Geoscience Reference
In-Depth Information
dimensionlessforce,N
q
.Thecharacteristicdisplacement,Y
f
,correspondingtomaximum
forceisshownforeachcurvewithanarrowhead.ThetermY
istheratioofthehorizontal
displacement, Y, toD.
Forcomparisonwiththemoistsandtestresults,thefiguresalsoshowforce-displacement
curvesfordrysandobtainedfromthisstudyandtestsbyTrautmannandO'Rourke(1983,
1985). The dry unit weight of the tests by Trautmann and O'Rourke (1983, 1985) was
16
4kNm
−
3
. The dry unit weights obtained during the dry sand tests by Turner (2004)
were 16.7 and 16
.
9kNm
−
3
.
The force-displacement curves for moist sand tests reached a peak at relatively small
displacement,typicallyatY
between0.1and0.2,andthendecreasedtoalowerconstant
value at larger displacements, typically at Y
of 0.2 to 0.3. The maximum dimension-
less force, N
q
, for all moist sand tests and the corresponding dimensionless displace-
ment,Y
f
,wereselectedattheinitialpeakinthecurve.AsshowninFigure17.14,force-
displacement curves for dry sand with similar dry unit weight as the moist sand tests did
notexhibitpeakbehavior.Maximumforcewasselectedforthesetestsusingahorizontal
asymptote to the force-displacement curve, and Y
f
was selected using Hansen's (1963)
90% criterion as described by Fellenius (1980). To compare moist and dry sand test
results at a second dry unit weight for H/D of 6, tests were also performed with dry
unit weights of 15.7 and 15
.
8kNm
−
3
, respectively, asdescribed by Turner (2004).
.
The force-displacement curves shown in Figure 17.14 illustrate several important fea-
tures of soil-pipe interaction. First, the test results for sand with 4% moisture are nearly
identical to the results for sand with 8% moisture, including maximum force, displace-
ment at maximum force, and curve shape. Second, for similar dry unit weight, tests in
moistsandexperiencedabouttwicethemaximumforceassociatedwithtestsindrysand.
Third, displacement at maximum force, Y
f
, was smaller for the moist sand tests com-
pared todrysand testsat thesame density. Moreover, theinitialcurve slope, or stiffness,
isgreaterforthemoistsandtestresults.Also,forthesamedryunitweight,themoistsand
force-displacementcurvesreachapeakvalueanddecrease,typicalofdense,dilativedry
sand, whereas the dry force-displacement curves approach a horizontal asymptote, typi-
cal of loose or medium dense drysand.
Figure 17.15 summarizes values of maximum force vs. dimensionless depth, as deter-
mined from the experimental data. Test results for dry, medium dense sand from Traut-
mann and O'Rourke (1983, 1985) are also shown, and an interpretive curve is drawn
through the moist test results and extrapolated to other H/D ratios. For H/D less than 6,
this extrapolation was performed by multiplying the dry sand test results by the ratio of
moist N
q
to dry N
q
determined at H/D of 6. For H/D greater than 8.5, the dry sand test
data were multiplied by the ratio of moist N
q
to dry N
q
determined at H/D of 8.5. The
interpretive curve between H/D of 6 and 8.5 was drawn as a line connecting the moist
sand data points.
The force associated with partially saturated sand is approximately twice that generated
under dry sand conditions. Direct shear test results show that increased shear resistance
