Environmental Engineering Reference
In-Depth Information
TDR-based conductivity measurements rely on the Giese-Tiemann (G-T) method,
which is based on the linear relation between
σ
0
and the static conductance of the
load,
G
s
[22]:
σ
0
=
K
p
G
s
(5.20)
where
K
p
is the probe constant. The value of
G
s
is evaluated from measurements
of the reflection coefficient at longer times (i.e., when the TDR has achieved the
steady state, approximately corresponding to the zero-frequency response), of the
probe inserted in the MUT,
ρ
∞
:
1
Z
TDR
1
−
ρ
∞
G
s
,
G
−
T
=
(5.21)
1
+
ρ
∞
where
Z
TDR
is the output impedance of the used TDR instrument. The value of the
probe constant can be estimated from the probe geometry:
=
ε
0
cZ
p
L
K
p
(5.22)
where
L
is the length of the used probe,
Z
p
is the characteristic impedance of the
probe,
c
is the velocity of light in vacuum, and
ε
0
is the permittivity of vacuum. In
(5.22), the evaluation of
Z
p
is not always an easy task. For this reason, to obtain
more accurate estimation of the electrical conductivity, the proportionality constant
K
p
is usually determined empirically, through multiple preliminary calibration mea-
surements on electrolyte solutions of known conductivity.
The G-T approach neglects the series resistance of the cable, connectors, and ca-
ble tester; as a result, for higher electrical conductivity values (i.e.,
2Sm
−
1
),
σ
0
>
0
.
this method underestimates
σ
0
. For this reason, in [28], the series resistor model
was introduced, thus considering the coaxial cable and the sample as two resistors
in series. The G-T method was modified as follows:
K
p
σ
0
=
(5.23)
(
1
/
G
s
)
−
R
cable
where
R
cable
is the resistance due to the cable. Castiglione-Shouse (C-S) proposed an
alternative approach for taking into account cable resistance; their method required
the correction of the measured reflection coefficient by scaling it with respect to the
reflection coefficients measured with the probe in air,
ρ
∞
,
air
, and short-circuited at
the distal end,
ρ
∞
,
sc
, respectively [8]:
ρ
−
ρ
∞
,
air
ρ
∞
,
air
ρ
∞
,
scaled
=
2
−
ρ
∞
,
sc
+
1
.
(5.24)
However, Lin et al. argued that, despite providing accurate results, the C-S method
is theoretically incorrect and that the series resistor model is indeed accurate [41].
In particular, Lin et al. suggested that the C-S scaling method actually accounts for
Search WWH ::
Custom Search