Design, specification and construction of filters - Geotechnical Engineering of Dams - page 351

9.2.1.1 Original USBR method
The original USBR method as described in USBR (1977) was:

5 to 40, provided that;
(b) the filter does not contain more than 5% fines passing 0.075 mm, and the fines
should be cohesionless.
(ii) D 15F /D 85B

(i) (a) D 15F /D 15B

5;
(iii) The grain size curve of the filter should be roughly parallel to that of the base material;
(iv) Maximum size particles in filter

75 mm to prevent segregation during placement;
(v) For base materials which include gravel particles, the base material D 15B and D 85B etc.
should be analysed on the basis of the gradation of the soil finer than 4.7 mm.

5 ensures that the filter is more permeable than the base
soil. Criteria (ii), (iii) and (ib) are designed to ensure that the filter is sufficiently fine to
control erosion of the base soil.
The authors' experience with applying this method was that, while it could be success-
fully applied to base soils which are clayey sands or sandy clays, it cannot be rigorously
applied to base soils with a high clay and silt content. In particular:

Criterion (i), i.e. D 15F /D 15B

- Rule (iii) cannot be followed (this is also the experience of Sherard and Dunnigan,
1985);
- Rule (ib) will almost always overrule (ia), often to the concern of inexperienced engi-
neers, who feel that the “mathematic rule”, i.e. (ia), should stand.

The authors' approach has been to be comfortable with rule (ib) overriding rule (ia), to
ignore rule (iii) and substitute in its place the US Corps of Engineers (1941) requirement
that the uniformity coefficient D 60F /D 10F

20.
However we would not now use this method, nor does the USBR.

9.2.1.2 Sherard and Dunnigan method
The USSCS (United States Soil Conservation Service) (Sherard et al., 1984a, 1984b and
Sherard and Dunnigan, 1985) carried out extensive laboratory testing to check filter cri-
teria. They used several different test apparatus to simulate a concentrated leak in a dam.
These included:

-Tests using the equipment shown in Figure 9.6 for base soils in the fine to coarse sand
(D 15B

0.075 mm to 2.36 mm) range;
-Tests using slot and slurry tests on silt and clay base soils, in which a small amount of
erosion was accepted as a successful filter;
-Tests using the “no-erosion” filter test apparatus shown in Figure 9.7 . For a filter to be
successful it was required that there should be no visible increase in the size of the hole
in the base soil, and only “very slight” erosion of the base soil;
- Sherard and Dunnigan claimed to have tested dispersive and non-dispersive silt and
clay soils, but in fact they only tested two dispersive soils (Foster, 1999, Foster and Fell,
1999a).

Based on these tests, Sherard and Dunnigan (1985, 1989) recommended the following:

1. For all soils with a gravel component (except Group 3 below), the filters should be
designed on the grading of that part of the soil finer than 4.75 mm;
2. Impervious Soil Group 1 (fine silts and clays): For fine silts and clays that have more
than 85% by weight of particles finer than the 0.075 mm sieve, the allowable filter for
design should have D 15F

9D 85B ;