LABOSSIERE STREET RESULTS

Table 2. Material properties used in numerical simula-

tions.

Conductivity Density Specific heat Latent heat

(J/hr m C) (kg/m3)

(J/m3)

LABOSSIERE STREET RESULTS

The 19951996 entries show the frost penetra-

tion simulation using the corrected thermal con-

The three winters when this project was run-

ductivity for the ledge and the uncorrected high

ning were not relatively cold ones. However,

thermal conductivity (k). This shows the dramatic

whatever the temperatures, if the numerical

difference between assuming pure ledge conduc-

model is valid, it should be able to mimic accu-

tivity and adopting a more realistic conductivity.

rately what was recorded. Figure 13 shows the

The measured maximum frost depth was 1.32 m

maximum frost penetration measured at the con-

(4.3 ft). The simulation using the corrected con-

trol string inserted in ledge on Labossiere Street,

ductivity gave a depth of 1.4 m (4.6 ft) and the

compared with the simulated frost penetration.

uncorrected conductivity resulted in a depth of 2

The winter of 19941995 has two entries. The first

m (6.6 ft), 43% higher than the uncorrected con-

shows the maximum frost penetration up to a

ductivity. This underscores the importance of

prolonged warming spell (with near-record

having accurate physical properties for the mate-

warm temperatures in January) of 1.04 m (3.4 ft)

rials on site.

with the simulation being 1.08 m (3.5 ft). The sec-

The 19961997 maximum frost penetration

ond maximum penetration happened after this

came around mid-February at 1.14 m (3.7 ft). The

warm period, in early March, and was 1.06 m (3.5

predicted depth was 1.1 m (3.6 ft).

ft); the numerical was 0.94 m (3.1 ft).

Figure 13. Actual and predicted frost penetration data. Winter of 19941995 has

two entries because of long warming spell in middle of winter. Effects of uncorrected

and corrected ledge thermal conductivity values are shown for winter of 19951996.

11