The design is based on the application of Fourier’s law of onedimensional
heat conduction under steady state condition. From
the definition of the one dimensional steady state, it is assumed
that the heat flow is considered to be in one direction, which
means that no lateral heat transfer occurs within the specimen.
Practically, it is difficult to achieve this condition because of the
ambient temperature interference (ATI) which produces additional
radial temperature distribution to the desired axial temperature
gradient [26]. It is essential to consider the effect of the ambient
temperature on longitudinal and radial heat flow. The common
method to minimize the effect of ambient temperature is to use
an insulation layer, keeping the ambient temperature constant
and controlling the temperature of sink disk. There are three options
for the sink temperature. The first is to keep the sink temperature
higher than the ambient temperature, which will produce a
contrived heat flow from the sink into the specimen. The second
option is to keep the sink temperature below the ambient temperature.
This will make the outer portion of the specimen cooler
which may give more chance of heat to inject radially into the
specimen. The third (and the best) option is to keep the sink temperature
near the ambient temperature; however the radial loses
will be linked to the efficiency of the insulation. In fact, to establish
unidirectional heat flow, it is required to have a mechanism that
separates and controls the two kinds of the heat flow (longitudinal
and radial heat flow). In order to minimize the effect of ATI and to
maximise heat flow in one direction, our apparatus is designed to
control the radial heat flow by constructing a new layer with
adjustable temperature (thermal jacket) and to keep the temperature
of the sink disc near ambient temperature. The level of the
thermal jacket temperature will control the amount and the direction
of the radial heat flow along the specimen length. As the temperature
of the thermal jacket is kept below the minimum
specimen temperature, heat will flow from the specimen to the
ambient. The amount of this heat depends on the difference in
temperature between the specimen and the ambient. On the other
hand, heat will flow from the thermal jacket into the specimen as
the thermal jacket temperature is kept higher than a certain level
of temperature. This means that there is a place where there is
no radial flow. Since the ambient temperature is lower than the
thermal jacket temperature, the heat generated by the thermal
jacket will flow from the jacket to the ambient not into the specimen.
The idea is just to keep the temperature of the thermal jacket
at the desired level and so that it loses its heat to the ambient. Another
source of heat leakage that in some designs can negatively
contribute to the measurement is the base heat loss. To eliminate
the effect of the base heat loss, in the equipment reported here
the heater is inserted between two identical specimens. Consequently,
the input power used in the calculations is divided by
two. However, in this case the symmetry of the specimens and
the apparatus itself are influential. Also, the heat flux, which is de-
fined as the amount of the heat that passes through a unit crosssectional
area, should be uniform across the specimen. To achieve
this, the heater and sink discs must be as flat as possible and made
of highly conducting and emissive material. It is important at this
point to highlight the effect of the contact resistance which is de-
fined as the resistance to heat transfer at an interface between
adjoining objects of different shapes or roughness due to poor
physical contact, on the estimation of heat flux using heat flux meters.
The poor contact between the soil and heat flux meters, in
some thermal cell configurations, can cause an underestimate of
the heat flux by a significant value. For example, if there is an air
gap of 5% of the thickness of the plate, this would lead to an underestimate
of the heat flux of upto 54% [28]. In the new thermal cell,
the heat flux is directly determined from the heat power input and
all temperatures used in the calculation of the thermal conductivity
are measured within the length of the specimen, which means
that any concern of the contact thermal resistance is not relevant.
The number, positions and the directions of the measuring thermocouples
are also very important, because the thermal conductivity
of the thermocouple material is very high compared with the soil.
This may cause unpredictable heat flow through the thermocouples
if not appropriately considered. Finally, as the diameter of
A. Alrtimi et al. / International Journal of Heat and Mass Transfer 72 (2014) 630–636 631
the specimen is limited to the U100 tube (103 mm), the length of
the specimen has a large effect on the amount of radial heat loss.
Short spec