You are on page 1of 10

Forensic structural engineering: Case study assignment (Group 11)

Tower of Pisa

Carlos Azua Gonzalez


Miguel Chavez Abril
Juan Chavez Olalla

4490894
4449185
4503252

Introduction
The Leaning Tower of Pisa is perhaps one the most well known cases in Forensic
Engineering among the category of structure-ground interaction related failures. The
documentation of the case presents two extremes: documentation during construction
appears to be scarce as it was obtained by indirect means [1], whilst the potential failure
analysis and remediation has been well documented (good examples in chronological order
[2], [3], [4], [5] among others). The following sections treat possible consequences which
triggered the instability of the Tower and how this case in Forensic Engineering can leave us
some lessons for new designs.
In the literature, complex constitutive models are commonly used to describe the non linear
soil behaviour [6], and the case of this Tower is not an exception. One can find authors
explaining the complex behaviour of the leaning Tower by means of Finite Element
calculations ([2], [7], [8]). In the sake of simplicity, however, this essay refers to linear models
to test presumable hypotheses.
Description of the structure and its behavior
The Leaning Tower of Pisa is about 60 m high and the shallow foundation (plinth) is 19,6 m
in diameter. The structure is a hollow cylinder which walls are built with rubble and mortar;
moreover, the staircase passes through the walls forming a spiral. The inner and outer
cladding is composed of marble. Figure 1 shows a rough cross section of the Tower. The
total weight of the structure is 141,8 MN ([1],[7]).

Figure 1. a) Cross section of the Leaning Tower of Pisa. b) Ground profile


Source: Burland, et al (2003).
The soil profile is divided in three main strata (Figure 1) described as follows. Stratum A is a
10 m thick layer of sandy silts and clayey silts. The last two meters of layer A consists of
dense fine sand. The horizontal variability of this stratum is believed to be the cause of the

inclination of the tower. Indeed, the material in the south of the tower is finer than the
material in the north which means that it is more prone to deformation. Stratum B consists of
marine clay up to 40 meter depth. Finally, Stratum C is a dense sand [7].
Description of the damage
Leaning of the tower started in the early years of construction, as observed in masonry
corrections of the lowest floors, then by 1990 the inclination of the foundation was 5.5
degrees. In order to understand the behavior of the Tower, it is necessary to know the
history of construction of it which is divided into 3 stages. First, from 1173 to 1178 the
construction started and reached the fourth cornice when it stopped. Second, from 1272 to
1278 the second stage started and reached the seventh cornice. Finally, from 1360 to 1370
the bell chamber was constructed. As a result of the staged construction of the tower, it did
not fall down. Indeed, around 100 years of consolidation took place between stages, so
water pressure in the soil was dissipated leading to an increment in soil strength ([1], [7]).
Although the structure shows some cracks and deterioration, it has enough bearing capacity
to withstand the working loads applied to it, so leaning of the tower is not related to
deficiency of the structure, but rather to deficiency in the soil. Moreover, deficiency in the soil
foundation mentioned before is caused by deformability (stiffness) rather than by strength of
the foundation material which is characteristic of fine deposits such as the ones found under
the Pisa Tower. This is evidenced by several physical and numerical experiments [7] in
which a toppling failure mechanism (Figure 2a) occurs rather than a general failure
mechanism (Figure 2b). In other words, the failure is localized in a small zone and not
moving the entire soil mass beneath the footing.

Figure 2. a) Local shear mechanism (gradual toppling). b) General failure mechanism.


Source: Coduto (2001)
Listed hypothesis
Masonry degradation is a common issue regarding tower collapse of such ancient
constructions [1]. The circular walls from which the tower was built, were weak from an aging
process (having significant cavities) which would increase the likeability of a collapse. A
main key aspect regarding the masonry element, is the fact that limestone was the main
tower constituent (mortar). Calcareous material such as this are to prone to weathering
under harsh external conditions. Some sections from this circular wall were thinner than
others, generating a more likely unstable framework (non uniform cross section of tower).
However, the Pisa Tower does not show evidence of failure due to structural capacity

although it shows some deterioration in the walls which could worsen the stability of the
tower when a higher tilting is reached, so the stresses of the structure reach a limit state and
could collapse (fragile structural behavior).
External conditions such as wind force and seasonal changes in water table affect the
stability of the tower. First, wind could have affected the structure due to its cyclic character
causing fatigue of the soil even though the structure is not slender (it is both rigid and short).
Second, a fluctuating water table affects the stress state of the soil and consequently its
strength. However, including these effects for calculating the leaning stability of the tower
implies a numerically more complex model which is not the aim of this study. Also, a soil
structure interaction model may be used, so that the interaction soil-structure and its effects
are considered such as radial tensile stresses generated in the tower structure and
subsequent deterioration, but as mentioned earlier the aim of this study is to provide a
simplified calculation method of stability.
Provided that the main materials used for the tower construction were load-bearing masonry,
cut stone and white marble, the weight was so large (approximately 14500 tons) that the soft
foundation was unable to support it. This is the first evident reason as to why the tower
started tilting towards a weaker soil zone (South). It must be noted that if such large
consolidation periods (construction was periodically stopped at many different stages) were
not allowed to happen, the collapse of the structure would have been unavoidable. A total
slow construction time of 200 years was probably the main reason why the structure was
able to survive over all this time. The strength gained during the years of consolidation made
the formerly weak soil able to prevent soil failure and general structural collapse. The motion
generated due to leaning instability is progressive rather than an imminent bearing capacity
failure and it occurs due a critical eccentricity acting on a relatively tall structure. This
eccentricity would thus be translated in the overturning moment that leads to the potential
collapse that was prior generated by small increases in the inclination. The potential failure
of this structural system will be focused in terms of the leaning instability mechanism due to
the geotechnical conditions at the foundation site.
Testing the most presumable hypothesis (model)
In order to understand the cause of the tower tilting, a consolidation model is used for the
soil beneath the foundation in both sides of the Tower (North and South). As mentioned
before, the soil in the North is coarser than the soil in the South. The magnitude of the effect
of the grain size is explained by means of the consolidation period which is the time that
water takes to reach an hydrostatic state after applying an external load to a saturated soil.
The period of consolidation may vary between a couple of months in rather coarse soils to
hundreds of years in very fine soils. A key parameter to compute the consolidation period is
the permeability of the soil. The permeability of the clayey silt in the South is estimated in the
order of 10-9 m/s while the permeability of the sandy silt in the North is estimated in the order
10-4 m/s [8]. As shown in figure 1, Horizon A corresponds to a low permeable soil in the first
10 metres below the foundation followed by a 30 m thick clay layers which are impermeable.
As a result, the parameter h in equation 1, which is the length of the flow path, is equal to 10
m. One can refer to [10] for a detailed explanation of the theory of consolidation:

T = 2 h2 / c

Equation 1

In the aforementioned equation T stands for the period of consolidation and c for the
coefficient of consolidation. Fine grained material content does not greatly alter the
compressibility (inverse of the stiffness) behaviour and unit weight of the material, so they
are considered constant in both sides of the tower. Thus, the ratio of the consolidation
periods depends only on the ratio of the permeabilities, so it is concluded that the period of
consolidation of the clayey silt in the south is 105 times larger than that for the sandy silt in
the North based on equation 1 and equation 2.
c = k Eoed /

Equation 2

In the later equation, k stands for soil permeability, Eoed for the soil constrained modulus
(inverse of the compressibility) and for soil unit weight. One outcome of this analysis, is
that increments in groundwater level affects the stress state of the soil in the South for longer
time than in the North while increase in strength is faster in the North than in the South for a
reduction in water level. Additionally, a critical situation is expected during construction when
large load increments are expected or even during the first years of operation rather than the
post construction period.
Shallow foundations such as that used for the Tower are commonly critical to variations in
pore water pressure in terms of stiffness and strength. In fact, one can refer to complex
theoretical soil models that capture the nonlinear stress-dependent soil behaviour, perhaps
the most famous one is the Duncan-Chang model [6]. Nevertheless, less sophisticated
models as the one presented by Burland et al [7] may also perform well to disregard
hypothesis in a Forensic analysis. The model is simplified as an inverted pendulum with a
concentrated mass W (Figure 3). The stability of the system is provided by the elastic hinge
in the base which simulates the foundation. The model is an uncoupled soil-structure
interaction model meaning that rotation and vertical deformation are independent of each
other. The system of equations that describes the behavior of the model is presented as
equation 3.

Equation 3

Figure 3. Simplified model of the Leaning Tower of Pisa


Source: Burland et al, (2003)
As shown by Burland et al [7] the factor of safety against rotation is calculated assuming the
soil under the Tower as an elastic half space (Equation 4) or by means of Winkler springs
(equation 5).

Equation 4

Equation 5
Accordingly, one can find in the literature that for the half space approach the term E / (1-2)
is estimated in 2,85 MN/m2. It must be noted that the irregularities in the geometry of this
structure were significant, so the center of mass is estimated at a 40% (22.5 m) of its height
(60 m) [7] which yields a factor of safety of 1,12 for elastic half space [7]. On the other hand,
for the Winkler approach the subgrade reaction coefficient k is calculated as the pressure
per unit settlement. As a result, an average settlement of the Tower which is estimated in 3
metres so, k = 141,8MN/ (0,25xx(19,6m)2) / 3 m 0.157 MPa/m [7]. Therefore, Winkler
springs yield a factor of safety of 0,36. Obviously, the Tower had not collapsed into the
ground by overturning which yields to disregard the Winkler foundation approach. Hence, a
realistic factor of safety appears to be slightly higher than 1,0 which confirms that the Tower
was in neutral state during the 90s and that the problems faced were related to a low
stiffness of the ground.
Consequences of failure
The results of this tilting instability mechanism (relatively close to a failure state at short-mid
term) was luckily only a socio-economical issue instead of fatalities or accidents. The first
aspect (social) was evident since the continuous tilting of the tower of the years leaded to a
safety unrest for the people living, working or passing by the structure with the fear of never

knowing when it it was going to inevitable collapse. The fact that the first approaches for
finding a solution (preliminar remediations) proved to be no efficient and even negative did
not help to mitigate the social unrest at all. The second main issue regarding this event were
both the preliminary adopted measures and even a closure to the public at a particular
period (1990-2001). By using counterweights at the entire perimeter base of the tower, this
generated an unpleasant sight for this key touristic attraction of Pisa. If no other measures
were taken, this would have caused perhaps a significant decline in the income generated
by tourism for this city. This concern came into reality after the tower was closed to the
public for more than a decade. The closure of the leaning tower undoubtedly brought
economical repercussions to Pisa.
Remediation
Long after the towers construction, the first modern solution was applied in 1935. This
consisted in soil-improvement technique by injecting grout underneath the structure. The
result of this attempt was to slightly increase the inclination angle, proving to be a negative
solution.
The first positive action for a remediation attempt was in 1992 regarding the masonry
element. From this, the first-storey was braced with steel tendons to relieve the strain on the
vulnerable masonry. At 1993, 600 tons of lead ingots were stacked around the base of the
north side to counterweight the given tilting. This post-tensioned removable concrete ring
casted around the base of the tower aided in reducing the already existing overturning
moment by a 10 percent approximately. This short term solution was still not sufficient to
revert the potential failure of the tower of Pisa, so other options were sought. [12]
Socio-political pressure occurred due to non-aesthetic looks of remediation solutions. In
1995: 10 underground steel anchors were used to yank the tower northwards in a non visible
way, but, this proved to be critical and brought the tower closer to collapse. These anchors
were going to be installed 40m deep from the tension cable placed at the towers base. The
groundwater tended to be so high, that it was frozen to not allow it to affect the excavation
due to flooding but this caused an expansion of the water which pushed upwards the tower.
After thawing, large voids generated further settlements. As a short-term way to mitigate this
generated issue, additional 300 tons were desperately added to the northern side as no
other immediate solution was found. [11]
On 1999, a soil removal operation at the northern side was undertaken with the use of
drilling systems of hollow-stemmed continuous flight auger, from which 77 tons of soil were
removed and thus the tower was straightened by 44cm. A better tilting recovery could have
been obtained but in order to maintain the esthetic view of the leaning tower, it was
preferred to leave it with the same state as the one observed at 1838 (previous tilted state).
The main quality of this technology implemented was that a minimum disturbance of the
surrounding ground was obtained during and after the extractions. The generated cavities
during the processes at the silty soil horizon A (Northen critical zone) were closed gently and
this as well made possible a full scale extraction over a significant area. After the
excavations were done, the cavities were filled with betonytic grout. This slurry-like material

is characterized for its very low permeability. The equipment employed for the main solution
of this geotechnical problem is seen in figure 4. [13]

Figure 4. Tool for the soil extraction at the Northern area (Horizon A).
Source: Burland et al (2000)
The second permanent solution that was implemented arose from the discovery of a 0.8m
thick cement-conglomerate ring at the bottom of the catino located around the base of the
tower. The high quality of this ancient concrete ring was connected to the masonry plinth
foundation by using a stainless steel reinforcement as seen in figure 5. Additionally, these
elements were reinforced by circumferential post-tensioning bars. The final result of this
action was to enhance the effective area of this shallow foundation and thus the factor of
safety against leaning instability improves as well. [13]

Figure 5. Steel reinforcement and post tensioned bars anchored to the Towers foundation.
Source: Burland et al (2000)
A last attempt to solve definitely at a long term this potential collapse was by introducing a
new drainage system right beneath the critical northen area. Monitoring from 2003-2009
showed that no further displacements were observed. This third permanent solution
consisted of three wells sunk at the northern side with radial pseudo-horizontal drains
running through beneath the catino of the structure. A significant drop in pore water pressure
was monitored at the critical seasonal fluctuations. This mechanism is presented in figure 6.
After the last solution, the remediation teams concluded that approximately for 200 more
years the tower will be able to remain stable. [13]

Figure 6.- Solution of radial drainage system with 3 wells at the northern side.
Source: Burland et al (2000)
Lessons learned from the case study and conclusions
The first aspect to be commented is regarding the construction approach and criteria
adopted at that time ( a given fundamental mistake). Not having prior information or
knowledge from the subsurface at which a structure of significant importance would be built
is a main first mistake done in this practice. It must be mentioned that the field of Soil
Mechanics have been developed increasingly in the last century since 1925 with the
exemplary work of Karl Terzaghi [14], which resulted in constructions not based on the
effective stress principle before this series of publications. The second and probably most
important one was the fact that even after already experiencing a progressive tilting at the
initial construction phase ( first two floors), the construction was not cancelled or reallocated.
By selecting a more suitable place this entire problem would have been avoided.
Another crucial aspect regarding structures built in soft soils ( highly compressible) is the
consolidation time for such soft material in addition to the total construction time. Allowing or
not allowing the soil to consolidate while construction is taking place will have heavy impact
on the stability of such structure i.e. a long enough period of construction is desirable to
minimize the effect of pore pressures in the strength and stiffness of the ground.
The potential failure of this ancient structure of great importance was able to analyzed
through simple elastic approaches as means to corroborate that without the applied
solutions, failure would have been imminent at mid-long term.
It should be mentioned that the analysis of soil-structure interaction as referred in this essay
is not fully coupled and a consolidation analysis was performed only qualitatively. As a
result, we have roughly proved the soil stability under a theoretical framework and related
the problems to a deficient stiffness of the ground. Further analysis is needed to provide a
better understanding of the main mechanisms that triggered the progressive inclination of
the Tower.

References:
[1] Marchi M (2008): Stability and strength of leaning towers. PhD Thesis Politecnico
Universita Degli Studi di Parma - Universita Di Bologna.
[2] Burland, J. B., and Potts, D.M. (1994): Development and application of a numerical model
for the Leaning Tower of Pisa, IS Prefailure Deformation Characteristic of Geomaterials,
Hokaido, Japan, 2, 715-738.
[3] Pepe, M. (1995): La Torre Pendente di Pisa. Analisis teorico-sperimentale detla stabilita
dellequilibrio, PhD Thesis, Politecnico di Torino,265.
[4] Potts, D. M. and Burland, J.B. (2000): Development and application of a numerical model
for simulation the stabilisation of the Leaning Tower of Pisa, Proc. John Booker Memorial
Symp. on the Developments in Theoretical Geomechanics (eds. By Smith and Carter),
Sidney, 737 - 758.
[5] Jamiolkowski, M.B. (2001): The Leaning Tower of Pisa: end of an Odessey, Terzaghi
Oration Proc. XV ICSMGE, Istanbul, 4, 1979 -2996
[6] Duncan , J.M. & Chang, C.Y. (1970): Nonlinear analysis of stress and strain in soil, J. Soil
Mehc. Found. Div., ASCE 96,1629-1653
[7] Burland, J. B., Jamiolkowski, M. , Viggiani, C.(2003): The stabilization of the Leaning
tower of Pisa. Soils and Foundations. Japanese Geotechnical Society. Vol. 43, No. 5, 63-80
[8] Vermeer, P.A., Bonnier, P.G.(2002): Creep Analysis of the Leaning Tower of Pisa. Report
by Stuttgart University and Plaxis B.V., The Netherlands.
[9] Coduto, D. (2001): Foundation Design. Principles and Practice.Prentice Hall. ISBN:
0-13-5897068
[10] Verruijt, A. (2010): Introduction to Soil Dynamics. Delft University of Technology, The
Netherlands. Theory and Applications in Porous Media. Vol 24. Springer. DOI
10.1007/978-90-481-3441-0
[11] Burland J.B. et al (2009). Leaning tower of Pisa: Behaviour after stabilization
operations. International Journal of Geoengineering case Histories Vol. 1, Issue 3, p 156.
[12] Burland, J.B., Jamiolkowski, M.B. and Viggiani, C. (2003). The stabilisation of the
Leaning Tower of Pisa. Soils and Foundations Vol. 43, 5, pp. 63-80.
[13]Burland, J.B., Jamiolkowski, M.B. and Viggiani, C. (2000). Underexcavating the Tower of
Pisa: Back to the future.GEOTECH-YEAR 2000, Developments in Geotechnical
Engineering, Bangkok,Thailand, Balasubramaniam,A.S. et al. Eds, pp. 273-282.
[14] Terzaghi, K. (1925): Erdbaumechanik auf Bodenphysikalischer Grundlage (Leipzig F.
Deuticke). Principle of soil mechanics Eng. News Record (1925), a series of articles.

You might also like