Millimeter waves for water content monitoring in
materials and media
Meriakri V.V.,
Chigrai E.E., Parkhomenko M.P.
Institute
of Radioengineering and Electronics Russian Academy of Sciences,
ABSTRACT: Specific advantages of millimeter (MM) wave aquamerty are discussed. MM
waves ensure better spatial resolution and better sensitivity to water than
microwaves. MM waves are less sensitive to conducting impurities as compared
with microwaves; in addition, they can be used for water testing in media that
are opaque for optical and infrared radiation. In this paper, the water content
measurement methods for some liquid and solid-state materials are described.
Examples of devices for water control in crude oil and alcohol are considered.
Usually, it is possible to measure water content in real time (including in
flow) with uncertainty less than 0.1%.
Keywords:
moisture monitoring, aquamerty, millimeter waves
1 Introduction
The microwave method shows the following advantages:
·
Nondestructive measurement
·
Using of frequency dependent
material characteristic such as:
o
high permittivity of water
o
relaxation processes in
polar liquids
o
small influence of ionic
conductivity
·
Microwave radiation
propagates through opaque dielectric media, such that the volume moisture
content can be measured, in contrast to infrared radiation, which is absorbed
near the surface
·
Microwave methods are fast
and do not ionosize the test material as do some nuclear methods
·
They operate under rough
environmental industrial conditions; dust and water vapor have no influence on
the measurement results
·
high accuracy of
measurements, high sensitivity, high slew rate
·
continuous and discontinuous
measurements and simple adaptation to automation processes can be realised
·
simple handling, servicing
and high reliability
·
calibration under processing
conditions
The
monitoring of the composition of materials (including materials containing
water) is an important problem of applied spectroscopy using optical, infrared,
and microwave wavelength bands.
The aim of this paper is to examine the peculiarity of application of the relatively new millimeter (MM) wavelength region to the water content determination in materials and substances.
There
is now extensive literature devoted to the interaction between MM waves and
different materials, including water [1] –[8]. The main conclusions in relation
to the water content determination with help MM waves are as follows:
1. MM waves ensure better sensitivity to the water
content and better spatial resolution than microwaves (absorption
of MM waves in
water a > 15 dB/mm is much
greater than practically in all monitored host materials; as wavelength l decreases, the absorption in water increases more
rapidly than the absorption in these host materials).
2. MM waves are less sensitive than microwaves to
conducting impurities in water.
3. MM waves can be used for testing materials that are opaque for
optical and infrared waves.
2 Measurement methods
For
wavelengths greater than 6 - 8 mm, the effective waveguide and resonator
technique for measuring the properties of low-loss materials was elaborated in [2],
[5] On the other hand, for wavelengths smaller than 0.5 mm, very good Fourier
transform and laser spectroscopy methods are available [1], [6]. However, there
are some difficulties in carrying out material investigation in the wavelength
interval from 5- 4 mm to 0.6- 0.5 mm. The reason is that the waveguide
technique is ineffective due to a decrease in waveguide dimensions and a gap
between a waveguide and sample walls; on the other hand, the optical technique
is ineffective due to the diffraction affecting the field structure and
preventing the use of geometrical optic approaches.
The
best way for measuring the material properties is to use a quasi-optical lens
beam waveguide [3] that transmits only the fundamental low-loss mode with small
enough cross-section dimensions of a beam and large losses for high order
modes. In this case, the wave incident on the plane-parallel specimen and the
wave on the receiving aperture are of the Gaussian type. Therefore, it is
possible to estimate the measurement errors due to a finite thickness of a
plane-parallel specimen, inclination of its boundaries to the beam waveguide
axis, as well as due to cross-section dimensions of the specimen and receiving
aperture. The methods for the determination of complex permittivity e* =e¢-ie¢¢ are mainly based on measuring the dependencies of real and imaginary
parts of transmission and reflection coefficients on frequency, specimen
thickness, and polarization of the wave.
The magnitude of errors in the real part of the transmission
and reflection coefficients measured by the beam waveguide technique is less
than 0.5% for e¢ ranging from 1.1 to 200 and e¢¢ ranging from 10‑4 to 1,
whereas, in conventional free space measurements, these
values may be greater than 10%.
3 Some characteristics of substances of interest
Tables
1 and 2 present the complex permittivity of liquids with maximal loss in the
near MM region [7]. These tables show that water has the highest value of e¢¢.
Table
2 gives the complex permittivity e =e¢- i e¢¢ and absorption a for
water depending on temperature at MM waves [7], [9]. These results are in a
good agreement with [10].
Table
2 shows that, for l= 10-8 mm, a is practically independent of temperature.
Investigations
of aqueous solutions of salts [7] showed that the absorption coefficient a of NaCl solution in water changes little with concentration
(e. g. from 57 dB/mm to 56.5
dB/mm for concentrations 1 and 3 moles/litre at frequency 320 GHz and for
temperature T = 18C). The value of a for
KCl solutions increases in this case from 60.4 to 66.1 dB/mm (negative
hydration).
Table
1 Properties of lossy liquids. (T=20C)
|
Liquid |
e¢ |
e¢¢ |
l, mm |
|
H2O |
6.3 |
8.8 |
2.0 |
|
H2O |
5.1 |
5.1 |
1.0 |
|
H2O |
4.7 |
4.0 |
0.70 |
|
H2O |
4.5 |
3.5 |
0.6 |
|
DMCO |
3.4 |
1.6 |
0.9 |
|
(CH3)2CO |
3.0 |
3.2 |
1.0 |
|
CH3COC2H5 |
2.7 |
2.2
|
1.0 |
|
CH3NO2 |
3.2 |
5.2 |
1.0 |
|
C6H10O |
2.6 |
1.1 |
1.0 |
Table 2 Complex permittivity and absorption coefficient for water vs
wavelength and temperature
|
l, mm |
T, C |
e¢ |
e¢¢ |
a, dB/mm |
|
10 |
20 25 30 |
23.5 26.7 29.8 |
31.9 33.1 33.7 |
15.5 15.3 15.1 |
|
8 |
20 25 30 |
18.1 20.6 23.2 |
28.0 29.0 30.8 |
18.8 18.9 18.9 |
|
4 |
20 25 30 |
9.1 10.0 10.9 |
16.0 17.5 18.9 |
29.5 30.8 31.9 |
|
2 |
20 25 30 |
6.3 6.5 6.8 |
8.8 9.7 10.5 |
41.1 43.7 46.1 |
.
Also we have investigated solutions of sugar in water. Table 3 shows the results of this measurements.
Table
3 Dielectric characteristics of sugar (W%) in water solutions (T= 18C, f =42.2
GHz)
|
Number |
W% |
e¢ |
e² |
tand |
a, dB/mm |
|
1 |
4.6 |
13.0 |
22.1 |
1.70 |
19.3 |
|
2 |
4.8 |
12.6 |
21.9 |
1.71 |
192.4 |
|
3 |
5.4 |
12.7 |
21.5 |
1.69 |
190.3 |
|
4 |
0 |
13.8 |
23.8 |
1.73 |
201.0 |
Here loss tangent tand= e¢/e¢¢.
Tables 4- 6 present the measured parameters materials of interest for aquametry [3], [4], [7] -
[9], [11], [12].
Table
4 Properties of low- loss liquids. (T=20°C)
|
liquid |
n |
tand ´ 103 |
l, mm |
|
cyklohexane |
1.424 |
0.50 |
0.63 |
|
octane |
1.396 |
0.74 |
0.63 |
|
decane |
1.407 |
0.83 |
0.63 |
|
nonane |
1.405 |
0.93 |
0.63 |
|
carbon
tetrachloride |
1.490 |
3.5 |
1.0 |
|
cysdecalin |
1.474 |
4.2 |
1.0 |
|
transdecalin |
1.461 |
0.6 |
1.0 |
|
benzene |
1.510 |
5.2 |
1.0 |
|
pentane |
1.380 |
1.9 |
1.2 |
|
toluene |
1.510 |
5.4 |
1.2 |
|
1,4
dioxan |
1.343 |
5.0 |
1.0 |
|
cooling
liquid FC-43 |
1.380 |
1.3 |
1.0 |
|
CS2 |
1.343 |
3.0 |
1.0 |
|
crude
oil |
1.470-1.570 |
0.8-1.4 |
2.0 |
Table
5 Building and natural materials
characteristics.(l=2 mm,
T=20C).
|
Material |
n |
tand ´ 102 |
l, r |
|
brick
(red) |
1.78 |
3.5 |
r=1.5 |
|
brick
(silic.) |
1.82 |
4.2 |
r=1.8 |
|
concrete |
2.40 |
5.5 |
r=1.7 |
|
asphalt |
1.50 |
8.0 |
r=1.3 |
|
sand |
1.55 |
2.5 |
r=1.8 |
|
soil,
18C soil,
-39C |
1.60
1.60 |
3.8
2.5 |
r=1.5 |
|
snow,
-36C snow,
-1C |
1.12
|
0.65 2.6 |
r=0.23 |
|
pine
tree wood |
1.4 |
3.4/2.0 |
r=0.5 |
|
glass,
window |
1.45 |
5.0 |
|
|
organic
glass |
1.60 |
1.5 |
|
|
marble |
1.50 |
1.0 |
|
|
ebonite |
1.67 |
1.1 |
l=0.6mm |
|
cardboard |
1.80 |
6.0 |
|
|
cautchuck |
1.66 |
30 |
|
|
glues |
1.57-1.72 |
1.0-2.0 |
l=1.7 mm |
|
phenolone |
1.8 |
3.9 |
l=1.7mm |
|
polymyde |
1.66 |
1.8 |
l=1.7mm |
|
veneer |
1.5 |
10 |
l=7.6mm |
|
plaster |
1.7 |
0.7 |
|
In
Tables 4 and 5 n is the real part of
the complex refractive index n*=n-ik and the loss tangent tand=2nk/(n
-k
).
In
Table 5, r is the density of a material (g/cm3). The lesser value of
tand for pine-tree wood corresponds to electrical field
perpendicular to wood fibers. The
water content in wood is less than 7%
Frequency
dependence of n in MM range is weak but tand varies considerably as frequency changes. For example tand of brick at l = 6.3 mm is only 0.01, tand of concrete is 0.005.
Table 6 gives the properties of cloths [8]. Hear l is cloth thickness, ½t½2
, ½r½2 are transmission and reflection power
coefficients
Table
6. Cloths materials properties. (l=1.6 mm,
T=20C).
|
Material |
½t½2, % |
½r ½2, % |
n- 1 |
l, mm |
|
cloths
for tents |
82-94 |
£ 1.3 |
0.2-
0.3 |
0.3
- 0.5 |
|
cloths
for coat, wool |
77-
84 |
£ 3.0 |
0.1-0.2 |
2-
4 |
|
cloths
for suits, wool |
85-
98 |
£ 1.0 |
0.2-
0.3 |
0.5-
1.0 |
|
silk |
89-
93 |
£ 1.0 |
0.28-
0.35 |
0.15-
0.25 |
|
leather,
natural |
79-
85 |
£ 5.0 |
0.22-
0.28 |
0.9-
1.5 |
|
leather,
artificial |
75-
89 |
£ 5.0 |
0.22-
0.28 |
0.7-
0.8 |
|
fur,
artificial |
75-
89 |
£ 3.0 |
0.04-
0.07 |
4.0-
12 |
|
astrakhan |
71 |
£ 1.0 |
0.14 |
35 |
|
cloths
for shirts |
92-
95 |
£ 5.0 |
0.18-
0.23 |
0.2-
0.3 |
All
materials in Tables 4 -6 have strong dependence of tand and êt ê2 on moisture, e.g. dry cloths (Table 6) have êt ê2 = 75-98%, while, for moisture 28%, ït ï2 =25-80%.
The
results of such
investigations allow one to clear up the possibility of testing the water
content in these materials, to estimate the accuracy of such control, and to
find the most appropriate measurement setup.
The
devices use a certain wave guiding structure connected to a material under
test. The measuring parameters (losses or phase constants of the propagating
wave) are changed due to the presence of this material.
Depending
on e¢ and e¢¢ of dry material and the water content W, structures are used where a
wave either propagates through the material, or is reflected from it, or some
part of the power of the wave penetrates into the material under test [5], [9],
[12]-[15].
We
investigated water emulsions [7], [9]. Of the greatest interest for practice is
the water content in a crude oil emulsion. In the MM wavelength range, crude
oil has e¢ = 2.10 -2.20 and e¢¢ = 0.002 -0.003 depending on sort. Losses in crude oil are changed from a =0.05 -0.07 dB/cm at wavelengths l = 8-7 mm to a = 0.13 -0.18 dB/cm at wavelengths l = 4-3mm as compared with a =
200 dB/cm and a = 300 dB/cm for water, respectively.
The
experimentally found characteristics of water in oil emulsions for different
content of water W depending on wavelength are presented in Table 7, T = 20C.
Table
7 Characteristics of water in crude oil emulsion vs water concentration and
wavelength
|
l, mm |
W,% |
e¢ |
e¢¢ |
a×10, dB/cm |
|
10 |
0.5 1.0 |
2.16 2.19 |
1.5 2.5 |
2.7 4.6 |
|
8 |
0.5 1.0 |
2.16 2.19 |
1.5 2.7 |
3.6 6.1 |
|
4 |
0.5 1.0 |
2.15 2.18 |
1.6 2.8 |
7.4 12.8 |
|
2 |
0.5 1.0 |
2.15 2.17 |
1.5
2.5 |
13.8
23.4 |
The data in Table 7 are in a good agreement with the Lichtenecker
formula at wavelengths greater than 5 -6 mm [9]
.
Here
e
and e
are the permittivities of
water and oil respectively, e is the
permittivty of emulsion.
In
the short-wave part of the MM range, these results are well described by the
Maxwell-Garnett model [9].
.
The
particle size of emulsified water was much smaller than the wavelength in the
medium.
Table
7 shows that the sensitivity to water in crude oil at the MM range is very
high. Therefore, it is possible to realize an aquameter for crude oil, capable
of measuring the water content in- flow irrespective of the oil sort and in a
wide interval of temperatures.
4 Devices for water monitoring in different media
Now
we consider some examples of devices for water testing in different materials
and substances.
One
of such devices is described in [9], [13]. This device uses waves with l = 8-10 mm in standard steel oil pipe with an internal diameter of 50 mm
and a central conductor. The device measures the absorption in a pipe filled by
crude oil under test. The measurement uncertainty DW better than 0.08% was achieved over an interaction length of about 40
cm in the moisture content range W = 0.1-1 %.
A
modified design [9] consists of rectangular exiting and receiving waveguides
with longitudinal slots. These waveguides are inserted at an angle j into a pipe containing flowing oil. The slots are hermetically sealed
by a dielectric. The angle j
corresponds to the Brillouin angle for the
mode in the
waveguides. For a slot length of 30 mm, a spacing between the waveguides of 30
cm, and a pipe diameter of 50 mm, the moisture sensitivity at frequencies 32
-35 GHz was found to be 13 -15 dB per 1% of water.
A
single slotted waveguide was used to monitor the presence of oil in water. In
this case, the sensitivity to oil was about 4 -5 dB per 1% increase in the oil
content for an oil concentration of W= 0.5 -2.0 %
The
same slotted waveguide was used to measure the alcohol content in water [9].
Table
8 presents the real and imaginary parts of the permittivity of water-alcohol
solutions.
Table
8. Permittivity of water-alcohol solution, (T = 20C, l = 9 mm)
|
W,% |
0 |
10 |
22 |
40 |
94,5 |
96 |
|
e¢ |
19.6 |
12.4 |
8.6 |
6.4 |
4.2 |
4.1 |
|
e¢¢ |
29.0 |
21.2 |
14.6 |
7.9 |
0.9 |
0.8 |
When
the alcohol concentration was W=8 -40%, the sensitivity at frequencies 35 -37
GHz was 0.30 -0.35 dB per 1% of alcohol. For W >40%, the sensitivity becomes
lower.
Therefore,
for low concentration of water in alcohol, a dielectric waveguide in contact
with a cell containing alcohol-water solution was used [14]. Measurements were
performed at frequencies of 25 -27 GHz. The dielectric waveguide was made of
high-resistivity silicon (e¢= 3.42, tan d= e¢¤ e¢¢= 0.001). The
mode was used in a
waveguide with dimensions of 1.4 ´ 2.8
mm. The length of a polyethylene cell with a liquid under test was 25 mm, and
the wall thickness was 1 mm. The losses introduced by the cell for water
concentration in alcohol in the range W= 4-10 % were 0.5 -0.7 dB per 1% of
water
The
next object of interest is powders. These materials are widely used, for
example, in metallurgy, and the water
content in such powders essentially influences the quality of output.
The
most interesting objects are metal-containing powders (e.g. iron concentrate,
charge), connecting materials (for example bentonite), and some other materials
used in the so-called powder metallurgy. For high-quality production, it is
necessary to guarantee that the humidity of powder should lie within a narrow
interval (no grater than fractions of a percent). In this and some other cases
it is necessary to check water content in-flow or periodically in a few time
intervals. Usial chemical methods using drying of material do not permit to
realize this testing.
Some
devices for the in-flow water content testing in powder media were elaborated
in [15]. One of them is a quasi-optical interferometer for measuring the
reflection coefficient ïr ï2 depending on moisture.
Another
device uses a waveguide supporting a slow wave. A part of power of this wave
penetrates into the material under test with complex permittivity e*, and, as a result, the transmission coefficient changes depending
on the material properties. Such type of moisture meters was realized at
frequencies 26-38 GHz [15]. The accuracy of water content determination is
about 0.2-0.1% or better.
In
addition, a device using a dielectric resonator was elaborated. The resonance
frequency and the quality depend on the material permittivity and moisture. The
resonator may rotate if either the material or the measuring device move.
These
measurements methods can also be used for the determination of water content in
timber and paper.
The
last device elaborated by us for the determination of water content in liquids
is a quasi-optical instrument for crude oil flowing in pipes under pressure.
This device uses short MM waves (W band) and allows one to measure the water
content from 0 to 3% vol. with an accuracy of 0.05% in real time.
5 Conclusions
MM
waves provide new possibilities for measuring the water content in media and
materials. However, these possibilities have not been extensively used in
practice. The reason is that the electrical properties of monitored objects
were not adequately investigated at MM wavelengths and, secondly, MM waves were
not widely used for moisture content monitoring despite the fact that such
monitoring was technically possible. The above examples of successful
application of MM waves to moisture content measurement suggest that we can
look forward to the wide application of MM waves in aquametry.
References
[1] Chantry G.:
Submillimetre Spectroscopy, Acad.Press, London-New-York, 1971.
[2] Afsar M.N.,
Birch J.R., Clarke R.N.: The measurement of properties of materials
Proc. IEEE (1986), vol.74, No.1, p.183 -190
[3] Meriakri V.V., Apletalin V.N., Kopnin A.N. et.
al.: Submillmere beam waveguide spectroscopy and its applications
in book
Problems of Modern Radio Engineering and Electronics, ed. .Kotelnikov V.A. Nauka Publishers, Moscow, 1985, pp. 179-197.
[4] Meriakri
V.V.: Spectroscopy of millimeter and
submillimeter ranges
Moscow
University Physics Bulletin (1992),Vol. 47, No. 3, pp. 81-88.
[5] Krazewski A.W.: Microwave aquametry- needs and
perspectives
IEEE
Trans. MTT (1991),Vol. 39, No. 5,
pp. 828-835.
[6] Afsar N.M., Button K.J.: Millimeter and submillimeter
wave measurements of complex optical and dielectric parameters of materials
International Journal of Infrared and Millimeter Waves (1981), Vol.2, pp.1029-1044.
[7]. Apletalin V.N., Garin B.M.,. Meriakri V.V.:
Dielectric properties of liquids in the submillimeter band
Radio
Engineering and Electronic Physics (USSR) (1983) Vol. l28, No.1, pp.1-15.
[8] Meriakri
V.V., Chigrai E.E.: Properties of materials for practical use at the MM and SMM
wavelengths
Digest
of the 18-th International. Conference.
on Infrared and Millimeter Waves, (Colchester, UK), 1993, pp. 68-69.
[9] Meriakri V.V., Chigrai E.E., Nikitin I.P.:
Monitoring the water content of media and materials with millimeter waves
Radio and
Communications Technology (Russia) (1996), Vol.1, No 2, pp. 92-96.
[10] Liebe H.J., Hafford G.A., and Manabe T.: A model
for the complex permittivity of water at frequencies below 1 THz
International. Journal. of Infrared and Millimeter Waves (1991), Vol.
12, No 2, pp. 92-96.
[11] Meriakri V.V.: Material properties in the
millimeter range
Proceedings of the3-th International Symposium «Physics and Engineering
of MM and SubMM waves», Vol. 1, 1998, pp. 121-123.
[12] Fernandes H.C.C.,. Meriakri V.V.: Materials for
application in MM and SubMM ranges
Proceedings of 1999 SBMO/IEEE MTT-S International. Microwave and
Optoelectronics Conference, (Rio de
Janeiro-RJ), pp. 200-2002.
[13] Gershgoren V.A., Gokhman V.B., Meriakri V.V. et.
al.: Microwave device for water content measurement in oil and oil products
Russian
Patent Bulletin (1992) 2073859 G01,
No 22/04
[14] Meriakri V.V., Parkhomenko M.P.: Application of
dielectric waveguide for water control in alcohol
Electromagnetic Waves and Electronic Systems (2000), Vol. 5, No 1, pp.
38-39, (in Russian).
[15] Meriakri V.V.,. Pangonis
L.I.: Microwave methods moisture testing into materials and media used in
metallurgy
Abstracts
of the 28-th October Mining and Metallurgy Conference, 1996,(Bor, Yugoslavia).
Corresponding
author Meriakri Viacheslav Viacheslavovich, Head of Laboratory of Millimeter
and Submillimeter Waves Measurements, Institute of Radio Engineering and
Electronics Russian Academy of Sciences, Vvedenski sq. 1, Fryazino Moscow
Region, 141190, RUSSIA.
Tel:
07095-5269266 FAX: 07095-7029572
e-mail:
meriakri@ms.ire.rssi.ru meriakri@ms.ire.rssi.ru
http:www.meriakri.de.vu