146
the temperature response of the copper cylinder increased. This was
due to the fact that the Biot number approached infinity, thus causing
the eigenvalue for an infinite Biot number to be used. Thermal
diffusivity was affected by almost a one to one correspondence by
changes in the slope of the soil sample temperature response.
Another source of analytical error would be using the solution for
an infinite cylinder instead of that for a finite cylinder. The
analytical solution for a finite cylinder was determined by the
principle of superposition of the solution for an infinite cylinder and
an infinite slab (Equation 4-15).
9 s *cyl*slab (4"15)
Substitution and rearranging terms yields a linear equation with the
slope of the line consisting of a linear combination of the eigenvalues
for a slab and cylinders.
0 = c
cyl
cylcslab exPt'Dt( ^
(4-16)
This procedure was incorporated into the analytical analysis and the
thermal diffusivity determined. No significant differences were
obtained.
Errors in temperature measurement would comprise the major source
of experimental errors. Conduction error occurs due the sensor passing
through regions of differing temperatures. Heat can then be conducted
from the soil sample to the water along the thermocouple. Gaffney
et al. (1980) stated that conduction error caused by the surrounding
fluid could be minimized by using 36 ga or smaller thermocouple and by