Measurement and Estimation of State Variables - Unsaturated Soil Mechanics in Engineering Practice - page 142

was about 7 mV. The thermal conductivity suction sensors
were calibrated in the laboratory at a temperature of 23 ◦ C:

the following temperature change correction equation was
proposed:

T 23 ◦ C = T s 1 + s 2 T + s 3 T 2

(4.10)

0 . 0014 T 0 . 5743

0 . 6065

+

T 23 ◦ C = T T (4.9)

where:

T =

change in temperature measured by the
TC suction sensor,

where:

T 23 ◦ C = corrected temperature change that can be

applied to the 23 ◦ C calibration curve,

T 23 ◦ C = corrected rise in temperature corresponding to

23 ◦ C and

T = ambient temperature recorded by the TC

suction sensor, and

T T = rise in temperature measured by the TC sensor

at the ambient temperature T .

s 1 , s 2 , and s 3 = constants that depend on the range of

suction being measured.

The temperature correction proposed by Nichol et al.
(2003) is dependent not only on the ambient temperature
but also on the magnitude of the soil suction. It is apparent
that more variables are taken into consideration in the
development of the Nichol et al. (2003) correction. The
correction constants vary with the level of soil suction and
suggested values are presented in Table 4.8.

The proposed correction equation was then applied to the
data presented in Fig. 4.52. The temperature-corrected soil
suction measurements are shown in Fig. 4.53. The results
show that the variation in the sensor reading was reduced to
about 2mV when the ambient temperature correction was
applied.
Shuai et al., (2002) undertook a series of tests where the
ambient temperature was varied from 8 to 23 ◦ Cinaseries
of steps. The output voltage was measured and the tem-
perature correction was applied to the recorded readings.
The constant suctions applied in each case were 10, 50, and
200 kPa. The results obtained when the applied suction was
50 kPa are shown along with the measured temperatures in
Fig. 4.54. The results indicate that the suctions remained
essentially constant when the proposed temperature correc-
tion was applied.
Nichol et al. (2003) observed that ambient temperatures
influenced the thermal conductivity suction sensor readings
and proposed a more rigorous model to take into account
the effect of ambient temperatures. His studies involved the
use of the GCTS thermal conductivity suction sensors and

Table 4.8 Temperature Correction Constants
for Nichol et al. (2003) Equation

Suction Range
kPa s 1 , ◦ C s 2 , ◦ C − 1 s 3 , ◦ C − 2

10 − 3 10 − 6

0-10 0.960 1 . 9 × − 0 . 8 ×

10 − 3 10 − 6 10-30 0.962 1 . 8 × − 0 . 7 ×

10 − 3 10 − 6 30-75 0.968 1 . 5 × − 0 . 6 ×

10 − 3 10 − 6

75-300 0.973 1 . 3 × − 0 . 5 ×

300-10 6 10 − 4 10 − 6 7 . 0

× −

0 . 3

×

0.986

26

25

24

23

22

916

914

912

910

908

Without temperature correction
With temperature correction

906

0 5 10 15 20 25 30 35 40 45

Time, h

Figure 4.53 Soil suction readings “with” and “without” applied temperature corrections (after
Shuai et al., 2002).