Académique Documents
Professionnel Documents
Culture Documents
Current design practice for axially loaded piles and piled rafts in
Germany
C. Vrettos
Division of Soil Mechanics and Foundation Engineering, Technical University of Kaiserslautern,
67663 Kaiserslautern, Germany
ABSTRACT: The new generation of German DIN standards for geotechnical works adopted the philosophy
of the Eurocode EC7 introducing the limit state design approach. A summary of the relevant national
standards for pile design and construction is given. The design rules and the respective code provisions are
presented in detail, both for ultimate and serviceability limit states. Two illustrative pile design examples are
described. The second part of the paper provides an overview of the guideline contents for the design and
construction of piled raft foundations.
3.3 Characteristic resistances The use of the correlation factor was not
customary in Germany and is a compromise with
It is customary in Germany to determine the axial respect to the future introduction of the Eurocode.
pile resistances not via calculation methods but on The values of the correlation factor in the
the basis of
DIN 1054:2005 are different from those given in the
- Static load tests
- Dynamic load tests EC7-1 in combination with the National Annex, cf.
- Empirical correlations Frank (2006). The procedure defined in the DIN
basically corresponds to the experience in Germany
with the global safety factor concept.
3.4 Pile resistance from static load tests In the event where the limiting resistance is not
Following the DIN 1054:2005, when piles are explicitly identified from the measured resistance vs.
designed from static load tests the measured settlement curve it is assumed that it occurs at a
resistance values R1m,j are obtained from the in-situ settlement of
experimental resistance vs. settlement/heave curves
(RSC/RHC). From these data one then determines s1 = 0.10 Db (3)
the characteristic RSC/RHC curves that are the basis
for the design and the capacity check. where Db is the diameter of the pile base.
3.4.1 Characteristic resistance for ULS 3.4.2 Characteristic resistance for SLS
Characteristic values of resistance for the limit state The transfer of the rules defined in Table 1 to the
GZ 1 are obtained from the measured ones R1m,j by derivation of characteristic resistance vs. settlement
dividing them with a correlation factor . This curves (RSC) for the serviceability limit state could
corresponds to the recommendation given in EC7-1. produce undesirable jumps in the characteristic
The characteristic value R1,k is obtained as the RSC. DIN 1054:2005 recommends a qualified
minimum of the pile load test results according to weighting of the results of pile tests at the SLS
equation (1): enabling the derivation of a characteristic RSC that
should be continuous.
R1,k = Min {R1m ,min / ; R1m / } (1) A practicable solution is the following: If the
measured RSC show relatively uniform dispersion
with values smaller than s N / R1m = 0.25 or
where R1m,min and R1m are the minimum and average
s N / Rm = 0.25 a single correlation factor is
values of measured resistance, respectively, is the determined using the average value Rm , whereas for
correlation factor as given in Table 1, and
values larger than 0.25 the correlation factor is
determined using Rm, min . Only for non-uniform
i =1( R1m R1m, j ) 2 /( N 1)
N
sN = . (2)
dispersion with values lying below and above the
limiting value 0.25 would jumps in the curve be
If more than one pile contributes to the load transfer produced. A meaningful correction is then applied to
to the ground, e.g. when using a rigid pile cap, and if the experimental curve that has to be verified by a
the dispersion coefficient s N / R1m 0.25 , it is geotechnical expert.
allowed to use from Table 1 the average value R1m of
3.4.3 Static axial load tests CASE and TNO formulae and the CAPWAP
The performance of pile tests of static or dynamic TNOWave, respectively. In the Rapid Load Test the
nature is standardised in Germany in the direct method is performed by means of the
Recommendations for Static and Dynamic Pile Tests Unloading Point Method. This procedure is however
of the German Geotechnical Society, Working not allowed by the DIN 1054:2005.
Group 2.1, DGGT (1998).
One pile test should be carried out for each
Table 2. Increase of correlation factor according to Table 1
uniform subsoil condition, whereas for foundations when evaluating dynamic pile load tests.
with more than hundred piles two pile tests are
recommended. For sensitive buildings and buildings 1) Calibration on static pile load tests
of geotechnical category GK 3 (for the definition of Location of pile Analysis procedure Increase
GK see below) the number of pile tests is defined in test
cooperation with the geotechnical expert. For Same site Extended method None
micropiles DIN 1054:2005 requires at least 2 piles to (e.g. CAPWAP)
be tested, at minimum at 3% of all piles. Direct Method = 0.10
The testing load shall be large enough in order to (e.g. CASE-formula)
reach the limit state GZ1 fulfilling one of the Different project Extended method = 0.05
following criteria: Direct method = 0.15
a) Limiting settlement: 2) Deduction from general empirical values = 0.15
R g1 = R ( s g ) (only extended method permitted)
Note: When applying the correlation factors from Table 1 to
with sg = 0.1D whereby D is the pile diameter for dynamic load tests, the double number of tests must be
uniform cylindrical piles, the diameter of the pile available in the first column in Table 1 (e.g. static N = 1,
base for piles with widened base, and the equivalent dynamic N = 2).
pile diameter (same area) for piles with rectangular
cross section. 3.6 Pile resistance from empirical correlations
b) Creeping behaviour:
R g1 = R ( k s ) This applies mainly to bored piles and corresponds
to the traditional practice followed in Germany.
where ks is an individually specified amount of
Characteristic values for base resistance qb,k and
creep.
shaft friction at each layer qs,k,i are directly
With respect to the necessary instrumentation
separate monitoring of shaft and base resistance is determined from values of soil parameters using
required when soil is strongly nonhomogeneous, appropriate charts and tables. The variation of pile
shaft resistance after reaching peak value drops to a resistances with settlement s underlying the design
much lower residual value, and when strict criteria procedure is depicted in Figure 1, with Rk(s), Rb,k (s),
are imposed on limiting displacement values. In and Rs,k(s) denoting the characteristic values of
standard cases recording of pile head settlement is total, base, and shaft resistance of the pile,
considered sufficient. respectively. The characteristic axial pile resistance
is
3.5 Pile resistance from dynamic load tests Rk ( s ) = Rb,k ( s ) + Rs ,k ( s ) = qb,k Ab + q s ,k ,i As ,i (4)
i
Under certain circumstances DIN 1054:2005 allows where Ab and As,i are the pile base area and shaft
the determination of pile resistances from dynamic surface in layer i.
load tests. Correlation factors have also to be
In Figure 1 sg denotes the limiting settlement with
considered as given in Table 1. The required number
s g = 0.10 Db
of dynamic tests as well as the increase in the value
of the factors is summarized in Table 2 in that is set equal to the settlement at the ultimate limit
dependence on the available data from comparable state GZ 1.
static tests and the testing procedure used. For the pile settlement at the SLS the relevant
With respect to the data reduction procedure it is parameter is the pile shaft diameter Ds.
distinguished between direct and extended methods. The limiting settlement for the mobilisation of
The former are based on one-dimensional wave shaft resistance is defined by the following equation
propagation theory and allow the direct on-site with settlements given in [cm] and resistance in
estimation of the bearing capacity reached, while the [MN]:
latter incorporate the complete model of the pile s sg = 0.50 Rs , k ( s sg ) + 0.50 3.00 (5)
embedded in the soil. Typical examples are the
Table 4. Pile base resistance for bored piles in cohesive soils
according to DIN 1054:2005.
Table 9. Partial safety factors for pile resistances for the R 2,k R 1,k R R 2,k R 1,k R
ultimate limit state GZ 1B according to DIN 1054:2005. s2,k min
s 2,k 2s 2,k
s 2,k
Pile resistance Factor LF1-LF3 Secant for
spring coefficient
Resistance in compression from s 2,k max
Pc 1.20
pile load test s1,k Rk s1,k Rk
Resistance in tension from pile
Pt 1.30
load test
s s
Resistance in compression and
(a) (b)
tension from empirical P 1.40
correlations
Figure 2. Resistance of single pile R2,k at SLS from a
characteristic resistance vs. settlement curve according to
DIN 1054:2005 for small (a) and considerable (b) expected
Finally, the bearing capacity for the ultimate limit differential settlements between piles.
state is checked requiring
Table 10. Ultimate values for pile shaft friction for design
example
s sg = 0.50 Rs,k + 0.50 = 1.20 cm . Figure 4. Resistance vs. settlement curve from empirical
correlations for the design example
Resistance R [MN]
0 2 4 6
0
Rk,rigid
-2 Rk,flex
Pile D = 0.90 m Rm,max
Settlement s [cm] Rm,min
-4
Figure 5. Design situation for example of pile design from
static load tests
-6
5.2.1 Characteristic pile resistance
The characteristic pile resistance vs. settlement
curve is determined according to DIN 1054:2005 as -8
follows. If it is assumed that the load from the
superstructure is not distributed to more than one
pile (i.e. load transfer from weak to strong -10
piles), the design departs from a flexible pile cap,
and the characteristic values of the resistance vs. Figure 6. Derivation of characteristic resistance vs. settlement
settlement curve is determined from the minimum curves form the pile load tests.
values of the test results Rm,min. The correlation
factor is = 1.05, cf. Table 1. The resulting pile
resistance is given in Table 13 and Figure 6.
E1, d = GG , k G + FQ , k Q =
If on the other hand the pile cap is rigid enabling the
redistribution of pile loads, the characteristic pile = 1.2 1.35 + 0.6 1.50 = 2.52 MN
resistance is allowed to be determined from the
average value from all pile tests conducted. For R1, d = R1, k / Pc = 3.76 / 1.20 = 3.14 MN > 2.52 MN
relative values of the data dispersion s N / Rm 0.25
the correlation factor is obtained by linear
b) Load distribution by rigid cap (rigid system):
interpolation and is applied to the entire resistance
vs. settlement curve thus avoiding discontinuities in
the curve. The resulting resistance vs. settlement E1,d = 2.52 MN
curve is depicted in Figure 6 with the numerical R1, d = R1, k / Pc = 3.92 / 1.20 = 3.27 > 2.52 MN
values summarized in Table 13.
5.2.3 Serviceability check
5.2.2 Bearing capacity check From the characteristic loads F2,k resulting from the
At the ultimate limit state GZ 1B the condition structural analysis we have
E1,d R1,d
E2,d = E2,k = FG,k + FQ,k = 1.20 + 0.60 = 1.80 MN
has to be verified. For load case LF 1 we have: The allowable characteristic pile resistance in the
a) Independently acting piles (flexible system): serviceability limit state is obtained from Figure 6:
R2,d = R2,k = 2.10 MN > 1.80 MN (flexible system) individual piles and between pile and raft as
schematically shown in Figure 7. Back analysis of a
R2,d = R2,k = 2.25 MN > 1.80 MN (rigid system) static pile test is commonly used to verify the
assumptions of the model and the values of the
Hence, for both cases the SLS check is fulfilled. parameters.
The serviceability limit check can be also
performed via the pile settlement: From Figure 6 we Ftot,k
obtain for the characteristic pile load F2,k = 1.80 MN
the settlement for a flexible system s2 = 1.50 cm,
while for a rigid system s2 = 1,30. Both values are
smaller than the allowable settlement of 2.00 cm.
(x,y)
Rpile,k,1 Rpile,k,j
6 PILED RAFT FOUNDATIONS
6.1 General and design methods
Piled rafts are a new foundation concept for
important high-rise buildings and have been
successfully used in Germany since the beginning of (x,y)
the 90s, Katzenbach et al. (2000). This foundation
type is a viable alternative to conventional pile or
raft foundations in competent ground. The combined z
foundation is able to support the applied axial
loading with an appropriate factor of safety at a
tolerable level of settlement under working loads. qs,1(z)
The implementation of this foundation type led to an qs,j(z)
abolishment of complicated settlement-correction qb,1 qb,j
techniques. During the last years the computational
methods available in combination with Interactions: 1 Pile -soil
measurements on real projects allowed the realistic 2 Pile-pile
modelling of the complicated bearing behaviour of 3 Raft-soil
that composite foundation system. The bearing 4 Pile-raft
behaviour is described by means of the pile raft
coefficient pr that defines what amount of the total Figure 7. Interactions in a piled raft foundation (Hanisch et al.,
force Ftot,k is carried by the piles. Hence, pr ranges 2001).
from 0 for a raft to 1 for a pile foundation. Due to
the strong nonlinearity of the pile bearing behaviour The design work consists in estimating the
the pile raft coefficient depends on the stress level deformation behaviour of the composite system and
and accordingly also on the amount of settlement of the distribution of the load into its two components,
the piled raft foundation. The piles can be loaded up pile group and raft. The available methods may be
to their ultimate bearing capacity, and are spaced divided into i) approximate analytical, ii)
strategically to achieve a more uniform settlement approximate numerical, and iii) refined numerical,
thus reducing sectional forces in the raft and leading the choice being dictated by the importance of the
to a more economical solution. project. An overview is given by Poulos (2001).
This new foundation concept has been introduced The aim of the design process is to optimize the
very recently in the DIN 1054:2005 in Section 8.5.5. position and the geometry of the piles. The solution
For the design reference is made to the of the soil-structure interaction problem is obtained
corresponding guideline (Hanisch et al., 2001) and by means of a pseudo-coupled procedure that is
to the fact that a project specific approval by the based on an interaction between the designers of the
building control authority is required. superstructure and the foundation system, resp. The
The guideline requires the computational model interface in this design procedure is defined in terms
applied to be able to simulate: i) the bearing of the modulus of subgrade reaction for the raft and
behaviour of a single pile taking into account the the spring coefficients for the individual piles.
shearing along the pile shaft and the compression at For the preliminary design, where different
the pile base, and ii) the interaction effects between foundation alternatives are compared, a flexible
simplified method is required to assess the influence - Homogeneous soil conditions with no abrupt
of the pile group configuration and of the soil changes in soil stiffness;
parameters. An approximate method that is an - Foundation loaded at its center of gravity with no
extension of the method outlined by dynamic effects.
Randolph (1994) has been presented by the author It should be noticed that the check of the external
(Vrettos, 2006). The method allows the bearing capacity of the individual piles is not
accommodation of the variable pile distance and required.
length as well as the different stiffness of central,
peripheral, and corner piles. 6.2.2 Internal bearing capacity
For the final design a nonlinear finite element All foundation elements shall be checked against
analysis with an appropriate soil model is usually material failure for all relevant combinations of
carried out for important projects. In the past, the actions. In particular, piles shall be designed for
majority of these calculations used an elastoplastic compression combined with bending and shear,
Drucker/Prager cap model mainly due to the fact tension during construction stages, and downdrag.
that this formulation is already implemented in The raft shall be checked for bending, shear, and
numerical codes. punching shear at the slab-column and slab-pile
connections.
The partial resistances of raft and piles depend on
6.2 Ultimate Limit State
the stress level induced by the actions. Therefore it
The guideline for piled rafts adopted the limit state is required that the pile raft coefficient pr be
design philosophy and distinguishes for limit state calculated both for the ultimate and the
GZ 1 between external and internal bearing capacity. serviceability limit state. The sectional forces of the
The methodology defined by the EC7-1 for piles can raft and of the piles are computed for the load
not be directly applied to piled rafts. distribution on piles and raft corresponding to the
pile raft coefficient. The more unfavourable results
6.2.1 External bearing capacity are then used for the structural design.
The calculation of the resistance vs. settlement
curves for the overall foundation system is carried
out using a computational model accounting for the 6.3 Serviceability Limit State
interaction effects and applying a global safety
Checking is carried out for limit state GZ 2 in
factor to the characteristic values of actions. analogy to the bearing capacity check by setting for
the partial safety factor a value equal to 1.
n
Fk ,i R1,tot,k (13)
6.3.1 External serviceability
i =1
The requirements of serviceability of the system are
with equal to 2.0, 1.75, and 1.50 for load case fulfilled when
LF 1, LF 2, and LF 3, respectively.
Assessment of the characteristic value of the total n
resistance R1,tot,k is made on the basis of the induced
E Fk ,i C .
(14)
i =1
total settlement, i.e. the point at which the settlement
starts to increase at a higher rate. The effects E dependent on the actions Fk,i are
If the check is not performed by an adequate computed by an adequate model based on
realistic computational model it is allowed in simple characteristic values of the material properties.
cases to calculate the total resistance R1,tot,k by means The resistance parameter C is in most cases the
of the characteristic value of the base resistance of tolerable settlement of the piled raft foundation. It is
the foundation raft, i.e. ignoring the vertical bearing defined e.g. from the requirements referring to
capacity of the piles. Dowel resistance of the piles connections of service pipes in the building, the
along the failure surface may be considered in the settlement of adjacent buildings, or the bending and
analysis. tilting of the superstructure itself. In practice, for the
Simple cases are defined as follows: serviceability of elevators but also for optical
- Geometrically uniform piled raft, i.e. identical pile reasons tilting shall be limited to 1:1000. It should
length and diameter, constant pile spacing, also be noted that according to observations - even
rectangular or circular raft, overhang of less than for homogeneous soil conditions, symmetrical
three pile diameters at the edges of the raft; building and loading - differential settlements may
reach values in the order of 30% to 50% of the total - Verification of the computational model and the
settlement. assumptions made regarding soil parameters;
DIN 1054:2005 specifies for the serviceability - Timely detection of critical situations;
check a value of the safety factor for the actions - Observation of the evolution of settlements during
equal to 1.0 both for permanent and variable loads. the construction.
However, relevant for serviceability are the The program includes measurement of:
settlement-inducing loads. It is therefore - Load-settlement behaviour of the foundation by
recommended here when consolidation effects are geodetic methods;
of minor importance to reduce the permanent load - Load share between raft and piles by recording
of the superstructure by the own-weight of the soil-raft contact pressure and pile loads;
foundation slab and to add in absence of detailed - The bearing behaviour of typical piles (central,
information 30% of the variable load. edge, corner) comprising the pile head and pile base
force as well as the distribution of skin friction along
6.3.2 Internal serviceability the pile shaft;
Piles have to be checked for the allowable crack - The variation of soil deformation with depth.
width. The raft has to be checked in addition for the In simple cases measuring only the load-
amount of differential settlements that may lead to settlement behaviour will be sufficient.
structural distortion of the superstructure.
7 CONCLUSIONS
6.4 Checking of design and supervision of
construction
The adoption of the Eurocodes by the national DIN
Design and construction shall be approved by a standards resulted in important changes in the design
geotechnical expert particularly qualified for this philosophy with a unified set of principles for all
job. The following works have to be done: geotechnical design. It motivates a systematic
-Examination of the ground investigations and thought about uncertainty in soil parameters and
laboratory testing program and the interpretation of makes clear distinction between ultimate and
the results serviceability limit states. In German standards the
- Plausibility check of the characteristic values of basic principle for selecting numerical values for the
soil parameters used in the computational model partial safety factors is maintaining the safety level
- Examination of the computational model adopted of the former global safety concept. The
for the design and of the numerical results using an methodology for piles outlined in the
independent calculation method DIN 1054:2005 standard follows past experience
- Set-up and supervision of the monitoring program, and provides guidance by specifying pile resistances
and verification of design assumptions against field for a variety of pile types and soil conditions.
observations. The design methodology for piled rafts is
The construction of the piled raft foundation shall embedded in the limit state design philosophy.
be supervised by a geotechnical expert assigned by Design work requires the close cooperation between
the client and approved by the building control the structural and the geotechnical engineer in order
authority. to optimize the foundation with respect to the
It should be noted that for quality control settlements and construction costs.
purposes the DIN requires that designs be subjected
to review by checking engineers who need a
special official registration. However, responsibility
for the design remains with the design engineer. REFERENCES