
UNCLASSIFIED
22. Price
$4.00
USCOMM-ERL USCOMM.DC 29042.P74