Stone shape should be blocky, not elongated, to
enable the stones to "nest" together more effectively
to increase their resistance to movement. We used
rocks within a certain range of shape factor, which
is defined roughly as the ratio of the longest di-
mension to the shortest dimension of the rock, to
limit the number of rocks that were too elongated
to be effective as riprap. The guidance provide by
U.S. Army Corps of Engineers (1991) is that no more
than 30% of the material should have shape factors
greater than 2.5, and no more than 15% should ex-
ceed 3.0.
The stone size is defined as the equivalent spher-
ical diameter D of a stone of a given weight W, and
it can be calculated using the standard equation for
Figure 10. Partially installed layers of sand, filter fab-
computing the volume of a sphere:
ric, and stones used to simulate a riprap-protected bank.
In this view, the model and supporting platform are at-
D = {6W/(πγs)}1/3
tached to the main carriage in the test basin.
where γs is the specific weight of a stone. Conven-
The model bank riprap installation simulated
tionally, stone size is expressed as Dn , where D
a typical field installation as much as possible. We
denotes the equivalent spherical diameter of the
placed a 38-mm- (1.5-in.-) deep layer of sand on
stone and the subscript n denotes the percentage
the bank (held in place by the transverse alumi-
of the total weight of the graded material that con-
num angle strips), covered the sand with a filter
tains stones of smaller weight. Similarly, the stones
fabric, and placed manufactured stones to model
can be graded according to the weight expressed
the riprap material (Fig. 10). The practice in many
as Wn, where W denotes the weight for which a
field applications is to apply the riprap over a fil-
percentage n is lighter on the cumulative distribu-
ter material, which prevents fine materials of an
tion curve. In the U.S. Army Corps of Engineers
underlayer from being washed through the voids
(1991), standardized gradations having a narrow
of a coarse upper layer. Where geotechnical con-
range in sizes (1.4 < D85/D15 < 2.2) are given in tab-
ditions permit, a filter cloth (or geotextile) is often
ular form. The manual recommends that the grada-
specified for the filter to lower construction costs.
tion limits should not be too restrictive to make the
This type of filter material offers less frictional re-
production cost high. In general, the guidance sim-
sistance to the riprap compared to a filter made of
ply suggests that gap-graded or widely graded
gravel. In previous laboratory studies on the an-
stone distributions may not be suitable as effective
gle of repose of riprap, the riprap was found to be
riprap. According to the manual, most graded
less stable when placed on filter cloth than when
riprap have ratios D85/D15 less than 3; for our tests,
placed on a simulated gravel filter. We used the
we followed this guidance for the model riprap to
filter cloth for the underlying filter material in our
have the ratio D85/D15 less than 3.
model, because the use of a filter fabric in our tests
We obtained manufactured stone from a local
will better simulate the conditions for the grow-
quarry (West Lebanon, N.H.) for the model riprap.
ing number of bank protection projects using geo-
The rock from this source is a metamorphic and
textile filters, and it will be conservative for those
has a density of 2600 kg/m3 (160 lb/ft3). The shape
riprap installations using a gravel filter.
of an individual rock was typically blocky to slight-
ly elongated with angular edges. We manually sift-
Sizes and gradation of stones
ed the rocks to reject rocks with shape factor great-
We used scaled-down stones in approximate
er than 2.5. The large-size stones were sieved into
gradations as suggested by the U.S. Army Corps
the desired size fractions because of the unavail-
of Engineers (1991) to model the riprap. In gener-
ability of standard sieves and the difficulty in han-
al, we followed the guidance given by the Corps
dling.
on the stone characteristics and installation tech-
Although stone size is often represented in terms
niques for the purpose of making model ripraps.
of the D50 or D30 size for open-water riprap instal-
The guidelines and the convention used in speci-
lations, we have found that, for the case of ice shov-
fying stone size are summarized here for reference.
8
Previous Page
