Professional Documents
Culture Documents
3-7 The Standard Penetration Test (SPT)
3-7 The Standard Penetration Test (SPT)
3-7 The Standard Penetration Test (SPT)
The standard penetration test, developed around 1927, is currently the most popular and eco-
nomical means to obtain subsurface information (both on land and offshore). It is estimated
that 85 to 90 percent of conventional foundation design in North and South America is made
using the SPT. This test is also widely used in other geographic regions. The method has been
standardized as ASTM D 1586 since 1958 with periodic revisions to date. The test consists
of the following:
1. Driving the standard split-barrel sampler of dimensions shown in Fig. 3-5a a distance of
460 mm into the soil at the bottom of the boring.
2. Counting the number of blows to drive the sampler the last two 150 mm distances (total
= 300 mm) to obtain the N number.4
3. Using a 63.5-kg driving mass (or hammer) falling "free" from a height of 760 mm. Several
hammer configurations are shown in Fig. 3-7.
The exposed drill rod is referenced with three chalk marks 150 mm apart, and the guide
rod (see Fig. 3-7) is marked at 760 mm (for manual hammers). The assemblage is then seated
on the soil in the borehole (after cleaning it of loose cuttings). Next the sampler is driven a
distance of 150 mm to seat it on undisturbed soil, with this blow count being recorded (unless
the system mass sinks the sampler so no Af can be counted). The sum of the blow counts for
the next two 150-mm increments is used as the penetration count Af unless the last increment
cannot be completed. In this case the sum of the first two 150-mm penetrations is recorded
as N.
The boring log shows refusal and the test is halted if
When the full test depth cannot be obtained, the boring log will show a ratio as
70/100 or 50/100
indicating that 70 (or 50) blows resulted in a penetration of 100 mm. Excessive equipment
wear, as well as greatly reduced daily drilling meterage, results when blow counts are high.
Standardization of refusal at 100 blows allows all drilling organizations to standardize costs
so that higher blow counts result in a negotiation for a higher cost/length of boring or a
requirement for some type of coring operation.
SPT testing prior to about 1967 (according to ASTM) only required the sampler to be
seated and then driven 300 mm. This stipulation could reduce the N count nearly 50 percent
4
Strictly, the driving distance should be 305 mm since the original test was based on a driving distance of 12 inches.
Owing to the approximate nature of the test it will be adequate to use a distance of 300 mm and divide it into two
150-mm increments. The rationale is that the driving depth is never exact, since one cannot drive using a fractional
hammer drop. If the last hammer drop produces more than 300 mm of penetration, it is still considered 300 mm.
Guide rod
Guide rod
Limiters
Anvil
or drive
head
Guide rod
Drill rod
since the first 150 mm of required seating produces substantial friction resistance on the
sampler for the next 300 mm. It is unfortunate that many current SPT correlations are based
on N values from this earlier procedure.
Both before and after ASTM standardization it was regularly observed that Af values from
adjacent boreholes or from using different equipment in adjacent holes were not reproducible.
Because of wide SPT use, this problem received much attention—first by Gibbs and Holtz
(1957), who considered that overburden pressure and length of drill rod were the principal
causes of nonreproducibility. Beyond this, not much was done until de Mello (1971) presented
a comprehensive literature survey that started a focus on the driving energy [Schmertmann
(1975)].
Discrepancies can arise from factors such as using a warped or worn driving shoe, pushing
a rock (usually detected by an experienced driller), and allowing a quick condition in the hole
bottom resulting from too rapid withdrawal of auger or bit plug or from a differential in water
level between GWT and in hole (or stem of hollow auger). A quick condition is avoided by
attention to the ASTM 1586 standard. The status of the drive shoe can be ensured by regular
inspection, especially after hard driving.
Proper attention to these causes of discrepancies leaves the input driving energy and its
dissipation around the sampler into the surrounding soil as the principal factors for the wide
range in Af values. It should be evident that the blow count would be directly related to the
driving energy, which is theoretically computed as follows:
v = (2gh)l/2 (b)
and substituting Eq. (b) into Eq. (a), we obtain
1W
Ein = ^-(2gh) = Wh (c)
where W = weight or mass of hammer and h = height of fall. This gives, for the standard
63.5 kg hammer and h = 762 mm (30 in.), the theoretical input driving energy of
Em = 63.5 X 9.807 X 0.762 - 474.5 (say, 475 J)
Kovacs and Salomone (1982) found that the actual input driving energy Ea to the sampler
to produce penetration ranged from about 30 to 80 percent; Riggs et al. (1983) obtained energy
inputs ranging from about 70 to 100 percent. These discrepancies appear to arise from factors
such as the following:
1. Equipment from different manufacturers. A large variety of drilling rigs are in current
use; however, the rotary auger of Fig. 3-3a with the safety hammer of Fig. 3-lb is the
most common in North American practice.
2. Drive hammer configurations of Fig. 3-7. The anvil also seems to have some influence on
the amount of energy input to the sampler.
3. Whether
a. the hammer uses an automatic trip with the drop height h controlled to within ±25 mm,
or
b. the system used is a rope-cathead (low-speed power takeoff pulley) (see Fig. 3-8) with
Ea dependent on
(/) diameter and condition of rope
(U) diameter and condition of cathead (rusty, clean, etc., and whether using 125- or
200-mm diameter—200-mm is common in North America)
(///) number of turns of rope around cathead, as 1 \, 2, 3, etc. It appears that a nominal
2 turns is optimum and in wide use. There may be some influence on whether the
rope is pulled from the top (1 \ turns) or from the bottom (2^ turns) of the cathead.
(iv) the actual drop height at which the operator releases the rope to allow the hammer
to "free" fall. Riggs (1986) suggests the operator commonly overlifts an average
of 50 mm (actual drop height = 810 mm). This results from the operator pulling
the rope into the spinning cathead (Fig. 3-8), visually observing the lift to a mark
Figure 3-8 Drilling operator performing the SPT test using a safety hammer with rope-cathead lift. Rope is com-
ing off the bottom of cathead and operator is observing for height mark on hammer guide rod. Helper in foreground
is taking count and observing penetration.
(see Fig. 3-7) on the guide rod, and then releasing the rope back toward the cathead
so it loosens and allows the hammer to fall. Reaction time and mark visibility result
in this overlift. The operator commonly obtains 40 to 50 blows/minute.
4. Whether a liner is used inside the split barrel sampler. Side friction increases the driving
resistance (and AO and is less without the liner (shown in Fig. 3-5a). It is common practice
not to use a liner. Also it would appear that N values should be larger for soils with OCR
> 1 (and larger relative density Dr) than for normally consolidated soils.
5. Overburden pressure. Soils of the same density will give smaller TV values if p'o is smaller
(as near the ground surface). Oversize boreholes on the order of 150 to 200 mm will also
reduce N unless a rotary hollow-stem auger is used with the auger left in close contact
with the soil in the hole bottom. Degree of cementation may also be significant in giving
higher N counts in cemented zones that may have little overburden pressure.
6. Length of drill rod. Above about 10 m the rod length does not seem critical; however,
for shorter lengths and Af < 30 it is. This effect was first examined by Gibbs and Holtz
(1957) and later by McLean et al. (1975) and others [see Sehmertmann (1979)], who used
a computer model to analyze the influence of rod length as well as other factors such as
sampler resistance.
From the several recent studies cited (and their reference lists) it has been suggested that
the SPT be standardized to some energy ratio Er which should be computed as
_ Actual hammer energy to sampler, Ea Ky _ _
t,r = x iuu (a)
Input energy, Em
There are proposals to compute E1n based on the measured hammer velocity at impact with
the anvil or as the measured energy in the drill rod just below the anvil. It would appear,
however, that using the theoretical value given by Eq. (c) for Em would be preferable as it is
not equipment-dependent.
Since there is a wide scatter in Er and the resulting blow count Af when it is reasonable to
expect there should be a unique N for the soil at some depth, it is suggested the drill system-
dependent Er of Eq. (d) be referenced to a standard energy ratio value E^. In this way a drill
rig with, say, Er = 45 would, on adjustment to the standard Er\,, compute approximately the
same N count as from a drill rig with Er = 70. There are several current suggestions for the
value of the standard energy ratio Erb as follows:
Erb Reference
The author will use 70 since the more recent data using current drilling equipment with a
safety or an automatic hammer and with driller attention to ASTM D 1586 details indicate
this is close to the actual energy ratio Er obtained in North American practice. If a different
standard energy ratio Erb is specified, however, it is a trivial exercise to convert to the different
base, as will be shown next.
The standard blow count Nj0 can be computed from the measured Af as follows:
ATy0 = CN X N X Tji x T]2 X 773 X 174 (3-3)
where 77/ = adjustment factors from (and computed as shown) Table 3-3
Nj0 = adjusted Af using the subscript for the Ert, and the ' to indicate it has been
adjusted
Cw = adjustment for effective overburden pressure p'o (kPa) computed [see Liao
and Whitman (1986)] 5 as
r (95J6\in
CN = ——
\ Po )
5
There are a number of overburden corrections for N (this reference lists six); however, this equation plots at very
nearly the average of all those proposed and is the simplest to use.
TABLE 3-3
Factors ^ 1 For Eq. (3-3)*
Hammer for r\i Remarks
Donut Safety
* Data synthesized from Riggs (1986), Skempton (1986), Schmertmann (1978a) and Seed et al. (1985).
11?4 = 1.00 for all diameter hollow-stem augers where SPT is taken through the stem.
Note that larger values of Er decrease the blow count N nearly linearly, that is, Er^ gives
N = 20 and Er9o gives N = 10; however, using the "standard" value of Erjo gives an N
value for use in Eq. (3-3) of N = 13 for either drilling rig. We obtain this by noting that the
energy ratio X blow count should be a constant for any soil, so
En XN1 = Er2 X Af2 (e)
or
1
N2 = ^ X N1 (3-4)
ErI
For the arbitrarily chosen Er\ = 70, this gives, in general,
Nl = 1IxN1
ErI
For the previous example of N2 for E^ = 20 = Er2 we obtain
1. Do nothing, which, with current equipment and conditions, may be nearly correct. This
may have an advantage of detecting increases of soil stiffness (Es) with depth, and upper
variations may indicate cementation or OCR > 1.
2. Adjust only for overburden pressure (all 17; = 1 and CN = some value).
3. Use Eq. (3-3). This is probably the best method but requires equipment calibration for
Er—both equipment and operator. It will also require regular recalibration of the individ-
ual drilling rigs to account for wear and general equipment changes with use. This pro-
cedure will probably become mandatory to extrapolate N data across geographic regions
where different equipment (and Er) is used.
Example 3-2.
Given. N = 20; rod length = 12 m; hole diam. = 150 mm; p'o = 205 kPa; use safety hammer
with Er = 80; dense sand; no liner
Required. What are the "standard" N[ and A^0 based on the following?
/95 76\ 1 / 2 = 0 6 8
Erb = 10 and Erb = 60 CN = ( - ^ l
N2 = #60 = ^ X 16 = 19
////
Example 3-3. Same as Example 3-2 but with sample liner and Er = 60.
CN = 0.68 as before
Vi = % = 0.86 T72 = 1
/95 76\ 1 / 2
CN = = 0
(l00~) '98 <V*™& PO = PC)
"2 = K0 = ^ x 15 = 17
////
3-8 SPTCORRELATIONS
The SPT has been used in correlations for unit weight 7, relative density Dr, angle of inter-
nal friction (/>, and undrained compressive strength qu. It has also been used to estimate the
bearing capacity of foundations (see Chap. 4, Sec. 4-10) and for estimating the stress-strain
modulus Es (see Chap. 5, Table 5-6).
For reasons given in the preceding sections many of these correlations are questionable.
Some are based on a small database or on specific soils. Where a large database was used,
there is the question of what Er was used, this being very critical since many databases were
obtained from published literature that might range from the early 1940s to the present for a
corresponding range of Er on the order of 35 to 80 percent.