145
to many small points. As the water content increased, a film was
formed around the soil particle causing a continuous contact surface.
Increasing the water content increased the heat capacitance and the
thermal conductivity as well by displacing some of the air from the
pore spaces. The increase in thermal conductivity and contact area
occurred at a faster rate than did the increase in thermal capacitance.
At some point, the rate of increase in contact area and thermal
conductivity becomes slower than that of heat capacitance; therefore,
the thermal diffusivity begins to decrease from the maximum.
Sources of experimental error arise from measurement errors and
analytical errors. Analytical errors can arise from errors in
determination of the slopes of the temperature response for the copper
cylinder and the soil sample and estimates of the volumetric heat
capacitance. Inclusion of points from the nonlinear portions of the
temperature response curve for either the copper or the soil samples
could cause a change in the slope. A miscalculation of heat
capacitance of the soil could influence the final determination of the
thermal diffusivity, also. To determine the relative influence of
these type errors the solution routine was run and varying each of the
parameters over a range experienced during the tests. The slope of the
temperature response of the copper cylinder and heat capacitance had a
minor effect upon the final value of the thermal diffusivity (Figure
4-4). A variation of plus or minus 50 % in the heat capacitance
exhibited a plus or minus 2 % change in the thermal diffusivity. A
thirty percent reduction in the slope of the temperature response of
the copper cylinder caused a 2 % increase in the thermal diffusivity.
The deviation in thermal diffusivity decreased to zero as the slope of