Environmental Engineering Reference
In-Depth Information
5.6.2.2
Measurements on Sand Samples
After the validation of the reference-model, the proposed method was tested on three
different types of construction sand, differing in granulometry (ranging from approx-
imately 1 mm to 4 mm) and composition (sand #1 and sand #2 were siliceous sands,
whereas sand #3 was a calcareous sand). Moisture measurements were performed
on sand samples brought at increasing (pre-established) levels of moisture, through
the gravimetric method. Each gravimetric water content progressively achieved was
then associated to the corresponding volumetric water content (
) through the mass
density of the sand. It is important to underline that samples were moistened al-
most up to water saturation level. In this way, for each type of sand, a wide range
of reference moisture levels was obtained (referred to as
s
1
θ
s
6, respectively).
The TDR80E04 was also used for performing 'traditional' comparative TDR
measurement on the same samples. For this purpose, the apparent relative permit-
tivity values of the reference samples,
,
s
2
, ...,
ε
app
, were evaluated (directly in TD) through
separate measurements of the apparent length (
L
app
) of the 15 cm-long three-rod
probe (Campbell CS630) inserted into the MUT. The
ε
app
values were evaluated
through (5.5) [2]. Successively, on the same sample, measurements were carried out
using the antenna as a probe. For the measurements with the antenna, the sample
was placed in a top-opened plastic box with dimensions 20 cm
4 cm. Full-
wave electromagnetic simulations of the radiation pattern of the antenna, confirmed
that boundary effects caused by the box were negligible.
For all the samples (
s
1
×
20 cm
×
,
s
2, ...,
s
6) of each sand, the
S
11
(
f
)
of the antenna in
contact with the MUT was evaluated.
As an example, the changes in the
S
11
(
magnitude curves for increasing values of
moisture of the MUT for sand #3 are shown in Fig. 5.17: starting from the antenna
in air, each curve refers to a different moisture level (
f
)
θ
) ranging from 0% (
s
1) to the
maximum considered moisture level (
s
6).
To have a tentative value of the relative permittivity of the MUT, traditional com-
parative TDR measurement of the
app
were performed, so as to associate these data
to the
f
res
measured for the same sample. Results are summarized in Table 5.7.
As expected, as water content is increased, the value of
ε
ε
app
progressively differs
from the value of the theoretical
ε
m
. This behavior is mainly attributable to two
main effects. The first is the fact that dielectric losses become more relevant as the
amount of water is increased. In fact, the measured
ε
app
carries a strong contribu-
tion of the imaginary part of the relative permittivity (
ε
), and cannot be merely
ε
. In this case, the theoretical model should include a correction
factor accounting for
assumed equal to
ε
[6]. Secondly and more importantly, as discussed in [6]
and as confirmed by measurements, the proposed method is best suited for monitor-
ing materials whose corresponding relative permittivity does not exceed 10-12. For
higher permittivity values, due to the intrinsic degradation of the resonant behav-
ior of microstrip antennas, a strong attenuation of the resonance effect is observed.
Therefore, the theoretical model described by Eqs. (5.33)-(5.35) becomes less ac-
curate in describing the changes in the resonant frequency as a function of the per-
mittivity. Deviation between
ε
m
and
ε
app
are highlighted, for sand #3, in Table 5.8.
Search WWH ::
Custom Search