ment structure. The final simulation included the
Model 3--class 3 and class 4; and Model 4--both
lower-heaving subgrade (sample 1232) under the
clay subgrades.
5-yr full-depth section (ML5-F-4), termed case
In all of these simulations, it was assumed that
f4w6ss, or "second subgrade." One of the 1206
the pavement properties were constant from one
subgrade simulations, also the 5-yr full depth sec-
pavement test section to another. Some of the test
tion (case f4w6ld) used a lower dry density of
sections will possess properties different from oth-
1.69 Mg/m3 (105.5 lb/ft3) for the subgrade, as com-
ers because of the experimental design of the paved
pared with what was set as the "normal" case, or
surface. Some variations in the asphalt pavement
1.89 Mg/m3 (118 lb/ft3).
properties could be considered in subsequent mod-
eling efforts, but we maintained the same asphalt
Material properties
pavement properties in all of these simulations.
Material properties input to the FROST pro-
gram are shown in Table 14. Note that the class 3
Results--flexible sections
special, class 6 special, and two subgrade materi-
Figure 19 is an example of output from FROST
als were tested as part of this study to determine
for the 5-yr full depth simulation (f4w6). The top
the information for input to the model. The class 4
graph shows the daily mean air temperature for
special and class 5 special materials were not
the period from 1 October 1959 to 14 November
tested, and their behavior was approximated using
1960. The center portion illustrates the predicted
data from previously tested materials that most
frost heave and the bottom graph contains the pre-
closely matched their specified size gradations. A
dicted frost and thaw penetration as functions of
subbase from taxiway A at the Albany, New York,
time. Frost output graphs for all the flexible cases
airport most closely matched the class 4 special
are compiled in Bigl and Berg (1996b). The freeze
subbase specifications; and a dense-graded stone,
season simulated was characterized by several
from Winchendon, Massachusetts, most closely
short freeze-thaw events early in the season, fol-
matched the class 5 special material.
lowed by a continuous severe freeze event with a
In this initial modeling series, the equations
fairly rapid spring thaw. Table 13 contains maxi-
used in TRANSFORM to calculate the modulus
mum frost heave and maximum frost penetration
of base, subbase, and subgrade materials were the
depths for each of the cases simulated. Simula-
gravimetric form of the frozen equations (Table
tions with shallower water table depths had greater
1a) and the unfrozen equations in Table 1b. For
amounts of heave and less frost penetration com-
the class 6 special material, we used the semilog
pared to simulations with a deeper water table
form listed in Table 1b. The asphalt concrete modu-
location. Sections that included the substitute for
lus was calculated with the Schmidt (1975) rela-
the class 4 special subbase (taxiway A subbase)
tionship.
had higher amounts of heave than those with other
As stated earlier, after these Phase 1 simula-
combinations of subbase materials.
tions were completed, we discovered that resilient
Figure 20 presents the moduli being calculated
modulus tests on the 1206 subgrade in the unfro-
by TRANSFORM and passed as seed moduli to
zen condition were probably in error due to a
NELAPAV for the f4w6 case. The plotted moduli
miscalibrated testing system. As a result, the unfro-
are the minimum values for each day, with the
zen subgrade moduli predicted in Phase 1 modeling
subgrade (top) being located within 0.3 m (1 ft) of
are likely to be substantially higher than exist in
the asphalt and subgrade (bottom) being the rest
the field, resulting in less damage than would have
of the modeled section. Note that the predicted
been obtained with more reasonable moduli.
asphalt moduli during the summer months are
In the NELAPAV program, we used the linear
smaller than those of the subgrade, which results
model for paving materials, and for the base, sub-
in high horizontal strains at the base of the asphalt
base, and subgrade when frozen. Nonlinear mod-
layer during the summer months. In later simula-
els of various forms (Table 2) were assigned to
tions of this study, we used another model for
unfrozen base, subbase, and subgrade as follows:
predicting asphalt modulus at temperatures above
1C.
Model 1 (modified to semilog form)--class 6;
29
Previous Page
