880
Proceedings of the 18
th
International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013
approach for practical purposes by Kany (1974). This approach
provides realistic settlements also at the edges of foundations.
But it is not able to describe more complex soil behaviour such
as hardening or softening (Muir Wood 1990).
een flexible and stiff
behaviour of the foundation system.
s
= 1/12 · E /E ·(d/L)
3
(1)
tructu
: Foundation length [m]
din
alculated
behaviour (e.g. Leussink et al. 1966).
whole centrifuge test program
can be found in Arnold (2012).
pringman et al. 2002). The test setup is given in
figure 3 and 4.
Figure 3: Setup of the centrifuge test (Laue and Arnold 2008).
me of the test setup in the drum centrifuge (Laue and
no pressure. For these
conditions, a flexible behaviour can be observed with maximum
str
s distributed near to the column.
um plate at 50g (Arnold and Laue 2009); right hand side:
Resolution of the tekscan measurements given in kPa (Arnold and Laue
aves stiffer as four unloaded
walls are placed on top of all sides. Details on this test can be
found in Arnold and Laue (2009).
ted at 50g. Protoype load: 10925 kN. Prototype
ttlement: Approx. 220 mm (Arnold and Laue 2009). The resolution is
eq
ituated in the area of the load-bringing
column and the distribution is more uniform over the whole area
1.2
System-stiffness after DIN-code
The DIN-code 4018 (1981) defines a system-stiffness (eq.1),
which allows distinguishing betw
K
b s
K
s
: System stiffness [-]
E
b
: Stiffness of the foundation s
re [N/m
2
]
g on the c
E
s
: Stiffness of the soil [N/m
2
]
d: Foundation thickness [m]
L
The behaviour of the foundation is distinguished (Meyerhof,
1979) depending on the value of K
s
with K
s
= 0 representing
flexible, 0.001 < K
s
< 0.01semi-flexible, 0.01 < K
s
< 0.1semi-
stiff and K
s
> 0.1 stiff behaviour. This allows choosing the
stress distribution for design depen
2 CENTRIFUGE MODELLING
Details about the centrifuge modelling can be found in
Schofield (1980) and Laue (2002). The centrifuge tests, which
are presented in this contribution, have been conducted in the
drum centrifuge at ETH Zurich (Springman et al. 2001).
Detailed information about the
2.1
Centrifuge test on a flexible raft foundation
80 Centrifuge tests were conducted for studying the loading
behaviour of flexible raft foundations (Arnold 2012). The stress
distribution between raft and soil was measured with tactile
pressure pads (S
Figure 4: Sche
Arnold 2008).
The loading of these tests was conducted on a 4 mm thick
square aluminium plate as foundation with a side length of 11.2
cm under an enhanced g-level of 50. The model foundation
represents a prototype foundation with a side length of 5.6 m
and a thickness of 20 cm. Figure 5 shows the measured stress
distribution for a load of 4.25 kN (equivalent to a prototype load
of 10625 kN) and a settlement of 5 mm (equivalent to a
settlement of 250 mm at prototype scale). The white areas show
the highest pressure, black areas show
esse
Figure 5: Left hand side: Stress distribution under a 4 mm thick
alumini
2009).
2.2
Centrifuge test on a stiffened raft foundation (4 unloaded
walls)
Figure 6 shows the stress distribution under a foundation
stiffened by four unloaded walls. The stiffened foundation has
also a thickness of 4 mm but beh
Figure 6: Stress distribution under a 4 mm thick aluminium plate
stiffened with 4 unloaded walls situated on the four edges of the
foundation. Test conduc
se
uivalent to Figure 5.
The stiffer stress distribution can be seen in Figure 6. Less
clear peak pressure is s
of the foundation slab.
3 FULL SCALE MEASUREMENTS
Detailed information about the full scale measurements is given
in Arnold and Laue (2010) and Arnold (2012). Two different
buildings were equipped with oil filled pressure plates
manufactured by Gloetzl (Schmidt 1991) to gain some
information about the load extent on the raft and the stress
distribution between raft and adjacent soil. One building is
