value produced by the CUMDAM program and
3. Most of the sections are predicted to fail due
multiplying by the number of applications received
to cracking of the pavement caused by tensile
by the sections during that time period (562,830
stresses at the bottom of the pavement much be-
per year). This application rate was based on an
fore they will fail due to rutting caused by exces-
initial estimate that the traffic across the test sec-
sive deformation of the subgrade. It is only the
tions would be 2,815,000 ESALs in five years, or
simulations of the full-depth sections (ML5-F-4,
5,630,000 ESALs in 10 years. Subsequent to con-
case f4w6, and ML10-F-14, f14w50) that indicate
ducting the Phase 1 series, a revised estimate is
rutting failures may occur before excessive fa-
that the 5-yr sections will receive 3,300,000 ESAL
tigue cracking.
applications in five years, (i.e., are designed to fail
Figures 2224 illustrate examples of the rela-
after 3,300,000 ESAL applications); the 10-yr sec-
tion between accumulating damage and time. Ad-
tions will fail at 6,600,000 ESAL applications.
ditional plots of heave, frost penetration, and cu-
Individual values listed in Table 15 vary any-
mulative damage for all the flexible sections are
where from very much below to very much above
compiled in Bigl and Berg (1996b).
the failure values described. By comparing values
Damage related to the horizontal strain criteria
horizontally across the table for any case, it can be
has different patterns with time for different mod-
seen that different models estimate substantially
els. Damage predicted by the MS-1 (Asphalt In-
different pavement lives for the same strain condi-
stitute) model has some increase during winter
tions.
thaw events and during the main spring thaw, but
A close review of the values in Table 15 illus-
a majority of its increase is in the summer period.
trates a variety of interesting occurrences. Some
Predicted damage from the MS-11 (Witczak)
of them are the following:
model has a consistent pattern of rising during
1. Cases with Mn/ROAD test section ML5-F-4
winter and spring thaw periods, and remaining
that employed the lower density subgrade (f4w6ld)
constant during the summer. The pattern of dam-
or the 1232 subgrade (f4w6ss) predicted the pave-
age accumulation from the Corps of Engineers
ment to fail much more quickly than when the
model is similar to the MS-1 model, especially
higher (optimum) density 1206 subgrade was used.
when the water table is shallow, causing larger
This is due to the fact that the higher density 1206
resultant deflections and strains. For simulations
subgrade used in most simulations maintained a
with deeper water tables, smaller amounts of dam-
resilient modulus during the summer months in
age accumulate during the summer.
excess of 276,000 kPa (40,000 lb/in.2), but the
Damage from models using the vertical strain
modulus values for the same period were about
criteria follow three general patterns: 1) a sharp
96,000 kPa (14,000 lb/in.2) for f4w6ld and
Table 16. Applications to failure (1000) from simulations of
about 6900 kPa (1000 lb/in.2 ) for f4w6ss.
rigid test sections.
The "low density" predicted moduli are
the closest to the values measured during
RCON = 650 RCON = 500 RCON = 650
the summer months with an FWD.
Case
SCI = 80
SCI = 80
SCI = 90
2. Variations in water table position in
5-Year
the same test section result in different
R5W6
52,601
507
45,244
predicted lifespans, especially if the wa-
R5W12
65,598
605
56,452
ter table is less than 3 m (10 ft) deep at
R6W4
157,655
1,505
135,622
one location and more than 6.1 m (20 ft)
R6W12
132,120
1,701
113,933
deep at another. Both the ML5-F-2 and
ML5-F-3 test section simulations illus-
10-Year
>5.6 108
>5.6 108
trate a shorter predicted life with a shal-
R11W6
18,761,000
>5.6 108
>5.6 108
R11W12
28,141,500
low water table depth. For the shallow
water table conditions, greater availabil-
Notes:
ity of water results in higher thaw weak-
RCON = concrete flexural strength, lb/in.2
ening (reduced modulus values) in the
SCI = surface condition index at failure
spring.
Traffic simulated at rate of 562,830 ESAL applications/yr.
32
Previous Page
