See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/248878834

Estimating Settlement of Sand Caused by Construction Vibration

Article in Journal of Geotechnical Engineering · November 1996

DOI: 10.1061/(ASCE)0733-9410(1996)122:11(920)

CITATIONS READS

42 2,813

3 authors, including:

Sergey Drabkin
New York University
15 PUBLICATIONS 188 CITATIONS

SEE PROFILE

All content following this page was uploaded by Sergey Drabkin on 15 October 2017.

The user has requested enhancement of the downloaded file.

 ESTIMATING SETTLEMENT OF SAND CAUSED BY
CONSTRUCTION VIBRATION

By S. Drabkin/ H. Lacy,Z and D. S. Kim3

ABSTRACT: Construction and maintenance of buildings and roads, and normal traffic activities, all generate
vibrations in nearby structures. Multiple investigations showed that in most cases the levels of vibration were
too low to produce direct structural damage based on the vibration limits in existing codes. In current practice,
recommended vibration criteria limits seldom take into account settlement of foundation soils due to repetitive
vibrations, particularly the cumulative effects of long-term repeated exposure. Many different parameters play
an important role in the evaluation of vibrational densification of granular soils. They include vibration char-
acteristics (amplitude, frequency, number of cycles), source of vibration, distance to that source, soil parameters
(attenuation characteristics, grain size distribution, density, moisture content), and state of stresses in soil layers
vulnerable to vibration. This paper describes a method to predict in-situ settlement of sands caused by pile
Downloaded from ascelibrary.org by Sergey Drabkin on 12/11/13. Copyright ASCE. For personal use only; all rights reserved.

driving or vehicular traffic with a laboratory developed polynomial model. Extrapolation from laboratory to in-
situ conditions are derived. Attenuation characteristics of soils are accounted for. Several case histories were
investigated in detail to show practical applications of the proposed method.

INTRODUCTION centage of thickness of soil layer. In his chart, settlement de-

Vibration of soils transmitted to surrounding buildings and pends on soil density expressed in terms of cone penetration
structures is an inevitable by-product of the operation of resistance and ground acceleration. It does not take into ac-
count state of stresses in a vulnerable zone and effect of ac-
construction machinery and traffic vehicles. Vibration is a
complex physical phenomena characterized by amplitude, cumulation of vibration cycles.
frequency, and duration (or number of cycles). Vibration In-situ experimental studies indicated subsurface conditions
amplitude is usually expressed in terms of strain, acceleration, and construction operations that were responsible for vibra-
and peak particle velocity. Throughout this paper vibration tion-induced settlements and estimates of ranges of such set-
amplitude v is defined in terms of peak particle velocity. tlements. Site specific tests using construction equipment is
Correlation between different definitions of amplitude, its now the major tool for prediction of possible settlement.
Laboratory studies of low-level vibration by Kim et al.
levels, sources, and typical related troubles is shown in Fig.
1. When the value of v is less than 2.5 mmls, it can be barely (1994); and Kim and Drabkin (1995) showed how settlement
depends, in a complex manner, on vibration parameters, state
sensed by a human body. Therefore such vibration levels do
not cause any complaints. Vibrations with v larger than 50 of stresses, and properties of sand in drained conditions. These
mmls can be dangerous to structural integrity of sensitive investigations of vibrational densification of granular materials
demonstrated that amplitude of vibration combined with any
structures.
one other characteristic (Youd 1970) is not sufficient for set-
The effects of low-level vibrations with v ranging from 2.5
to 50 mmls is more complex by far. Many case histories (Lacy tlement assessment.
This paper combines the laboratory model developed by the
and Gould 1985; Leathers 1994) demonstrate that operation of
construction equipment generating such vibrations are not only authors and in-situ observations for development of a method
suitable for parametric assessment and prediction of low-level
disturbing for humans, but also cause settlement of structures
vibration-induced settlement.
on sand.
Approximate determination of dynamic settlement is per-
formed now based mainly on vibration amplitude. Barkan POLYNOMIAL MODEL OF VIBRATION-INDUCED
(1962) introduced an idea of critical acceleration (threshold of SETTLEMENT
vibrocompaction) at which and lower than which soil under
given conditions retains its stability. He characterized the stiff- Pile driving causes vibrations with a frequency and ampli-
ness of soils under dynamic action with modulus of defor- tude that has been measured on the ground in the field. A range
mation. This modulus is less than static Young's modulus of similar vibration amplitudes have been applied to triaxial
when vibration acceleration exceeds the critical value. The specimens in the laboratory and the effects measured at in-
practical applicability of modulus of deformation for settle- creasing cycles. Traffic and other vibration sources have been
ment prediction is limited, because it depends on all factors treated similarly.
defining Young's modulus (Hardin and Black 1968) and such A model of vibration-induced settlement for small to inter-
vibration characteristics as amplitude, number of cycles, and mediate vibration levels was developed using multifactorial
frequency. experimental design. Such factors affecting vibration-induced
Massarsch (1992) suggested assessing settlement as per- settlement as vibration amplitude, deviator stress, confining
pressure, soil gradation, number of vibration cycles, relative
'Res. Prof., Polytechnic Univ., Brooklyn, NY 11201. density, and moisture content were varied in different test runs
2Partner, Mueser Rutledge Consulting Engrs., 708 3rd Ave., New York,
NY 10017.
in the ranges shown in Table 1. In each particular run, all input
'Prof., Dept. of Civ. Engrg., KAIST, 373-1, Ksung-Dong, Yusung-Ku, factors were constant. Sand properties are shown in Table 2.
Taejon, Korea. A vibratory frame was designed to shake a soil specimen
Note. Discussion open until April I, 1997. To extend the closing date with a height of 150 rom within a triaxial cell. A program was
one month, a written request must be filed with the ASCE Manager of devised using a design of experiment methodology. This al-
Journals. The manuscript for this paper was submitted for review and lowed the investigation of many factors influencing settlement
possible publication on May 11, 1995. This paper is part of the ]our1Ul1
of Geotechnical Engineering, Vol. 122, No. II, November, 1996. using a relatively small number of experiments performed at
©ASCE, ISSN 0733-9410/96/0011-0920-0928/$4.00 + $.50 per page. drained conditions (Kim et al. 1994; Kim and Drabkin 1995).
Paper No. 10719. For settlement evaluation, the natural values of factors
920 I JOURNAL OF GEOTECHNICAL ENGINEERING I NOVEMBER 1996

J. Geotech. Engrg. 1996.122:920-928.

 VIBRATION AMPLITUDES
IN TERMS OF: o om
, .
1.0 10.0
Strain. % I
0.1 1.0
Acceleration (60 Hz). 8
Peak particle velocity (60 Hz).
in in.lsec. 0.1 1.0 4.0
inmmlsec. I I I
2.5 25 100
SOURCES: Construction Earthquakes
U1tra1011ic operations Blasts
waves Traffic
Pile driving Pile driving
TROUBLES Dynamic Dynamic
NEEDED TO BE ADDRESSED compaction compaction
BY CML ENGINEERS:
None People's Cosmetic Stroctural
complaints due damage to damage
to sensitivity sensitive without
to vibrations structures settlement
Settlement on Stroctural
Downloaded from ascelibrary.org by Sergey Drabkin on 12/11/13. Copyright ASCE. For personal use only; all rights reserved.

sands can cause damage

structural damage to very
sensitive
strocturus

FIG. 1. Correlation between Vibration Amplitude (Expressed In Terms of Strain, Acceleration, and Peak Particle Velocity) and Typical
Civil Engineering Applications

TABLE 1 Tested Ranges of Factors TABLE 2 Properties of Tested Sands

Designa- Maximum dry Minimum dry
tions 0 50 density density
(coded Sand (mm) (kglm 3 ) (kg/m3 )
No, Factors Tested Ranges values) (1 ) (2) (3) (4)
(1 ) (2) (3) (4) Fine 0.5 1,704 1,507
1 Vibration amplitude 2.5-18 mrnIs (v = 0.1- XI Mix 0.7 1,702 1,860
0.7 in.lsec) Coarse 1.7 1,702 1,577
2 Deviatoric stress 14-104 kPa (s = 2-15 X2
psi)
3 Confining pressure 69-207 kPa (p = lO- X, TABLE 3. Coding of Variable Factors
30 psi)
4 Sand mixture Coarse, medium, fine x. No. Factors Coding of factors
5 Number of vibration cy- iN = 60-500,000 X, (1 ) (2) (3)
cles
6 Moisture content Dry, saturated 1 Vibration amplitude Xl = -1 + (v - 0.1)/0.3
X.
7 Initial relative density Loose, medium dense X,
2 Deviatoric stress X2 = -1 + (s - 2)/6,5
3 Confining pressure X, = -1 + (p - 10)/10
4 Coarse sand x. =-1
4 Fine sand x. = 1
should be replaced by their coded values Xi and substituted in 5 Number of vibration cycles X, = -1 + (N - 60)/26,997
the polynomial equation 6 Dry sand X. = -1
In Y = 2,27 + 1.19xI - O,71xi + 0.49X2 - O,68x~ - O.80X3
6 Saturated sand X. = 2
7 Initially loose sand x, = -1
7 Initially medium dense sand x, = 2
+ 1,09x~ - 0.46x. + O.06x; + 0.45x, - O.38x;
- O.19x6 - O.lOx? (1)
specimen's height h is equal to its maximum value ho, which
where Y is settlement expressed in 0,0254 mm (0,001 in.), The corresponds to density Po of granular material. After applica-
coding is performed considering ranges of quantitative factors tion of confining pressure p, deviator stress s (defined as a
in Table 1 as varying from -1 to 1. The coding formulas and difference between vertical axial stress and confining pres-
coded values of qualitative factors are given in Table 3. When sure), and vibration, h become smaller for loose and medium
the observed values of quantitative factors exceed the ranges dense sands. The minimal height of a specimen Ii, exists that
specified in Table 1, they should be coded as 1. corresponds to the maximum possible density p, Without fric-
Eq. (1) does not account directly for short-term increase of tion between particles, Po would be equal to p, but because of
pore water pressure (PWP) that accompanies pile driving in friction forces the infinite number of metastable states exists
vulnerable zones below the water table (Leathers 1994). An with densities p < p (or height h > Ii),
increase of PWP influenced the settlement of saturated speci- Vibration applied to the specimen causes separation between
mens. Though the tests were performed with an open drainage some particles, decrease of friction, and breakage of metasta-
valve, the cross-sectional area of drainage pipe was much ble state. The resulting settlement is a change of h with time,
smaller than that of the specimen (8 and 3,800 mm2 , respec- The value of h determines how close the current state of the
tively). Therefore, the dissipation of PWP in saturated speci- specimen is to the equilibrium state Ii,
mens had similarities with in-situ behavior of saturated sands, The settlement h as a function h = h(p, s, v, t), where tis
The response surface analysis of (1) given by Kim and time of vibration, was theoretically derived by Boguslavskii
Drabkin (1995) illustrated many aspects of vibration-induced et al. (1995) based on energy barrier criteria, They showed
settlement. The polynomial model in (1) provides reasonably that the work of external forces was the driving mechanism
close assessment of settlement, but it does not explain the na- of settlement. Vibration energy characteristic is proportional to
ture of vibrational densification of sands, v 2 and has the similar effect upon sand particles as temperature
Before application of external stresses and vibration, the upon atoms in the theory of rate processes (Glasstone and
JOURNAL OF GEOTECHNICAL ENGINEERING / NOVEMBER 1996/921

J. Geotech. Engrg. 1996.122:920-928.

 Laidler 1941). Therefore the specific direction of vibration is
not significant for densification of sands. Hence vibration char- (4)
acteristics determined in experiments with a vibration table are
applicable for evaluation of in-situ settlements caused by con- Woods and Jedele also recommended values of a for dif-
struction vibrations. ferent soils atf= 5 and 50 Hz. These values resulted from the
averaging of large amounts of data at different sites, hence
EXTRAPOLATION METHODS FOR ASSESSMENT OF they may differ from that existing at the investigated site.
IN-SITU SETTLEMENT WITH POLYNOMIAL MODEL When variations of v with r are known they can be used to
determine 0: for different frequencies of vibrations generated
One- and lO-layer extrapolation methods were used to es-
by operating equipment with the equation
timate the settlement of 3-40 m of sand by applying the model
developed for 150 mm specimen. The one-layer method con- In[( 'l!2lvI)(rzlrl)"]
siders the vulnerable zone as one layer with the influencing 0:= (5)
factors assessed in the middle of the layer. Then it was as-
sumed that settlement A of a layer with thickness h is directly Another method of assessment of 0: was suggested by Haupt
proportional to the settlement y of the tested specimen (1986)
Downloaded from ascelibrary.org by Sergey Drabkin on 12/11/13. Copyright ASCE. For personal use only; all rights reserved.

27rDf
A=L
150
H (2) 0:=-- (6)
t C
where all values are in millimeters. where D = material damping in percent, and C = wave-prop-
The lO-layer extrapolation method considered the vulnera- agation velocity. For the elastic range of soil deformations, D
ble zone divided into 10 equal layers. Each layer was evalu- is about 5%. The propagation velocity of Rayleigh waves in
ated for the conditions existing in its middle, and the settle- dry sand varies from 100 to 300 mls and in saturated sands
ment of each layer was summed. This method accounted for from 300 to 700 mls (Massarsch 1992).
variations of vibration amplitudes and increased in-situ stress Calculated with (6), 0: = 0.0031 mm- 1 for saturated sand at
with depth. frequency 50 Hz and C = 500 mls. Tabulated by Woods and
Both methods replace the settlement of three-dimensional Yedele (1985), 0: = 0.00031 mm- I • The values of a differ 10
specimens with the settlement of a one-dimensional vulnerable times. Therefore, direct evaluation of 0: from field tests is rec-
zone. This is valid only if all sand in this zone is vibrated. In ommended.
the case of pile driving, settlement large enough to damage An example of three typical vibration environments where
structures is usually limited to the horizontal distances equal evaluation of vibration-induced settlement might be performed
to the depth piles being driven below the founding level of is shown in Fig. 2: surface traffic vibration (highway), in-depth
the structure (Dowding 1991). Settlement caused by operation traffic vibration (subway), and in-depth construction vibration
of such construction machinery as pavement beakers, curb (pile driving).
breaking loaders, and jack hammers should be small in mag- To calculate vibration-induced settlement using the predic-
nitude and areal extent as they do not vibrate large volumes tion model, seven influencing factors need to be evaluated. In
of soil. The same consideration limits the validity of test pile the current practice, vibration amplitude is monitored on the
driving for direct prediction of settlement. Potentially danger- ground surface next to the adjacent structure of concern. The
ous sources of vibration are driven piles, dynamic compaction, one-layer method considered all 20 m of sand as one vulner-
and surface (highways) and subsurface (subway) traffic. able layer with constant values of factors corresponding to the
Vibrations lose energy during their propagation through the middle of the layer. Vibration amplitude (factor 1) was the
ground. The decay of amplitude of vibrations with distance same as on the surface (5 mmls). The total number of vibration
can be attributed to geometrical and material damping, which cycles from each source (factor 5) was 500,000.
may be described by the equation (Woods and Jedele 1985) The 1O-layer method considered a vulnerable zone consist-
(3) ing of 10 layers with the thickness of 2 m each (Fig. 3). Var-
iations in vibration amplitude with depth along the vulnerable
where VI and 'U2 = vibration amplitudes at distances rl and r2 zone was determined for 0: = 10- 5 mm- 1 using (3). The effects
from a source of vibration; n = a coefficient depending on type from pile driving was approximated by a set of discrete
of propagated wave (n = 2 for body waves along the surface; sources located at each layer and generating vibrations with
n = 1 for body waves in the ground; n = 0.5 for Rayleigh an amplitude of 5 mmls. The number of vibration cycles from
waves); and 0: = a coefficient of attenuation (units of l/dis- each such source was increased proportionally to the increase
tance), which is affected by the material damping of soils and of confining pressure. Vibration amplitude from the driven pile
the vibration frequency f
The frequency range of usual construction vibration sources Peak particle velocity
5 mmlsec
is 5-25 Hz. Assuming a wave velocity in soil of about 500 Elevalion, m IS m HIGHWAY
mls or less, the minimal wave length is about 30 m. Low-level o;--------k==Jl9===:!:~===i~L----I-
vibrations usually affect soil at distances smaller than the wave Fill Y= 1700kglm 3
length or in the near zone of the source and the earth's surface. 4

Hence, Rayleigh waves are mostly responsible for transfer of Sand: fine, ~ SUBWAY
medium dense
energy from a vibration source. Therefore, n is assumed to be y= 1750 kglm 3 PILE Driving
0.5 in the following calculations.
12 Vulnerable
Youd (1970) showed that frequency variations did not in-
fluence densification of granular materials when an entire spec- 16
imen is vibrated with the same amplitude. Therefore, (3) does
not have a frequency term. But the vibrational wave distri- 20
bution in the ground is frequency dependent, and 0: takes care
of that dependence. When 0:1 is known for fl> an unknown 0:2 24
for h. can be computed using the following formula by Woods FIG. 2. Typical Urban Environments Considered in Parametric
and Jedele (1985): Study

922/ JOURNAL OF GEOTECHNICAL ENGINEERING / NOVEMBER 1996

J. Geotech. Engrg. 1996.122:920-928.

 -I

--- -------
0
-2
~
.... //
.... -3
// / -v-
Total settlement
(mm):

--
-4 Highway --24.7 -
-s fIiahway

s ..!II -5
// / -<>-
Subwoy
JJ / Subway 29.7
'-12
Pile driviaa
]
'"
-6
-7
(J / ---
Pile drivin&.--594 _
·16 .....0.‫סס‬OO I pmfl
..-0.5 -8 I
·20 -9 I
-10
L
·24 o 10 IS 20
2 3 4 6 Settlement, mm
Vibnliall omplillldo, IIIIII!-
FIG. 4. Variations In Vibration-Induced Settlement with Depth
FIG. 3. Variations In Vibration Amplitudes with Depth tor Dlt- (Layers) for Different Vibration Sources
Downloaded from ascelibrary.org by Sergey Drabkin on 12/11/13. Copyright ASCE. For personal use only; all rights reserved.

ferent Vibration Source.

EVALUATION OF CASE HISTORIES OF
VIBRATION·INDUCED SETTLEMENT
decreased on the surface with increase in penetration depth. It Back Bay Site In Boston (Massachusetts)
decreased with depth from highway and increased with depth
towards the subway. The densification of sand layers during pile driving was re-
The effect of traffic on vibration amplitude was studied by ported by Leathers (1994). The project site was located in the
many researchers (e.g., Sutherland 1950; Barneich 1985; and middle of a block of existing buildings in the Back Bay section
Long 1993). Peak particle velocities measured at 3-6 m from of Boston. The site was about 33 by 67 m in plan. Two ex-
the vehicle depend on weight (type) of vehicle, its velocity, isting buildings were located immediately adjacent to the pile
and road surface roughness (defined as a ratio of height of driving area.
road bulges to the square root of distance between them). A Subsurface conditions were explored by 10 borings drilled
bus or truck with a total weight of 10 t moving at 65 kmIh on prior to pile driving. The primary soil strata were granular fill,
a typical urban road with a roughness 0.2 will cause surface organic silt, silty clay with zones of clayey sand and gravelly
vibrations of 8 mmls at the distance of 3 m. Train traffic vi- sand, and glacial till. The thickness of lower sand stratum was
bration was about 6.3 mmls at 15 m from the track. A subway about 24 m. A generalized subsurface profile is shown in
generated an amplitude of about 5 mmls at the vertical distance Fig.5.
of about 6 m from the tunnel (measured on the structure). A total of about 180 precast 360 X 360 mm concrete piles
Variations of in-situ stress conditions depend mostly on with lengths ranging from 29 to 39 m were driven for this
depth. For the center of the vulnerable layer, a deviator stress project. The ICE 640 diesel hammer had a rated energy of 54
kN-m.
s (factor 2) was determined by subtracting a mean effective
The peak particle velocity of vibration was measured at two
stress p from a vertical effective stress CTu;, where p = (1 +
locations on the exposed foundation of the adjacent building.
2Ko)/3*CTu; was determined as factor 3 (confining pressure) by
Maximum vibration amplitudes ranged from 6.4 to 15 mm1s
assuming an earth pressure coefficient at rest Ko 0.5. All = (Leathers, personal communications, 1994). The decrease of
other factors were considered to be constant: factor 4 was vibration amplitude with increasing depth of pile penetration
equal to 1 (fine sand); factor 6 was 2 (moist sand); and factor is shown in Fig. 6. Calculated with (5), Q: was -4.26 X 10- 5
7 was 2 (dense sand). mm-\ which was smaller than Q: = 10-5 mm- I recommended
In-situ observations of low-level vibration-induced settle- by Woods and Jedele (1985).
ment is a complicated task. The number of cycles required to Settlements observed after pile driving ranged from 18 to
observe the full effect of vibration is very large. 500,000 cy- 54 mm at different site locations. Settlements measured at the
cles of vibration from a subway, and especially a highway, ground surface and the top of the sand stratum were nearly
may accumulate for 5-25 yr to cause settlement of 25 mID or identical. Settlement measurements were performed during
more, which may damage a structure. Settlement is not usually and after driving of several piles. The settlement occurred dur-
monitored for such a long period of time and the related dam- ing driving, but did not continue once driving stopped.
age is associated with other reasons (aging of structure and so Direct in-situ measurements of vibration induced sand den-
on). Only pile driving presented examples of documented case sification using the SPT method were unsuccessful. The SPT
histories of settlement, because required number of cycles is blow count from the preconstruction borings ranged from 12
accumulated quickly, and settlement monitoring lasts for to 44 blows per 300 mm. They were the same in a few borings
weeks, not decades. investigated after the pile driving. These observations could
Variations in vibration-induced settlement with depth (Fig. be foreseen because small changes of density that accompany
4) are source dependent; pile driving can result in settlement low-level vibration should not be discernible with such a
that is twice as large as from surface traffic vibration. The method as SPT.
input of individual upper layers to settlement can be bigger The expected settlement was evaluated using the prediction
than that of lower layers even though vibration amplitude reg- model and one- and lO-layers extrapolation schemes. The nat-
istered in upper layers may be smaller than in deeper layers. ural and coded values of factors are shown in Fig. 5. The good
Relatively large settlement of upper layers is explained by matching of observed and calculated settlement confirmed
smaller confining pressures at the corresponding depths. Leathers's hypothesis that observed settlement was due to den-
For proper settlement assessment, vibration amplitude sification of the sand stratum during pile driving.
should be monitored not only at the ground surface but also
within the ground. Otherwise, vibration amplitude should be Southern Brooklyn Site (New York City)
evaluated considering attenuation characteristics of soil as well Differential settlement of aeration tanks was described by
as vibration paths. Lacy and Gould (1985) and Lacy et al. (1994). The 5 m high,
JOURNAL OF GEOTECHNICAL ENGINEERING / NOVEMBER 1996/823

J. Geotech. Engrg. 1996.122:920-928.

 80 m wide tanks were supported by timber piles as shown in factors are shown in Fig. 7. Calculated settlement of 58.9-
Fig. 7. Closed-end 10.75 in. (273.1 mm) pipe piles were driven 74.3 mm matched well the observed settlement.
to depths exceeding 40 m in close proximity to the tanks to Vibrations were monitored with a seismograph on the sur-
support the new structure. Pile driving in medium dense, fine face of both structure and the adjacent ground. Peak particle
to coarse sand was performed with Vulcan 08 impact hammer. velocities on the structure were always less than 2.5 mmls. On
Settlement was noticed when about 100 new piles had been the ground the vibration amplitude ranged from 2.5 to 23 mml
installed. When settlement exceeded 25 mm, all fluid was re- s. Estimates of at from ground and surface measurements
moved from the tanks, reducing loading by about one half. ranged from 0.02 to 0.05 mm- I • These values indicate atten-
Nevertheless, the settlement continued and reached 70 mm uation in organic soils forming the upper strata on site. The
during the driving of additional piles. Augered cast-in-place vulnerable zone was formed by medium dense fine sand, and
piles (ACIP) replaced the remaining piles to be driven, which the Vulcan 08 impact hammer generated frequencies of about
were generally those closest to the existing structures. 20 Hz. Hence an assumed value of at = 10- 5 mm- I was used
The possible settlement was assessed using one- and 10- in 1O-layer extrapolation method.
layer extrapolation schemes. The natural and coded values of The effectiveness of tank emptying for vibrational settle-
ment prevention was evaluated. The state of stresses was de-
Elevation, m termined in the middle of the vulnerable zone at the depth of
Downloaded from ascelibrary.org by Sergey Drabkin on 12/11/13. Copyright ASCE. For personal use only; all rights reserved.

Vibr.AmpL Meas.Sett. CalcSettl'Calc.SettL"

Fill
(nunlsec) (rnm) (rnm) (rnm) 25 m. The surcharge load on the ground surface was about
-4 6.4-15.0 18-54 23-78 28-85 100 kPa. This was the pressure of 5 m of water and the tank's
Silty
clay
weight. At the depth of 25 m, the influence of that load re-
Factor Natural Coded
-12 Value Value duced to 20% of its value, or 20 kPa according to the chart
6.4150 -0.5..0.64 provided by the U.S. Navy in 1971 as cited by Holtz and
Medium
dense 65 kPa 0.14 Kovacs (1981). The vertical stress at the depth was about 400
130 kPa -012
sand and
Vulnerable Coarse -I
kPa. Thus, emptying of the tank provided less than 5% of
gravelly
sand
zone 500000 cycles 0.85 pressure decrease. Correspondingly, no effect on vibration-in-
Moist 2
duced settlement was observed, This estimation would be im-
Dense 2
possible without insights in mechanisms governing vibrational
Range o f t
-36
--.i.
pile tips
·Sett1ement was calculated for in-situ
stress conditions in the middle of the
densification of sands.
Rock vulnerable zone
"Settlement was calculated for the
vulnerable ZODe divided into 10 layers
Cedar Creek Site In Wantagh (New York)
FIG. 5. Case History of Deformations In Sand Layer during Evaluation of vibration effects on settlement was performed
Pile Driving at Back Bay Section of Boston by Lacy (unpublished report, 1986) to permit the use of a
vibratory hammer to drive sheeting in close proximity to ex-
6 r------------------., isting structures, The study consisted of pile-hammer testing
and monitoring of vibration-induced settlement during pile
driving at the Cedar Creek site.
Several effluent tanks were planned to be constructed close
to the existing 2.4 m diameter effluent and 2.1 m diameter
outfall pipes as shown in Fig. 8. Subsurface conditions are
shown in Fig. 9 as interpreted from borings made in the vi-
cinity of the pipes. The upper 4-5 m layer was hydraulically
placed lose to very compact gray and brown medium sand
with a trace of silt and gravel. The sand's density varied sub-
stantially from boring to boring. Next followed 1-2 m of or-
2L------'-----'----'-------'-----' ganic silty clay, peat, and medium dense sand. The pipes were
OSlO IS 20 2S founded below this layer. Under the pipes to the elevation -12
Dcpth,m m is medium compact gray fine to coarse sand with a trace of
FIG. 6. Change of Vibration AmplitUde Registered at Surface silt and gravel. This layer was vulnerable to vibration densi-
with Increasing Depth of Pile Penetration fication.

Vibr.Ampl. Meas.Sett. Calc.SettI.' Calc.SettI·,

(mm/sec) (rom) (rom) (rom)
Aeration tank
lOOkPa
Elevation, m 17.5 70 49 61
o r-------t-ri~~:....r- Factor Natural Coded
Fill Value Value
--SZ-
Organic
-10
Existing wooden I 17.5 mm/sec 0.96
Sand, fine to
coarse,
piles 2 154 kPa 1
-20 3 306 kPa I
medium
4 Fine I
dense Vulnerable
5 500000 cycles o. &5
trace of woe
- 30 6 Moist 2
gravel
7 Dense 2
40 New pipe
'Settlement was calculated for in-situ
piles
stress conditions m the middle of the
-50 vulnerable zone
"Settlement was calculated for the
vulnerable zone divided into 10 layers

FIG. 7. Case History at South Brooklyn Site, New York

924/ JOURNAL OF GEOTECHNICAL ENGINEERING / NOVEMBER 1996

J. Geotech. Engrg. 1996.122:920-928.

 Pile-Hammer Testing given range of depths, the observed amplitudes were approx-
A pile-hammer test program was developed to evaluate the imately equal, thus at the distances up to 10 m they were
vibration amplitudes from the ICE 812 vibratory hammer and considered uniform in the settlement prediction model.
the Vulcan 65C impact hammer at different frequencies of vi- The subsurface amplitudes' measurements for the same dis-
bration, at different distances, and depths. Vibrations were tances varied respectively from 0.6 to 36.5 and from 2.9 to 10
monitored at the ground surface using three component seis- mm/s.
mometers connected by shielded cables to a seven channel Vibrations produced by an air-steam impact hammer Vulcan
permanent recording oscillograph. Only two locations were 65C were monitored while the test sheeting was driven to a 9
monitored simultaneously because three oscillograph's chan- m depth at the location shown in Fig. 8. The hammer operated
nels were needed for one location. The surface measurements at 60 blows per min. The driving time was about 12 min. The
were taken at distances from 3 to 18 m from driving. Subsur- hammer was operating at only about 50% of its rated speed
face monitoring was performed inside a 76 rom diameter 8.5 and a much reduced hammer energy. The positions of trans-
m long casing. The geophone was side clamped at the level ducers on the surface and in the ground remained the same as
of -7.6 m. The casing was placed in the borehole positioned for the vibratory hammer test. The observed v ranged from
as shown in Fig. 8. 0.6 to 1.6 mmJs at all distances to the surface and subsurface
Testing of the effects of vibration was performed in several transducers. No settlement was observed during test driving.
Downloaded from ascelibrary.org by Sergey Drabkin on 12/11/13. Copyright ASCE. For personal use only; all rights reserved.

stages. At first, a single sheet was driven to the geophone level The testing program on the Cedar Creek site is shown in
with. the vibratory hammer continuously at a frequency of ap- detail, because it demonstrated both positive and negative as-
proxImately 23 Hz. The total time of driving, including ham- pects of intuitive experimentation. The program succeeded in
mer start-up, was about 53 s. Measurements were taken with its main purpose to define the general range of possible vibra-
the subsurface geophone and with the seismometers on the tions generated by different vibration hammers at the given
surface 3.1 m from the sheeting. Afterwards, vibrations were site. Therefore, the necessary recommendations concerning
monitored while a sheet was vibrated with frequencies ranging pile driVing could be made.
from 15 to 25 Hz at fixed depths of 3.1, 4.6, 6.1, and 7.6 m. Subsurface measurements did not adequately determine am-
The range of amplitudes on the surface at a distance of 3.1 plitudes existing in the vulnerable zone. Their low values may
m from the hammer varied from 1 to 59 mm/s. At a distance be explained by the positioning of the geophone inside the
of 6.1 m, they varied from 1.2 to 27 mm/s. Much higher vi- casing. Other case histories (see for example South Brooklyn
bration amplitudes (25-59 mm/s) were observed at 15 and 18 site) demonstrate that transducers located on the structures
Hz than at higher frequencies (1-12 mm1s). During installa- show smaller vibration amplitudes than transducers placed on
tion of the sheet piling, the hammer was operated at the higher the ground surface. Both surface and subsurface measurements
frequencies. As a result, no direct influence of vibration fre- allowed the authors to derive a = 0.0008 mm - I characterizing
quencies was observed, except dUring the start-ups and shut- the strata outside of vulnerable zone. The depth of the geo-
downs of the hammer, when much larger vibration amplitudes phone's installation was not sufficient to characterize attenu-
were generated compared to normal working conditions. In the ation in the vulnerable zone.

50m
Existing 2.1m Evaluation of Settlement Using Polynomial Model
outfall pipe

Sheetp,les Based on the test driving data, settlement was evaluated

~O
PZ 22
using a one-layer extrapolating method. Calculations with the
1O-layer extrapolation method were unnecessary at this site
because of the small thickness of vulnerable zone.
Settlements were evaluated for vibration amplitudes of 5
mmls and 10 mm/s. The vulnerable 7 m zone was below a 5

00
® Surface
transducers
A. Subsurface
geophone's
location
+ Sheet pilesPZ22
dr,ven for test program
m sand fill layer, hence confining pressure in the middle of
the zone was 92 kPa and the deviator stress was 46 kPa for
the at-rest earth pressure coefficient condition. The levels of
all factors are shown in Fig, 9. The predicted settlement due
to driving of a small number of test piles was about 2 mm
and could be within the error margins of settlement observa-
FIG. 8. Settlement Monitoring Location Plan at Cedar Creek tions. The calculated settlement after driving of sheeting along
Site the pipeline ranged from 9 to 15 mm. Therefore, the polyno-

-
Settlement Vibr.Ampl. Meas.Sett. Calc.Settl-
Elevation, m
monitoring (rnmlsec) (mm) (rnm)
yO
Fill: hydraulically 5-10 13-19 4-12
placed medium sand, Factor Natural Coded
-2
density varied from
Value Value
loose to dense in
-4 different borings I 5-10 rnmIsec -067.002
Organic silty clay, Pipe 2 46 kPa -0.28
peat, sand 3 92 kPa -0.67

t
-6 Sand: 4 Fine I
medium, loose to 5 500000 cycles 0.85
medium dense, 6 Moist 2
-8 Vulnerable
density 164 glcm 3 7 Dense -I
zone
saturated
-10

-12 1 -Settlement was cakulated for in-situ

stress conditions in the middle of the
vulnerable zone

FIG. 9. Cedar Creek Site

JOURNAL OF GEOTECHNICAL ENGINEERING / NOVEMBER 1996/925

J. Geotech. Engrg. 1996.122:920-928.

 Vibr.Ampl. Meas.Sett. Calc.Settl.· Calc.settl.··
PILE DRIIVING (nun/sec) (mm) (mm) (mm)

Elevation, m 17.5 250 80-100 100-120

Concrete Factor Natural Coded

pier Value Value
·20
I 17.5 nun/sec 0.96
2 188 kPa I
zone
-40 3 374kPa I
of fine silty
4 Fme I
sand
5 500000 cycles 0.85
-60

-80
1 6
7
Moist
Loose
2
-I

·Settlement was calculated for in-situ

stress conditions in the middle ofthe
vulnerable zone
··Settlement was calculated for the
vulnerable zone divided into 10 layers
Downloaded from ascelibrary.org by Sergey Drabkin on 12/11/13. Copyright ASCE. For personal use only; all rights reserved.

FIG. 10. Case History of Pier Settlement in Lesaka

~
mial model was able to predict settlement better than test pile
driving.
I
Building 2-story Historic

building ~m building
Lesaka Site (Northern Spain)
1.5-15 m

--=~ile
Picornell and del Monte (1985) discussed a case of pile-
driving-induced settlement of a pier foundation. The building Seismograph 2 driving
of a steel mill factory was supported by cast-in-place concrete locations woe at
Tri-boca tower
1.08 m diameter piers embedded to a depth of 20 m into sandy construction site
soil. The vertical pressure on each pier did not exceed 534
kPa. New equipment foundations had been designed on steel ~
H-piles to be driven to bedrock. Upon driving steel H-piles,
FIG. 11. Plan of Trl-Beea Site in Manhattan
one of the pier foundations settled 250 mm.
The investigation of this settlement included field plate-load
tests, pier-load tests, and laboratory consolidation tests. They Tri·Beca Site in Manhattan (New York City)
showed that static design loads would cause only minimal set-
tlement (less than 9 mm). At the time the settlement was no- The settlement that accompanied pile driving at the Tri-beca
ticed, a visual inspection did not reveal the presence of cracks site in Manhattan was described by Lacy et al. (1994). A 52
or any other feature on the ground surface that would suggest story residential building, Tri-beca Tower, was planned in
some type of cave-in effect. Using an analogy between this close proximity to two other buildings (Fig. 11). The original
settlement and the settlements observed at the Embarcadero foundation project considered installation of 178 mm outside
site (Clough and Chameau 1980) and at the Leningrad site diameter open-ended 30 m long pipe piles.
(Dalmatov et al. 1968), dynamic compaction induced by the The subsurface conditions are shown in Fig. 12. Medium
pile driving on the sand layer was considered to be the only compact fine to medium sand was expected to densify due to
cause of settlement. Analysis of the vibration-induced settle- vibrations. Test piles were driven to evaluate expected settle-
ment using the polynomial settlement model raises doubt as ment of an adjacent two story building and a historic landmark
to that conclusion. six story building. Extrapolation from field measurements of
In their investigation, Picornell and del Monte provided nei- ground settlement predicted 25-50 mm settlement for the two
ther observed vibration amplitudes and duration of pile driving story building and 13-25 mm settlement for historic building.
nor the type of driving equipment. A maximum v = 17.5 mm/ The values of vibration amplitudes during test pile driving are
S was assumed. The mean effective stress (confining pressure) now not available. They were limited by 25.4 mm/s to prevent
in the middle of vulnerable zone at a depth of 35 m was 562 direct damage to the structures from vibrations.
kPa assuming an earth pressure coefficient at rest as 0.5. The Two seismographs were used to monitor vibrations from
deviator stress was 188 kPa. The stresses exceed those used construction activities. Locations of monitoring stations are
for the development of the polynomial model. Thus, the com- shown in Fig. 11. Seismograph 1 station was on the floor at
bination of maximum tested level of confining pressure and a subbasement level of the historic building. The distance from
deviator stress was chosen to get the possible upper bound of the driven piles was 30-40 m. During the study period, all
settlement. The number of vibration cycles were also maxi- peak particle velocities measured were below 2.5 mm/s. Seis-
mum. The natural and coded values of factors are shown in mograph 2 locations were changed within the indicated area.
Fig. 10. Its distance from the driven pile was 1.5-15 m. The vibration
The calculated settlement using both computational methods amplitudes registered by this seismograph varied from 2.5 to
is about 120 mm, which is two times smaller than was ob- 18 mm/s.
served at the site. Vibration-induced densification could not be The observed settlement of the two story building was 38-
totally responsible for settlement of this magnitude. Still, pile 69 mm at different stages of construction. There was no set-
driving was definitely responsible for settlement. The expla- tlement of the historic building.
nation is probably the combined effect of vibration on both Settlement of both buildings was evaluated using the poly-
soil densification and the reduction of skin friction on the nomial prediction model using one-layer and 1O-layer extrap-
piers. These piers were not supported by bedrock, so settle- olation methods assuming ex = 10- 5 mm- I • The predicted set-
ment of surrounding soil could cause pier settlement. The tlement for the historic building was less than 10 mm and that
quantitative effect of vibration on stability of friction piles in matched the observations. For the two story building, the val-
sands was not properly investigated. ues of factors used in both methods, predicted, and observed
926/ JOURNAL OF GEOTECHNICAL ENGINEERING / NOVEMBER 1996

J. Geotech. Engrg. 1996.122:920-928.

 Vibr.Ampl. Meas.Sett. Calc.Settl· Calc.settl··
(mmlsec) (rom) (rom) (rom)
5-13 38-69 12-49 18-65

Factor Natural Coded

Value Value
1 5.. 13 -0.67 ..0.37
2 53 kPa -0.13
3 106 kPa -0.46
4 Fine 1
5 500000 cycles 0.85
6 Moist 2
7 Dense 2
'Settlement was calculated for in-situ
stress conditions in the middle of the
vulnerable zone
"Settlement was calculated using
1O-layers extrapolation scheme
Downloaded from ascelibrary.org by Sergey Drabkin on 12/11/13. Copyright ASCE. For personal use only; all rights reserved.

FIG. 12. Case History at li'1·Beca Site In Manhattan

TABLE 4. Case Histories of Vibration-Induced Settlement settlement prediction model takes into consideration all these
SETILEMENT (mm) factors. Site investigations should provide sufficient informa-
tion to determine these factors for proper assessment of settle-
Calculated
Vibration ment.
amplitude Observed One-layer 10-layer To extrapolate the test results obtained from a 150 mm spec-
Investigated sites (mm/s) in-situ scheme scheme
imen to in-situ settlement, two extrapOlation methods were
(1 ) (2) (3) (4) (5)
tested. Settlement was calculated using both methods and the
Back Bay (Leathers 1994) 6.4-15 18-54 18-63 24-78
Brooklyn, South (Lacy et aI.
results matched well with settlement observed at the sites.
1994) 17.5 70 59 74 Studied case histories showed that usually the use of
Brooklyn, West (Lacy et al. the one-layer method was sufficient for estimation of settle-
1985) 2.5-15.2 61 5-56 8-63 ment. The lO-layer extrapolation showed the effect of differ-
Cedar Creek (Lacy 1986) 5-10 13-19 9-15 - ent layers on settlement and provided the upper limit of ex-
Embarcadero (Clough and Chao
meau 1980)
Leningrad (Dalrnalov el aI.
1-5 8-51 8-58 - pected settlement. That approach may be recommended for
evaluation of effects of pile driving in highly nonhomogeneous
1967) 2.8 6-11 4.5 8.1
Lesaka (Picomell and del soils.
Monte 1985) 17.5 250 III 117 The maximum densification happens when high deviatoric
Northbrook pipeline (Linehan stress is combined with low confining pressure. In case of high
1992) 2.8 38 37 48
Tri-beca (Lacy el aI. 1994) 2.5-18 38-69 15-112 27-135 confining pressure, the model will be valid approximately up
to double values of stresses observed in the field as compared
to laboratory tested. We do not recommend use of the model
settlements are shown in Fig. 12. They either match or exceed beyond that range.
the observed settlements. The range of observed and calculated settlements caused by
Settlement of the two story building would have been larger, soil densification was between 5 and 135 mm. With increasing
but pile driving was stopped after settlement reached the above confinement or initial density, settlement is substantially re-
values. Underpinning jacks were installed. The driven piles duced. An increase of stress anisotropy causes increase in set-
were replaced by ACIP piles. They were installed by drilling tlement. At the Lesaka site, 250 rom settlement was caused
holes with a continuous flight hollow-stem auger. Then the most probably by the combined effect of soil densification and
auger was slowly withdrawn while continuously pumping partial loss of support by friction piles due to reduction in
grout to prevent collapse of the hole. The completed grout friction forces caused by vibration.
column formed a cast-in-place pile. To reduce settlement due Test pile driving is needed for providing vibration ampli-
to removal of soil, care was taken to remove as small amount tudes generated by equipment and vibration attenuation in the
of soil as it was possible. The installation of ACIP piles also ground at a given site. The vibration amplitudes measured in
caused some settlement, but its value was smaller than would soil were much higher than measured on structures.
be anticipated for pile driving, considering that the ACIP piles The subsurface measurement of vibration is complicated,
were generally closer to the building than the driven piles. and may produce unreasonably low values when done in cas-
ings. The measurements should be performed inside of the
Other Sites vulnerable zone. When in-situ data are not available, values
of a in Woods and Yedele (1985) provide reasonable estimates
Drabkin (1995) analyzed sites in West Brooklyn (Lacy of vibration attenuation of soils.
et aI. 1985), Embarcadero (Clough and Chameau 1980), Len- Extrapolation of settlement observed in pile-hammer tests
ingrad (Dalmatov et at 1967), and Northbrook (Linehan may give inaccurate settlement predictions because the influ-
1992). The results are summarized in Table 4, and show good ence of several factors is neglected. The total number of cycles
agreement between calculated and observed settlements. during test pile driving is much smaller than during construc-
tion. The volume of sand involved in densification is small;
SUMMARY AND CONCLUSIONS thus, the observed settlement is also small. Therefore, the use
of a polynomial prediction model is preferable.
Vibration-induced settlement on granular soils is affected by The effect of vibration on the stability of friction piles em·
a combination of various factors such as vibration character- bedded in sands without reaching bedrock or other dense soils
istics, attenuation characteristics of soils, in-situ stress condi- is not currently sufficiently investigated to provide quantitative
tions, soil type, and the thickness of the vulnerable zone. A correlations. For Lesaka, South Brooklyn, and West Brooklyn
JOURNAL OF GEOTECHNICAL ENGINEERING / NOVEMBER 1996/927

J. Geotech. Engrg. 1996.122:920-928.

 sites, the loss of friction during pile driving could be partially Leathers. F. D. (1994). "Defonnations in sand layer during pile driving."
responsible for the observed settlements. Vertical and Horizontal Deformations of Foundations and Embank-
ments, Vol. I, Proc. of Settlement '94. A. T. Yeung and G. Y. Felio,
Traffic-induced vibration is characterized with the same vi-
ASCE, New York, N.Y.• 257-268.
bration amplitudes as have caused settlement from pile driv- Linehan. P. W., Longinow, A.• and Dowding, C. H. (1992). "Pipe re-
ing. The thickness of the vulnerable zone may be comparable sponse to pile driving and adjacent excavation." J. Geotech. Engrg.,
for elevated highways supported on piles. But the accumula- ASCE, 118(2).300-316.
tion of the necessary vibration cycles is much slower than in Long, L. T. (1993). "Measurements of seismic road vibrations." Proc.,
pile driving, and settlement may continue for 15-20 yr. No 3rd Int. Conf. on Case Histories in Geotech. Eng., 677 -680.
Massarsch. K. R. (1992). "Static and dynamic soil displacements caused
documented case histories were located. Traffic-induced-vibra-
by pile driving." Keynote Lecture, 4th Int. Conf. on the Application of
tion soil densification should be considered in long-tenn plan- Stress- Wave Theory to Piles. 1-10.
ning in urbanized areas. Picornell. M.• and del Monte. E. (1985). "Pile driving induced settle-
Parametric studies have demonstrated the potential useful- ments of a pier foundation." Proc. of Symp.. vibration problems in
ness of the discussed model for prediction of vibration-induced geotech. engrg., ASCE Convention, G. Gazetas and T. Selig, eds.• New
in-situ settlement of sands. The polynomial model provides a York, N.Y.. 174-186.
Richart, J. E. Jr., Hall, 1. R. Jr., and Woods, R. O. (1970). Vibrations of
better evaluation of expected settlement than widely used test soils and foundations. Prentice-Hall Inc., Englewood Cliffs, N.J.
pile driving. Sutherland, H. B. (1950). Proc., 30th Annu. Meeting of the Hwy. Res.
Downloaded from ascelibrary.org by Sergey Drabkin on 12/11/13. Copyright ASCE. For personal use only; all rights reserved.

Board,406-419.
APPENDIX I. REFERENCES Wiss. J. F. (1981). "Construction vibrations: State-of-the-art." J., Geo-
tech. Engrg. Div., ASCE, 107(2), 167-181.
Barneich, J. A. (1985). "Vehicle-induced ground motion." Proc. Symp.,
Woods, R. D., and Jedele, L. P. (1985). "Energy-Attenuation relation-
Vibration Problems in Geotech. Engrg., Proc., ASCE Convention, G.
ships from construction vibrations." ASCE Symp. on Vibration Prob-
Gazetas and T. Selig, eds.. ASCE, New York, N.Y., 187-202.
lems in Geotech. Engrg., ASCE. New York, N.Y., 229-246.
Boguslavskii, Y.. and Drabkin, S. (1995). "The kinetics of powder set-
Youd, T. L. (1970). "Densification and shear of sand during vibration."
tlement caused by low level vibration and elastic stresses," Physica A,
J.• Soil Mech. and Found. Div.• ASCE. 96(3), 863-880.
222,75-86.
Clough, G. w., and Chameau, J.-L. (1980). "Measured effects on vibra-
tory sheetpile driving." J. Geotech. Engrg. Div., ASCE, 104(10), APPENDIX II. NOTATION
1081-1099.
Dalmatov, B. I., Ershov, V. A., and Kovalevsky, E. D. (1967). "Some The following symbols are used in this paper:
cases of foundation settlement in driving sheeting and piles." Proc.,
Int. Symp. on Wave Propagation and Dynamic Properties of Earth C = wave propagation velocity;
Mat., Univ. of New Mexico with New Mexico Section of ASCE, Al- D = material damping (%);
buquerque, N.M., 607-613. f = vibration frequency (Hz);
Dowding, C. H. (1991). "Pennanent displacement and pile driving vi- H, thickness of in-situ sand layer (mm);
brations." Proc., 16th Annu. Member Conf. of the Deep Found. Inst., h specimen's height (mm);
J. F. PavIa, ed., Sparta, N.J.
ho maximum height of specimen before application of ex-
Drabkin, S. (1995). "Low level vibration induced settlement of granular
soils," PhD dissertation, Polytechnic University, Brooklyn, N.Y. ternal stresses and vibration (mm);
Glasstone, S., Laidler K. J., and Eyring, H. (1941). The theory of rate ii minimum height of specimen after application of exter-
processes. McGraw-Hili Book Co., Inc., New York, N.Y. nal stresses and vibration (mm);
Hardin, B. 0., and Black, W. L. (1966). "Sand stiffness under various earth pressure coefficient at rest;
triaxial stresses." J. SMF Div., ASCE, 92(2), 27-42. coefficient depending on type of propagated wave;
Haupt, W. (1986). "Bodendynamik, grundlagen und anwendungen." p = confining pressure (psi) (6.9 kPa);
Vierweg und Sohn, BraunschweiglWiesbaden, Gennany, 53-105. rio r2 distances from source of vibration (m);
Holtz, R. D., and Kovacs, W. D. (1981). "An introduction to geotechnical s = deviator stress (psi) (6.9 kPa);
engineering." Prentice-Hall, Inc., Englewood Cliffs, N.J.
Kim, D. S., and Drabkin, S. (1995). "Investigation of vibration induced
t = duration of vibration (s);
settlement using multifactorial experimental design." Geotech. Testing v = vibration peak particle velocity (in.ls) (25.4 mmls);
J., 18(4),463-471. Vlo V 2 = vibration amplitudes at distances r, and r2 (in.ls) (25.4
Kim, D. S., Drabkin, S., Laefer, D., and Rokhvarger, A. (1994). "Pre- mmls);
diction of low level vibration induced settlement." Vertical and Hor- Xi = coded values of factors;
izontal Deformations of Foundations and Embankments, Proc. of Set- y= settlement (0.0254 mm);
tlement '94, A. T. Yeung and G. Y. Plmo, eds., ASCE, New York, N.Y., Y settlement (mm);
806-817.
Lacy, H. S., and Gould, J. P. (1985). "Settlement from pile driving in
IX = coefficient of attenuation affected by the material damp-
ing of soils (l/distance);
sands." Proc. of Symp.. Vibration Problems in Geotech. Engrg., ASCE
Convention in Detroit, Mich., G. Gazetas and T. Selig, eds., ASCE.
11 = settlement of in-situ sand layer (mm);
New York, N.Y.. 153-173. P specimen's density (kglm3 );
Lacy, H., Moskowitz, J.• and Merjan, S. (1994). "Reduced impact on Po minimum density of specimen with h = ho (kg/m 3);
adjacent structures using augered cast-in-place piles." Preprint No. p maximum density of specimen with h = ii (kglm 3 ); and
940612. Transp. Res. Board. Washington, D.C. CJ'u = vertical effective stress (kPa).

928/ JOURNAL OF GEOTECHNICAL ENGINEERING / NOVEMBER 1996

View publication stats J. Geotech. Engrg. 1996.122:920-928.