Missouri University of Science and Technology

Scholars' Mine
International Conference on Case Histories in (2004) - Fifth International Conference on Case
Geotechnical Engineering Histories in Geotechnical Engineering

Apr 13th - Apr 17th

Settlement of Shallow Foundations Constructed

over Reinforced Soil: Design Estimates vs.
Measurements
Michael Majchrzak
Kleinfelder Inc., Pleasanton, CA

Marshall Lew
MACTEC Engineering Inc., Los Angeles, CA

Ken Sorensen
Kleinfelder Inc., Sacramento, CA

Tom Farrell
Farrell Design-Build Co. Inc., Placerville, CA

Follow this and additional works at: http://scholarsmine.mst.edu/icchge

Part of the Geotechnical Engineering Commons

Recommended Citation
Majchrzak, Michael; Lew, Marshall; Sorensen, Ken; and Farrell, Tom, "Settlement of Shallow Foundations Constructed over
Reinforced Soil: Design Estimates vs. Measurements" (2004). International Conference on Case Histories in Geotechnical Engineering. 41.
http://scholarsmine.mst.edu/icchge/5icchge/session01/41

This Article - Conference proceedings is brought to you for free and open access by Scholars' Mine. It has been accepted for inclusion in International
Conference on Case Histories in Geotechnical Engineering by an authorized administrator of Scholars' Mine. This work is protected by U. S. Copyright
Law. Unauthorized use including reproduction for redistribution requires the permission of the copyright holder. For more information, please contact
[email protected].
 Settlement of Shallow Foundations Constructed Over
Reinforced Soil: Design Estimates vs. Measurements
Michael Majchrzak, CE, GE Marshall Lew, PhD, CE, GE Ken Sorensen, CE, GE Tom Farrell, MS, CE, GE
Kleinfelder Inc. MACTEC Engineering Inc. Kleinfelder Inc. Farrell Design-Build Co. Inc.
Pleasanton, CA Los Angeles, CA Sacramento, CA Placerville, CA

ABSTRACT

Faced with difficult soil conditions for the support of two 6-story office towers in Dublin, CA and of a 6-story parking garage in
Sacramento, CA, engineers recommended the use of Geopier Rammed Aggregate Piers to reinforce the soil for the support of high
bearing capacity spread footings instead of deep foundations. Foundation selection for both sites was influenced by long-term
settlement performance, schedule and cost savings, and seismic uplift resistance. Rammed aggregate piers were installed to strengthen
upper weak and compressible soil layers resulting in a substantially stiffer soil layer on which shallow, high bearing capacity spread
footings were constructed. Several rammed aggregate piers were also installed with steel anchors to resist seismically induced
overturning forces. Design parameter values were confirmed by full scale aggregate pier modulus tests and uplift tests and a 24 hour
load test at the Dublin site. Total settlements were estimated to be less than 1½ inches for the Sacramento site and less than 1 inch for
the Dublin site. Measured settlements are less than 1 inches total for both projects with differential settlements less than ½ inches,
confirming the design approaches and soil properties used for design. Site selection, rammed aggregate pier design methodology,
modulus and uplift load test results, and measured settlement performance are presented for two projects in California.

INTRODUCTION see Fig.1. The Block 224 garage, located in Seismic Zone 3,
was built for the State of California as part of the Capitol Area
The support of buildings using shallow foundations is East End Project in 2000. The garage houses 753 parking
generally the first consideration of geotechnical engineers spaces and offices on half of the first floor.
when preparing foundation recommendations because of
construction costs, reliable performance, and ease of Kleinfelder Inc. of Sacramento performed a site investigation
construction. However, geotechnical engineers are often faced and identified compressible alluvial soil to depths of 30 and 35
with poor soil conditions that can increase construction costs. feet. Kleinfelder recommended 75 foot long driven concrete
In the past decade, Geopier rammed aggregate piers (RAP) piles to control excessive building settlements because
have gained wide acceptance for strengthening and reducing conventional spread footings could not meet a settlement
the compressibility of soft clays, undocumented fills, and requirement of less than 1.5 inches. McCarthy wanted to
loose sands. The implementation of Geopier RAPs in reduce costs and schedule to win the design-build bid for the
California’s seismic Zones 3 and 4 requires the use of uplift
elements to resist seismically induced overturning forces.

The two case histories discussed illustrate the performance

offered by RAP support of shallow high bearing capacity
footings. These case histories also illustrate that the use of the
RAP foundation system resulted in cost savings within the
same performance standards as other conventional foundation
systems. RAP design methodology, construction, modulus
and uplift tests, and measured settlements are presented.

SACRAMENTO, CA - PROJECT DESCRIPTION

The structure built by McCarthy Building Companies, Inc. is a

6-story parking garage with Class A offices on the first floor, Fig.1. 1301 P Street Parking Garage, Sacramento, CA

Paper No. 1.64 1

project and considered RAP soil reinforcement supporting produces much less noise and vibrations than pile driving,
high bearing capacity footings as a design-build foundation local businesses were less impacted during RAP installation.
alternative to the driven pile design. The RAP foundation Construction was set to begin in February 1999, historically
system was estimated to save at least 60 days on the schedule one of the wettest months of the year in the area.
and $100,000 in construction costs. McCarthy contracted
Consolidated Engineering Labs of Sacramento (CEL), to Building dead plus live loads for the cast-in-place concrete
evaluate the use of the rammed aggregate piers for the project. structure ranged from 138 to 835 kips on gravity columns and
McCarthy was awarded the project and CEL required full- 1600 to 3600 kips at seismic resisting shearwalls. Net uplift
scale load testing prior to completing the foundation design. forces at shearwall ends equaled a maximum of 740 kips.

Penetration Resistance B-1 DUBLIN, CA - PROJECT DESCRIPTION

0NSPT2 USCS
0 The structures built by DPR Construction Inc. of Redwood
CL-Fill City, CA consist of two identical 6-story Class A offices with
a pedestrian bridge in between, Fig. 3. Located in Seismic
3 Zone 4 and in an old army depot, the site had undocumented
fills and old building foundations beneath the ground surface.
ML
10 3

17

ML
20 4
Depth below pad grade (ft.)

8
Fig.3. 6-Story Class A Offices in Dublin, CA
30 7
SC The project was planned to get started early in December
49 2000. DPR and the developer were looking for ways to
reduce costs and build the project through the winter. They
considered RAP soil reinforcement with high bearing capacity
40 78 GW spread footings for the 6-story steel moment frame buildings.
CL
Kleinfelder Inc. of Pleasanton, CA had performed a site
13
investigation which revealed compressible clay soil to depths
of 50 feet. Kleinfelder recommended three foundation
50 14 options: 1) 7 feet of overexcavation and recompaction with
conventional spread footings, 2) 65 foot long pre-cast concrete
Fig.2. Sacramento Site - Soil Profile driven piles and grade beams, and 3) RAPs with high bearing
capacity spread footings. Figure 4 shows a comparison of the
three options. The grading option was deemed to be risky
The site is underlain by undocumented fill to depths of 7 feet,
then alluvial, soft to medium silty clay and sandy silt to 30 Driven piles, pile caps
and grade beams
Overexcavate / recom pact
and spread footings
Geopier soil reinforcement and
high capacity spread footings
feet, then alluvial, dense sand and gravel to about 42 feet, then
medium to stiff clay to about 65 feet and then dense gravels to
7,000 ps f
the maximum depth explored of 80 feet. Figure 2 shows the 3,500 ps f
N=5 - sandy clay
soil profile and standard penetration resistance to a depth of 50
feet near the RAP modulus load test location. The upper medium sti ff silty cl ay

undocumented fill was placed during the late 1800’s and early N=5 sandy silt

1900’s to raise the city above flood waters. Groundwater was Medium dense clayey s and
observed between 8 and 12 feet below the ground surface
during RAP installations. Bounded on three sides by city Stiff to very sti ff silty clay
streets and an alley on the fourth, the site is located in a busy
section of downtown Sacramento. Because RAP construction Fig.4. Foundation systems considered for the Dublin site

Paper No. 1.64 2

because of wet weather conditions. The cost for piles and RAMMED AGGREGATE PIER CONSTRUCTION
grade beams was high compared to the rammed aggregate
piers and footings. RAP and high bearing capacity spread The sequence of RAP construction is shown in Fig. 6. RAPs
footings were selected by the developer and DPR because of are installed by drilling 24-, 30-, 33-, and 36-inch diameter
performance, cost savings, and schedule advantages. shafts and ramming thin 12-inch lifts of well-graded, crushed,
aggregate in the shafts to form very stiff, high-density
The Dublin site was underlain by undocumented clayey fill aggregate piers. The ramming equipment consists of 18 ton to
with old shallow concrete foundations to depths of 7 feet, then 27 ton hydraulic excavators equipped with 2,000 to 4,000
by 12 to 16 feet of medium stiff silty clay, then by soft and pound hydraulic break hammers and specially modified
stiff silty clay to the maximum depth explored of 50 feet. beveled tampers. The hydraulic hammer delivers between 1 to
Figure 5 presents the soil profile, Standard Penetration Test 2 million ft.-lbs. of ramming energy per minute to the beveled
resistance, and Cone Penetration Tip resistance near the RAP tamper at 300 to 500 blows per minute. Figure 7 shows the
load test site. Groundwater was observed at depths of 22 feet typical installation equipment. The ramming action increases
during installation of rammed aggregate piers. the lateral stress in the surrounding soil and increases the
stiffness of the stabilized composite soil mass.
Building dead plus live loads for the steel moment frame
structure ranged from 260 to 970 kips. Being in Seismic Zone
4, maximum service level seismic forces were as high as 1,010
kips. Net uplift forces equaled 599 kips at the buildings core
moment frames.

Penetration Resistance CPT tip resistance (tsf)

0NSPT2 USCS 0 20 40 60 80 100
0 0
5 CL-Fill A. B. C.
Fig.6. Rammed Aggregate Pier Construction Process
8
A. Drill 30-inch and 33-inch RAP shafts
CL-CH B. Ram 2-inch crushed rock into bottom bulb
10 10 10 C. Ram 12-inch lifts of 3/4 or 1½ inch aggregate
base rock to 6” above the design elevation

10 The beveled tamper densifies, forces, and embeds the crushed

aggregate laterally into the sidewalls of the shaft. The result
of RAP installation is a significant strengthening and
20 5 20 stiffening of subsurface soil. In addition, high lateral stresses
Depth below pad grade (ft.)

Depth below pad grade (ft.)

and shear strengths are able to support high bearing capacity

spread footings with minimal settlement. Rammed aggregate
7 piers exhibit friction angles varying from 48 to 52 degrees and
ML dry densities from 140 to 148 pcf (Fox and Cowell 1998).
30 5 30
Depending on pier depth, soil conditions, and construction site
conditions, typical RAP installation rates vary from about 35
12 to 60 elements per day. Slower rates are typical for
installations that require temporary casing. A discussion of
SC quality control testing for rammed aggregate pier installations
40 5 40 can be found in the ICBO ES Report ER-5916 (2002).

RAP installations were performed during the winter months of

1999 and 2001 at the two sites. Two 22 ton hydraulic
CL excavators equipped with 3,500 pound hydraulic break
7 hammers and specially modified beveled tampers were used to
50 50 ram the aggregate lifts at both sites.

At the Sacramento site, drill depths below pad grade were 20

feet for 36-inch RAP uplift elements and ranged between 11
Fig.5. Dublin Site – Soil Profile and 20 feet for 30-inch RAP bearing elements. Crews

Paper No. 1.64 3

 P

DRILL DEPTH = GEOPIER

SHAFT LENGTH (GSL)
UZ = GSL + 1 DIA.
Upper Zone

1 (DIA)
PRESTRESS
ZONE
UZ

2B
Bottom of UZ
LZ

LZ = 2B - UZ
Lower Zone
Fig.7. Typical RAP Installation Equipment

installed 428 RAPs at the Sacramento site during inclement Bottom of LZ

weather in February and March 1999. McCarthy was able to
begin excavation and construction of spread footings after Fig.8. Upper Zone and Lower Zone for Spread Footing
about 50% of the rammed aggregate piers were installed.

At the Dublin site, drill depths below pad grade were 22 feet Upper zone settlement calculations
for 33-inch RAP uplift elements and ranged between 11 feet
and 25 feet for 30-inch RAP bearing elements. Seven hundred As compressive loads are applied to RAP supported footings,
and twenty-eight RAPs were installed during winter weather the stiff piers attract a greater portion of footing- bottom stress
conditions at the Dublin site in January and February of 2001. than the softer matrix soil. The distribution of stress depends
on the ratio of stiffness of RAPs to matrix soil and on the ratio
of area of the RAPs to the gross footing bottom area. Upper
RAMMED AGGREGATE PIER DESIGN CALCULATIONS zone calculations are based on a spring analogy (Lawton and
Fox 1994 and Lawton et al. 1994) and are described in the
Rammed aggregate pier construction increases vertical and following equations:
horizontal stresses in the matrix soil and increases the stiffness
of soil and fill which significantly reduces foundation 1. Footings are assumed to be perfectly rigid relative to the
settlements (Lawton and Fox 1994, Pitt et al. 2003). In high foundation materials. Thus, the stresses applied to the
seismic and high wind zones, the resistance to lateral and composite foundation materials depend on their relative
uplift forces is required. Brief discussions of RAP design stiffnesses and area coverage. From static equilibrium, the
methodologies for settlement control, uplift resistance, and total load on the footing (P), expressed as the product of
lateral resistance are presented and comparisons of calculated applied composite stress (q) and footing area (A), is resisted
to measured settlements are discussed. by a total upward resisting force in the rammed aggregate
piers (Qg) and soil (Qs) materials:

Total Settlement Control P = qA = Qg+Qs = qgAg+qsAs (1)

Design calculations for estimating total settlement of shallow where qg is the stress on top of the RAP, Ag is the area of the
foundations supported on RAPs are well described in the RAPs below the footing, qs is the vertical stress on the matrix
literature (Lawton and Fox 1994, Lawton et al. 1994, Fox and soil, and As is the area of the matrix soil below the footing.
Cowell 1998, Minks et al. 2001, Hall et al. 2002). The design
procedure computes foundation settlements by considering 2. The settlement of the RAP will equal the settlement of the
settlement in the upper, RAP reinforced zone and settlements matrix soil due to the rigid footing condition. The upper zone
in the lower, unreinforced zone of soil below the RAP bottom settlement (suz) of the foundation(s) can be written in terms of
bulb or prestress zone. Figure 8 shows the upper zone (UZ) RAP top stress (qg) and stiffness modulus (kg) or in terms of
and lower zone (LZ) in section. The total settlement is the matrix soil stress (qs) and soil stiffness modulus (ks):
computed as the sum of the upper zone and lower zone
settlement values. suz = qg/kg = qs/ks (2)

3. Equation 2 can be rewritten to express the matrix soil stress

in terms of the RAP top stress and the ratio of the pier and

Paper No. 1.64 4

matrix soil stiffness modulus values by expressing the Table 2: Design Calculations for Footing M-11 at Sacramento
stiffness ratio (Rs) as Rs = kg/ks: Calculation / Property Eq. / Symbol Value
Column D+L load (kips) P 710
qs = qg(ks/kg) = qg/(kg/ks) = qg/Rs (3) Footing width (ft) B 10
Ave. bearing pressure (ksf) q = P/B2 7.1
4. Combining Equations 1 and 3 and by expressing the area No. of RAP Np 7
ratio (Ra) as Ra = Ag/A and rewriting qg in terms of q: RAP diameter (ft) d 2.5
Area replacement ratio (Ra) Ra=Np(d/2)2/B2 0.34
qg = [qRs / (RaRs+1-Ra)] (4) RAP stiffness modulus, (pci) kg 350
Soil stiffness modulus, (pci) km 10.4
Upper-zone settlements are computed using Equations 2 and 4
Stiffness ratio Rs=kg/km 34
which depend on the applied footing stress, the relative
stiffness of the RAP and matrix soil, the area ratio of the RAP top-stress, qg (ksf) qRs/(RsRa-Ra+1) 19.6
RAPs, and the RAP stiffness modulus. RAP stiffness modulus UZ settlement (in) suz=qg/kg 0.39
values are measured in the field with full scale modulus load Rap shaft length (ft) Hs 12
tests. Table 1 presents typical values of kg and allowable RAP UZ thickness (ft) Huz = Hs+d 14.5
and soil composite bearing capacity (qc) based on the standard Zone of footing influence (ft) Ht = 2 B 20
penetration resistance and undrained shear strength of LZ thickness (ft) Hlz = Ht - Huz 5.5
unimproved soil or fill. Mid-depth of LZ (ft) z = Huz + Hlz/2 17.25
Normalized LZ depth (ft) z/B 1.73
Westergaard influence factor Is 0.1
Table 1. RAP stiffness modulus - kg1 and composite qc LZ settlement (in) Consol eqtn 0.45
kg /qc for Silty kg /qc for Silt kg / qc for Total footing settlement (in) Sest = suz + slz 0.84
Nspt Su
Sand and Sand and Clay Peat Max. actual settlement (in) Sactual 0.54
N (psf) (pci) / (ksf) (pci) / (ksf) (pci) / (ksf)
3 500 165 / 5.0 125 / 4.5 75 / 3.5
6 1,250 225 / 6.0 175 / 5.0 110 / 4.0 Based on the soil conditions for the Dublin site (Kleinfelder
9 1,750 260 / 7.0 210 / 6.0 125 / 5.0 2000), a RAP stiffness modulus of 255 pci and a high
12 2,300 285 / 8.0 250 / 7.0 - allowable dead plus live bearing capacity of 6,500 psf were
16 3,000 310 / 8.5 260 / 7.0 - used for design. Table 3 shows the RAP settlement design
25 4,500 325 / 9.0 275 / 7.5 - calculations for a 499 kip gravity column.
>25 5,000 360 / 10.0 300 / 8.0 -
1. For 30-inch diameter RAPs supporting spread footings with
a minimum Ra = 30% (Cowell and Fox 1998). Table 3: Design Calculations for Footing M5A at Dublin
Calculation / Property Eq. / Symbol Value
Column D+L load (kips) P 499
Lower Zone Settlement Estimates Footing width (ft) B 9
Ave. bearing pressure (ksf) q = P/B2 6.2
Estimates of lower zone settlements below the bottom of the No. of RAP Np 5
rammed aggregate pier bulb are computed using conventional RAP diameter (ft) d 2.5
geotechnical settlement analysis procedures well described in Area replacement ratio (Ra) Ra=Np(d/2)2/B2 0.30
the literature and texts (Terzaghi and Peck 1967 and Bowles RAP stiffness modulus, (pci) kg 255
1988) combined with soil elastic modulus values interpreted Soil stiffness modulus, (pci) km 10.0
from the results of in-situ testing data or from the results of Stiffness ratio Rs=kg/km 25.5
laboratory oedometer consolidation tests. Lower zone RAP top-stress, qg (ksf) qRs/(RsRa-Ra+1) 18.6
calculations are also based on a distribution of stress radiating
UZ settlement (in) suz=qg/kg 0.51
from the perimeter of the RAP zone of improvement, which
Rap shaft length (ft) Hs 14
can extend 0.5 to 1 diameter from the drilled edge, depending
UZ thickness (ft) Huz = Hs+d 16.5
on the soil type. A Westergaard or Boussinesq analysis
(Bowles 1988) is used for estimating lower zone footing stress Zone of footing influence (ft) Ht = 2 B 18
influence factors (Is). LZ thickness (ft) Hlz = Ht - Huz 1.5
Mid-depth of LZ (ft) z = Huz + Hlz/2 17.25
Based on the soil conditions for the Sacramento site Normalized LZ depth (ft) z/B 1.92
(Kleinfelder 1998), a RAP stiffness modulus of 350 pci and a Westergaard influence factor Is 0.08
high allowable dead plus live bearing capacity of 7,200 psf LZ settlement (in) Consol eqtn 0.36
were used for design. Table 2 shows the RAP upper zone Total footing settlement (in) Sest = suz + slz 0.87
settlement design calculations for a 710 kip gravity column. Max. actual settlement (in) Sactual 0.75

Paper No. 1.64 5

Fig.9. Typical Detail of a 4-bar RAP Uplift Element

for cohesionless soil - fs = kpsv’ tan (∅’s) (6)

RAP Uplift Elements for cohesive soil - fs = su (7)

Rammed aggregate pier uplift elements were installed to resist where sv’ is the effective vertical stress and kp is the Rankine
seismic overturning and uplift forces. An uplift anchor passive pressure coefficient (Lawton et al. 1994). The
consists of two or four 75 ksi threaded steel bars bolted to a 1- ultimate uplift capacity (Tult) is computed by integrating the
inch thick A36 steel plate. Design lives greater than 100 years unit uplift resistance (fs) over the perimeter area (As) of the
are achieved with oversized bars, special poly coatings, RAP plus the weight of the pier (Wpier):
galvanization, and electrical isolation from the footing
reinforcement. Uplift resistance is developed by perimeter Tult = fs As + Wpier (8)
shearing resistance along the element and is enhanced by the
high lateral stresses at the edges of the shaft (Wissmann et. al. Typical allowable capacities of 50 to 80 kips are developed.
2001 and Caskey 2001). Figure 9 shows a typical detail for a The allowable capacity is typically increased by a factor of
4-bar RAP uplift element. The design procedure computes the one-third or more for seismic loads. The ultimate uplift
unit resistance to vertical movement (fs) as the product of the capacity of RAPs were computed using Equations 6, 7 and 8
effective horizontal earth pressure (sh’) and the tangent of the and confirmed by performing uplift load tests at the two sites.
unimproved soil friction angle (∅’s) or as its undrained shear
strength (su):

Paper No. 1.64 6

RAP Elements and Lateral Resistance

Rammed aggregate piers attract a greater amount of stress

because they are stiffer than the matrix soil under applied
loads (Pitt et al 2003). The combined high friction exhibited
by RAPs and the high normal forces on the piers increases the
ultimate sliding coefficient (fult) for the entire footing bottom.
The design procedure used to estimate the ultimate sliding
friction of a RAP supported footing considers the area ratio,
the stiffness ratio, and the RAP friction angle (∅’g), Eq. 8.

fult = [RsRatan(∅’g)] + [(1-Ra)tan(∅’g)] (8) TELLTALE

(RaRs + 1-Ra)

A detailed discussion of the statics behind Eq. 8 can be found

in Wissmann et al. 2001. A safety factor between 1.5 and 2.0
is typically applied for design to obtain the allowable sliding
coefficient (fall). The allowable load resistance is typically
increased by a factor of one-third or more for seismic loads.

Table 4. Typical values of fall for RAP/soil composites

Soil Classification Typical ∅′s fall1
Sand and Gravel 28° - 45° 0.52 – 0.60
Silt and Clay 20° - 30° 0.51 – 0.52
1. Values computed for Rs = 15, Ra = 33%, and FS = 2

RESULTS OF CONSTRUCTION

Modulus Test Configurations

Figure 10 shows a modulus test section and a photo of the test Fig.10. Section and Photo of RAP Modulus Test Set-up
set up. The test set up consists of a compression element, two
uplift elements, and a reaction frame. The compression
element is loaded to 150% of the maximum top-of-pier stress
matrix soil. This is done by observing the deflections of
calculated from Eq. 5. The load is applied against the reaction
telltales installed into the bottoms of the RAPs. As shown on
frame and resisted by the uplift elements. A telltale is
Fig.12 at the Dublin site, the bottom of the test pier is shown
installed at the bottom of the modulus test pier, just above the
to move only slightly while deformations at the top of the pier
bottom bulb, to facilitate the measurement of bottom-of-pier
increase at a growing rate. This behavior is interpreted to
deflections. During testing, the deflections at the top and
indicate that the RAP is bulging outward.
bottom of the RAP are measured.
A 24 hour test was performed at the Dublin site, where the
load of 117% times qg (105 kips) was applied for 24 hours.
Modulus (Compression) Test Results
The total deflection after 24 hours was 0.03 inches.
Figures 11 and 12 present the results of the modulus tests
Table 3. Results of Modulus Tests for both sites
performed at the Sacramento and Dublin site respectively.
The purpose of the modulus test is to verify the RAP stiffness Sacramento site: 36-inch dia. x 12 foot GSL
modulus (kg) used for design calculations in Eq.(s) 2 and 4 at Design top-of-pier stress (qg) = 18,224 psf
117% of the design top-of-pier stress, qg. Table 3 presents the 117% x (qg) Deflection Tested Modulus
tested RAP deflections, tested top-of-pier stress, and tested 21,322 psf 0.29 inches 510 psi/inch
RAP stiffness modulus for both sites.
Dublin site: 30-inch dia. x 18 foot GSL
Although the purpose of the modulus test is to verify the RAP Design top-of-pier stress (qg) = 18,649 psf
stiffness modulus used for design calculations, the tests may 117% x (qg) Deflection Tested Modulus
also be used to add insight into how the RAP behaves in the 21,819 psf 0.23 inches 670 psi/inch

Paper No. 1.64 7

 0 5 10 15
R A P T o p S tre s s (k s f)
20 25 30 35 40
MEASURED SETTLEMENT PERFORMANCE
0 .0 0
0 .0 0
0 .0 5
0 .0 8
0 .1 0
0 .1 7
0 .1 4 Two sites with compressible undocumented fill and soft clay
0 .1 8
0 .2 0
0 .2 7
0 .2 2 in c h e s a t 1 0 0 % d e s ig n lo a d = 1 2 8 k ip s
0 .2 3
soil were reinforced with rammed aggregate piers to support
high bearing capacity footings. The measured results of total
Average RAP Deflection (inches)

0 .2 3 0 .2 9 in c h e s a t 1 1 7 %
0 .3 0

0 .4 0 0 .3 8
0 .3 9
0 .4 3 settlements after the addition of live loads are presented.
0 .4 5 in c h e s a t 1 5 0 % d e s ig n
0 .5 0 lo a d = 1 9 3 k ip s
0 .5 3

0 .6 0 0 .5 7

0 .7 0
Sacramento Site
0 .8 0

0 .9 0
0 .8 5 in c h e s
McCarthy was able to begin excavation and construction of
concrete spread footings after about 50% of the RAPs were
a t 1 9 5 % d e s ig n
T o p o f R A P d e fle c t io n fo r 3 6 - in c h d ia . x 1 2 f o o t G S L b e a r in g e le m e n t lo a d = 2 5 0 k ip s
1 .0 0
installed. As the first floor columns of the concrete structure
Fig.11. Modulus test results for Sacramento site
were being poured, initial baseline top-of-footing surveys
R A P T o p S t r e s s ( k s f)
were measured. Once the sixth floor pour was completed,
0 5 10 15 20 25 30 35 40 brass settlement monitoring monuments were set at 12
0 .0 0
0 .0 1
0 .0 7
0 .0 4 T e llta le d e fle c t io n
0 .0 9 in c h e s
locations which included gravity columns with dead plus live
0 .1 0
0 .1 7 ( 1 0 0 % d e s ig n
s tr e s s ) loads ranging from 138 kips to 835 kips and at two shearwalls
0 .2 0
0 .2 8
0 .1 9 (1 1 7 % d e s ig n
s tre s s ) with dead plus live loads of 1,200 and 1,800 kips at the each
Average RAP Deflection (inches)

0 .3 1
0 .3 0
0 .3 7
0 .3 1 0 .3 5 ( 1 5 0 % d e s ig n
s tr e s s )
end. Measurements were obtained until settlements flattened
0 .3 5
0 .4 0

0 .5 0
0 .4 2 after 3 years from initial baseline readings.
0 .5 0
0 .5 2 0 .5 3 0 .5 3 a t 1 8 0 k ip s
0 .6 0
2 4 h o u r lo a d t e s t a t 1 0 0 % d e s ig n The results of foundation settlement surveys are plotted
s tre s s = 1 8 ,6 4 9 p s f o r 9 2 k ip s .
0 .7 0 M e a s u r e d 0 .0 3 in c h e s a ft e r 2 4 h o u rs . against time in Fig.14. The results indicate that the foundation
0 .8 0 settlements have ranged between 0.25 and 0.75 inches with
0 .9 0
T o p o f R A P d e fle c tio n fo r 3 0 - in c h d ia . b y 1 8 fo o t G S L b e a rin g e le m e n t
both the maximum value and the average of the values less
1 .0 0
T e llta le d e fle c tio n a t th e b o tto m o f th e R A P
than the design estimates.
Fig.12. Modulus Test Results graph for Dublin Site
D+L=138 kips M1

Uplift (Tension) Test Results D+L=421 kips M3

D+L=357 kips M5 N

D+L=710 kips M2 D+L=710 kips M4

The results of the uplift tests for both sites are presented in
Fig. 13. At the design uplift load of 83 kips for the
Sacramento site, a deflection of 0.19 inches was measured. At shear wall
D+L=1200 kips each end M7
the design uplift load of 75 kips for the Dublin site, a D+L=835 kips M11 D+L=510 kips M8 D+L=379 kips M6
shear wall
deflection of 0.37 inches was measured. Temporary
elongation of the 75 ksi threaded bars accounts for about one- D+L=210 kips M12 D+L=1800 kips each end M10
D+L=210 kips M9
half of the measured deflection which is supported by the
rebound and the final deflection shown in Fig. 13. Uplift tests
are performed up to 200% of the allowable uplift capacity.
Time (month-year)
May-99

May-00

Aug-01
Nov-98

Nov-99
Sep-99

Sep-00

Dec-00

Dec-01
Mar-99

Mar-00

Feb-01

Jun-01

Feb-02
Jan-99

Jan-00

Apr-01

Apr-02
Oct-00

Oct-01
Jul-99

Jul-00

1.00 0.0
Dublin Site - 33-inch dia. x 15 ft G SL with four #8 bar anchor

0.90 Sacram ento Site - 36-inch dia. x 14 ft G SL with two #10 bar anchor 0.90

0.80 0.81
Average RAP Uplift Deflection (inches)

0.77
1st floor
0.70 concrete pour
0.69 0.66
-0.5
0.60
0.58
Settlement
(inches)

0.51 6th floor

0.50 0.51 0.49 concrete pour
Garage opens
0.42 Feb 2, 2000
0.40
0.37 0.37
1 year of service 2 years of service
0.32 0.34
0.30 (full live load) (full live load)
-1.0 Jan 25, 2001 Feb 15, 2002
0.25 0.24 0.25
0.20 0.208
0.19 0.18
0.13 0.14 M-1 (138 kips) M-2 (710 kips) M-3 (421 kips)
0.10
0.08 0.07 M-4 (710 kips) M-5 (357 kips) M-6 (379 kips)
0.00
0.03 M-7 (1200 kips) M-8 (510 kips) M-9 (210 kips)
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 M-10 (1800 kips) M-11 (835 kips) M-12 (210 kips)
Uplift Load (kips) -1.5

Fig.13. Uplift Load Test Results at both sites Fig.14. Surveyed locations and settlements Sacramento site

Paper No. 1.64 8

The footing settlement calculated in Table 4 was 0.84 inches SUMMARY AND CONCLUSIONS
and actual measured settlements were about 0.54 inches. If
the tested modulus value of 510 pci is used then the calculated Two projects at soft soil sites in Sacramento and Dublin,
and measured settlements are in closer agreement. California, utilized the Geopier rammed aggregate pier system
to support shallow high bearing capacity spread footings for
six-story building structures. The projects and the soil
Dublin Site conditions have been described. The RAP elements were
installed to average depths of 16 feet below footing bottoms to
The Dublin project started at the end of December 2000 reinforce the undocumented fill and soft to medium-stiff clay
during winter rains. Load tests were performed during and natural clay soils at both sites. The RAP foundation
inclement weather for about a week. Settlement surveys system replaced 65 and 75 foot concrete driven pile designs.
started after erection of 3rd and 5th floor steel before concrete Settlement surveys conducted during and after the
decks were poured. This accounts for about 20% to 25% of construction of the buildings revealed that the RAP-supported
the dead loads. Several locations where monitored including footings settled ¼ to ¾ inches under the applied total loads.
gravity columns with dead plus live loads ranging from 300 The measured settlement values are in good agreement with or
kips to 600 kips and at moment frame mats with dead plus live less than the design estimates.
loads of 1,500 and 2,300 kips. Settlements were measured up
to the end of construction and once more after live load was Explanations to why the measured settlements are lower than
applied. predicted values include: 1) the real benefit from increased
lateral stress in the reinforced soil is not accounted for in RAP
The results of foundation settlement readings for the Dublin design procedures (Handy 2001), 2) the reduction in vertical
site are plotted against time in Fig.15 The footing settlement stress to the lower zone due to the positive group interaction
calculated in Table 5 was 0.87 inches and actual measured effects of the RAP system are not included in the design, 3)
settlements ranged between 0.7 to 0.75 inches. The measured conservative estimates were made for the consolidation
settlements are in good agreement with the estimated values. behavior of the lower zone soil, 4) predicted upper zone
settlements are based on lower “more conservative” RAP
stiffness modulus values when compared to actual tested
modulus values, and 5) secondary compression in the lower
zone may still occur over time.
D+L=398 kips M1A
D+L=398 kips
M5A and M9B
These case histories illustrate that the use of RAP soil
reinforcement to support high bearing capacity footings
resulted in cost savings within the same performance
standards as other conventional foundation systems.
D+L=398 kips
M4A and M1B

ACKNOWLEDGEMENTS
D+L=648 kips M2B
D+L=398 kips D+L=398 kips
M3A and M3B D+L=597 kips M4B
M2A and M5B The authors are grateful to Dr. Kord Wissmann of Geopier
Foundation Company and Bill Kenney and Aaron Taylor of
Time (month - year)
Farrell Design-Build Companies, Inc. for their assistance with
Nov-01

Dec-01
Sep-01
Jun-01

Feb-02
Apr-01

Apr-02
Jul-01

0
the settlement and geotechnical data and helpful comments.
The authors express their gratitude to Dr. Nathaniel S. Fox for
his careful review and helpful comments. These projects
would not be possible without their investors who are the State
of California and corporate businesses, we are grateful for the
-0.5
opportunity to report good performance of their investments.
Settlement
(inches)

Bldg A concrete decks

complete. Bldg B begins
interior tenant improvements. Live loads After 3 mos. of REFERENCES
-1 introduced Live load

M1A (398 kips) M2A (398 kips) M3A (597 kips)

Bowles, J.E. (1988) Foundation Analysis and Design. 4th
M4A (297 kips) M5A (499 kips) M1B (297 kips)
M2B (648 kips) M3B (597 kips) M4B (597 kips)
Edition. McGraw-Hill, Inc. New York, 1004 pp.
M5B (398 kips) M9B (499 kips)
-1.5
Caskey, J.M. (2001). “Uplift Capacity of Rammed Aggregate
Pier Soil Reinforcing Elements.” Masters Thesis. University
Fig. 15. Footing plan and settlements Dublin site
of Memphis. December 2001.

Paper No. 1.64 9

Farrell Design-Build Companies, Inc. (January 1999). Block Wissmann, K.J., and FitzPatrick, B.T., and Lawton, E.C.
224 Parking Garage Geopier Design and Construction (2001). “Technical Bulletin No. 4 – Geopier Lateral
Recommendations, Sacramento, CA. Resistance.” Geopier Foundation Company, Inc. Scottsdale,
Arizona.
Farrell Design-Build Companies, Inc. (June 2000).
Corporate Headquarters - Geopier Design-Build Submittal,
Dublin, CA. Symbols used in order of appearance:

Fox, N.S. and Cowell, M.J. (1998). Geopier Foundation and P dead plus live load downward force on a footing
Soil Reinforcement Manual. Geopier Foundation Company, q applied bearing pressure
Inc., Scottsdale, Arizona. A area of the footing bottom
Qg load resisted by rammed aggregate pier
Hall, K.M., Wissmann, K.J., Caskey, J.M., and FitzPatrick, Qs load resisted by soil
B.T. (2002). “Soil reinforcement used to arrest bearing qg top stress on rammed aggregate pier
capacity failure at a steel mill.” Proceedings, 4th International Ag area of rammed aggregate piers below footing
Conference on Ground Improvement. Kuala Lumpur, qs bearing stress on soil
Malaysia, 26–28 March. As area of soil below footing
suz upper zone settlement
Handy, R. L. (2001). “Does Lateral Stress Really Influence kg stiffness modulus of rammed aggregate pier
Settlement.” ASCE Journal of Geotechnical and ks stiffness modulus of unimproved soil
Geoenvironmental Engineering, Vol. 127, No. 7. Rs stiffness ratio
Ra area ratio
ICBO ES Report ER-5916 (Reissued September 1, 2002) Nspt standard penetration test blow counts
ICBO Evaluation Service, Inc. • 5360 Workman Mill Road, su undrained shear strength
Whittier, California 90601 • www.icboes.org Hs length of drilled shaft below footing bottom
Huz thickness of upper zone soil
Kleinfelder, Inc. (June 1998). Geotechnical Investigation Ht thickness of total zone of stress influence
report for Proposed Parking Structure at 13th and P Street, Hlz thickness of lower zone soil
Sacramento, CA. Is stress influence factor at mid-depth of lower zone
fs vertical rammed aggregate pier shaft resistance
Kleinfelder, Inc. (February 2000). Geotechnical Investigation sh’ effective horizontal earth pressure
report for Proposed Corporate Headquarters, Dublin, CA. kp Rankine horizontal earth pressure coefficient
sv’ vertical effective stress
Lawton, E.C., and Fox, N.S. (1994). “Settlement of structures
∅’s effective soil friction angle
supported on marginal or inadequate soils stiffened with short
aggregate piers.” Vertical and Horizontal Deformations of Tult ultimate uplift resistance
Foundations and Embankments, A.T. Yeung and G.Y. Fello Wpier weight of rammed aggregate pier
(Editors), American Society of Civil Engineers, 2, 962-74. ∅g’ rammed aggregate pier friction angle

Lawton, E.C., Fox, N.S. Fox, and Handy, R.L. (1994).

“Control of settlement and uplift of structures using short
aggregate piers.” In-Situ Deep Soil Improvement, Proc.
ASCE National Convention, Atlanta, Georgia. 121-132.

Minks, A.G., Wissmann, K.J., Caskey, J.M., and Pando, M.A.

(2001). “Distribution of Stresses and Settlements Below Floor
Slabs Supported by Rammed Aggregate Piers.” Proceedings,
54th Canadian Geotechnical Conference. Calgary, Alberta.
September 16–19.

Pitt, J.M, White, D,.J., Gaul, A., Hoevelkamp, K. (2003).

Highway Applications For Rammed Aggregate Piers In Iowa
Soils. Iowa DOT Project TR-443.

Terzaghi, K., Peck, R.B. (1967). Soil mechanics in

engineering practice. 2nd edition, John Wiley and Sons, New
York.

Paper No. 1.64 10