characterization forms included estimated per-
age (as shown in Fig. A1b). Gravimetric water
cent of unsurfaced trail section covered by still
content of nonsubmerged material ranged from 8
water, initial rut depths, etc. (App. F).
to 34%, and averaged approximately 19% with a
standard deviation of 4.4.
Thaw depths on the trail ranged from 10 to 41
cm (4 to 16 in.) and averaged approximately 25
TEST SITES AND DESCRIPTION
cm (10 in.). Thaw depths outside the travelway
BASED UPON PRECONSTRUCTION
(alongside the woods) were appreciably less, prob-
SITE CHARACTERIZATION TESTING
ably due to shading. Additionally, the right wheel
path seemed to exhibit greater variability in thaw
Wooded trail
In its original condition, this narrow 550-m-
depth than did the left wheel path. Orientation
(1800-ft-) long trail through the woods was im-
and shading may have been responsible for this
passable by vehicles. The travelway consisted
as well. For the wooded trail, where construction
of a thin (0 to 10-cm- [0 to 4-in.-]) vegetative/
and testing spanned several days, additional soil
organic mat atop a sandy soil. Gradation curves
moisture and thaw depth were measured on an
of the sandy soil are shown in Appendix A (Fig.
interim basis to document changing conditions in
A1a). Terrain ranged from flat to gently slop-
the soil.
ing. Although relatively flat and wet every-
Based upon CBRCLEGG, the material in the
where, the trail did not appear to be uniform
wheel paths was slightly stiffer than that along
(indicating a possible unequal frame of refer-
the centerline. This is probably a result of com-
ence for comparing test section performance
paction due to past use/trafficking. Histograms
evaluation. For example, 5 cm [2 in.] of rut in a
showing the distribution of gravimetric water con-
tire chip section on a dry subgrade cannot be
tent and CBR are shown in Figures 1ad. Typical
compared to 5 cm [2 in.] of rut in a chunkwood
CBRDCP profiles for stations 6+00 and 18+00 at
section on a wet subgrade).
left, right, and center of the travelway are shown
The sampling/testing grid for the wooded trail
in Figure 1c. CBR in this figure was calculated
is shown in Figure A1c. Site characterization tests
using DCP values, and plotted using the Pave-
were conducted and samples were taken, with a
ment-Transportation Computer Assisted Struc-
few exceptions, at 7.6-m (25-ft) intervals along
tural Engineering (PCASE) program "DCP" (U.S.
Army 1995). The figure shows substantial strength
centerline and at 30.5-m (100-ft) intervals in the
variability both horizontally and vertically. Again,
right and left wheel paths. Site characterization
this provides additional evidence pointing toward
evaluation forms were completed for each 15.2-m
a possible unequal frame of reference for test sec-
(50-ft) section from station 0+00 to station 18+00
tion performance evaluation.
(all stations are indicated in feet). Additionally,
centerline elevations were measured every 15.2 m
(50 ft).
Sloped trail
The sloped site consisted of sections of two
The plan and profile are shown in Figures A1d
intersecting trails. Both trail sections (collectively
and e, respectively. Two construction crews
termed the sloped trail) had erosion gullies. Prior
worked simultaneously at either end of the trail,
to site characterization and construction, the
building toward the middle. The control sections
sloped trail was both graded to remove the ero-
near the center of the trail experienced minimal
sion gullies and filled to yield a relatively uni-
disturbance prior to trafficking, while control
form 16 to 18% grade. The corner was incorpor-
areas on the south end of the trail became impass-
ated as a control section. Because of recent use,
able during the construction phase and required
the sandy travel surface was not covered with an
improvements. These NOGO situations are fur-
organic layer as were the wooded and pentagonal
ther discussed in a subsequent section dealing
loop trails. The gradation of the sandy material is
with vehicle mobility.
shown in Appendix A, Figure A2a. The sloped
The wooded trail consisted of a saturated, thaw-
trail site is shown in Figure A2b, and the plan and
ing soil layer over frozen ground. The ground-
profile are shown in Figures A2c and A2d, respec-
water table was near the surface. Well-defined
tively.
ruts in both wheel paths often held standing
Based upon soil sample tests, gravimetric water
water. An estimated 25% of the entire 550-m- (1800-
contents ranged from 6 to 16%, and averaged 11%
ft-) long trail was covered by standing water; each
with a standard deviation of 1.9. Figure A2e shows
15.2-m (50-ft) section ranged from 0 to 75% cover-
7