Structural Analysis of Historical Constructions - Modena, Lourenço & Roca (eds) © 2005 Taylor & Francis Group, London, ISBN 04 1536 379 9

Technologies for the prestressing rings of the Leaning Tower of Pisa

A. Lodigiani ALGA SpA. Milano, lta/y

G. Macchi University, Pavia, ltaly

ABSTRACT: Prestressing techniques proved to be very useful in providing structural safeguard measures to the Leaning Tower of Pisa. In fact , during the II years of stabilization works the International Safeguard Committee used pretensioned cables, strands and wires for the application of active circling forces at the criticaI leveI ofthe First Loggia and to the foundation plinth, both for temporary strengthening measures and permanent interventions.

In some cases innovative implementation techniques had to be studied. In all cases the techniques proved to be efficient and durable and provided solutions of small invasiveness and perfectly reversible when required.

TNTRODUCTION

The lnternational Committee appointed by the Italian Governrnent in 1990 for the "Safeguard and Restora­tion of the Leaning Tower of Pisa" assigned the contract for the works to a Consortium of contractors (including the companies Bonifica, Ismes, Italsonda, Rodio and Trevi) called "Consorzio Progetto Torre di Pisa" .

Among the temporary and permanent structural measures four interventions required the use of the post-tensioning technology.

The first temporary and reversible circumferen tial prestress, made in 1992 on the first Loggia zone, could be defined as an urgent "first aid" intervention, aimed at preserving the life ofthe Monument during the time necessary for the definition and realization of its final stabilization. The second temporary and reversible prestressing work, made in 1993, concerned the con­struction of a temporary ring of concrete segments pressed against the bottom of the Tower surface by means of unbonded internaI post-tensioning tendons.

The beam was designed to support the lead ingots necessary to counterbalance the Tower appropriately during the time necessary to perform the stabilization works on the Tower foundation.

The third work, realised in 200 I, consisted in the replacement of the first temporary ring of 1992 with permanent, but reversible, circumferential pre­stressing belts made of a stainless steel wire.

771

The fourth permanent intervention, made in 200 I, was the permanent construction of a concrete ring under the Catino surface that was blocked to the Tower foundation by means ofpost-tensioned internaI unbonded tendons.

All the strengthening works, according to the Com­mittee specifications, were performed with a proven and reversible prestressing technology that allows the gradual and monitored tensioning of the tendons, the control and adjustment oftheir stress, the replacement ofthe tensile elements.

Furthermore, any deterioration or damage of the Tower surface had to be avoided.

2 TEMPORARY CIRCUMFERENTIAL PRESTRESS OF THE FIRST LOGGIA (1992 Scientific Responsible: Pro! G. Macchi)

The first temporary circumferential post-tensioning intervention was based on a feasibility study of Pro­fessors F. Leonhardt and G. Macchi (November 1990) that made evident the urgent necessity to provide "an active transversal prestress" to improve the strength of the criticaI zone located at the first Loggia and avoid "outside buckling of facing stones" .

The strengthening was made through the installa­tion (fig. I) and jacking at 132 kN of n° 10+8 0.6" hot dip galvanised monostrand tendons encased in a clear

1 J, M [ I

, .,.. I

r--

~ ~611

1,

If

./

"

t Vertical representation for printíng only

Figure 1. Temporary circling.

PVDF tube inert to atmospheric polJution and to UV rays ageing.

The monostrand tendons circling the first Log­gia walJ (na I O units) and the underneath Drum (8 units) were tensioned (fig. 2) and anchored on an intermediate monostrand stressing anchorage.

It was designed by VSL lnternational and could slid on a PTFE pad to pratect the marble surface of the Tower (fig. 3).

The strand tai ls, cut after the tensioning ofthe rings in sufficient length to alJow the control, adjustment and releasing of their force at any required stage, were encapsulated in the galvanised anchor blocks. The entire system was studied in order to assure a long term corrosion protection (~30-40 years). The usual greased/waxed unbonded strand was discarded to eliminate any risk of staining the marble surfaces. The use ofthe stainless steel strand was set aside both

Figure 2. Tendon tensioning.

Figure 3. Intermediate anchorage detail.

for the impossibility of producing the smalJ quantity required and for relaxation reasons. The instalJation, tensioning and releasing procedures were tested, prior to perform the work on the Tower, on a fulJ-scale ring instalJed on an existing concrete water tank hav­ing a diameter similar to the one of the Tower. The monostrand tendons, removed in November 2000 to alJow the final circling with the 4 mm diameter stain­less steel wire, were perfectly preserved as welJ as the anchorages and the PVDF tubes (fig. 3).

3 THE CONCRETE COUNTERWEIGHT-HOLDER BEAM (1993 Scientific Responsible: Pro! G. Macchi)

In 1992 a first check of the monitoring data of the inclination showed that the speed ofthe out-of-plumb movement of the Tower was increasing.

The Committee felt that such a dangerous move­ment had to be urgently stopped in order to alJow time enough for a well based and safe choice of the final stabilization technique.

772

North

Figure 4. Counterweight-holder beam - Plano

n04 6T15 lendons

Figure 5. Counterweight-holder beam - Sections.

Figure 6. Lead ingots.

On this purpose, a proper concrete ring beam (fig. 4), provided with the necessary shorleningjoints, pressed against the Tower bottom like a vice through n04+2 6T 15 prestressing tendons (fig. 5), on which to load the counlerweight, was built.

The ring beam, designed to support a load of 1000 t, was gradually loaded with 600 t of lead ingots (fig. 6).

The concrete beam, shorlly called lead-holder, designed by Bonifica according to the sketches of F. Leonhardt, was built in a way aimed at prevenl­ing any damage to the marble e1adding. The inter­nai unbonded post-tensioning was realised through tendons formed with six galvanised, greased, and individually sheathed 0.6" strand.

Their tails were protected against corrosion with grease and encased in hot dip galvanised caps in order to assure the possibility of controlling and adjusting their tensioning when necessary.

AI lhe end ofthe underexcavation in 2001, the coun­terweight was removed, lhe concrete beam was lolally dismantled and proved to be a perfectly reversible intervention.

4 CIRCUMFERENTIAL PRESTRESSfNG AT THE FIRST LOGGIA WITH A 4 mm DIA. STAfNLESS STEEL WlRE (2000/01 Scientific Responsible: Pro! G. Macchi)

When the stabilisation works (realised as known through the progressive and monitored extraction of soil cores from lhe ground under lhe Northern sector ofthe Calino), were eIose to gel lhe required reduction of the Tower lean, the Committee decided to remove the temporary ringing slrands installed on the firsl Loggia zone and to proceed, at G. Macchi proposal,

773

to the permanent but reversible circling ofthat criticaI zone with narrow stainless steel wire made belts of the same capacity ofthe previous strand rings. The change aimed at less visual impact and better durabil ity. The guideline stated that the circumferential prestress had to be realised with a proven technology using stain­less steel resistant to the atmosphere corrosion, had to be applied gradually and controlled by monitoring, should not damage the Tower surfaces pressed by the tensioned wires, should have a very limited impact on the Monument, had to be revers ible and, during his installation should not damage the Tower neither disturb the peculiar ambient noise coming from the tourist activities in the Piazza.

The technical design and the implementation of the prestressing rings was assigned, through the Consorzio Progetto Torre di Pisa, to ALGA and the Author devel­oped the final design of the intervention and re lated special machinery and managed the installation works with Prof. Macchi 's constant and active supervisiono

The guiding idea was to build a proper winding machine that could continuously lay a stainless steel wire under tension on theTower surface using the avail­able mechanical faci lities and according to the specific characteristics of the Monument. This could be made with the technology utilised along several years in the past for the externaI post-tensioning of the cylindrical silos or tanks.

A cold-drawn (4 mm dia.) austenitic stainless steel wire (NTR 50, equivalent to ASTM A276 S20910), with an excellent resistance to the atmosphere corro­sion (better then AIS! 316L stainless steel) and very high mechanical resistance (its yie ld point is about the double of AISI 3 16L stainless steel), was chosen fo r the circl ing.

Figure 7. Drum circling.

The wire dimension, was defined according to the maximum stressing force suppl ied by the winding machine (1 0 kN), to the mechanical characteristics of the cold-drawn wire (resistance 2: 1400 N/mm2 ,

yield point 2: 1200 N/mm2), to the stressing tension « 700 N/mm2

) applicable to the wire without induc­ing undesirable and invaluable relaxation phenomena in the stainless steel and according to the diameter of the tensioning device too.

The construction of the winding machine was quite complex due to the limitations caused by the architecture of the Tower and its geometrical irregu­larities. A winding machine totally driven by a D.e. motor (4 kW) powered through four sequeI batte ries (24V-200Ah-500A each one) that assure an autonomy of ~3 h at max. stressing force (two sets of batter­ies were provided), was built in order to avoid to disturb the usual buzzing in the Piazza with the contin­uous enraging noise of a gasoline engine. The winding machine (fig. 7) was designed in order to run on a cir­cular steel I-beam placed on the small ring outside the columns (wide 194 mm only at column 15*), to pro­vide the stressing belt on the cylindrical surface (height 390 mm only at column 7*) ofthe fi rst Drum and then to wind the base ofthe first Loggia moving (fig. 8) into the tight corridor between the columns and the Tower wall (wide 893 mm only at column 8). The max. speed ofthe winding machine at ful! traction was ~ 15 m/min and a proper inverter allows to regulate it. An out hanged cabin (fig. 7) dimensioned for two operators, allows to make ali the operations related to the Drum circling. The movement of lhe machine was realised through a set of gears (fig. 9) running on a one inch steel chain located on I-beam (fig. 10) for the Drum winding, and on the Loggia wall for its base ringing.

Figure 8. Loggia base circling.

774

The stressing force (8 kN), regulated by a proper inverter, was applied to the stainless steel wire and regulated through a 500 rnrn dia. pulley whose free rotation around a vertical pin was contrasted by a fric­tion mechanism connected to a force transducer that allows to monitor (fig. li) the stressing force in the wire continuously.

The vertical movement ofthe pulley was finely reg­ulated through an endless screw that allows to lay down the wire spires closed one against the other.

The winding machine was tested, adjusted and cali­brated on a concrete wall sector, on purpose realised in ALGA factory, having the same diameter ofthe Tower Drum. A block of marble similar to the Tower cJadding, was properJy located in the concrete template to test the anchoring ofthe wire cJamps and the deviation pins (fig. 12).

EXPANOER -r=::;:aG;::;=:Y;~~~=:~~Tl OEVICE

1" srEEL CHAIN

Figure 9. MechanicaI detail.

Figure 10. RaiI detail.

Figure 11 . Force monitoring.

GEARS

775

The Tower surfaces wrapped by the tensioned wires, were painted at first, according to the instruction of the experts of the Central Institute for Restoration of Rome, with a protective soluble acrylic resin (Paraloid B-72) that is normally used to consolidate the marble surfaces deteriorated by the atmosphere corrosion.

The influence of the resin solution on the friction of the tensioned stainless steel wire was tested too, with positive results, on the concrete template. The counterbalancing wheels installed on the cabin with mechanical shock absorber to compensate the geomet­rical irregularities and eccentricity ofthe Drum, were covered with a proper rubber outrider for not damag­ing the marble cJadding during their running along the cylindrical surface.

In arder to contain the maximum width ofthe stress­ing belt on the Drum in no more than 145 rnrn (fig. 13) and to make the ornamental band of black marble (fig. 14) that decorates the comice completely vis­ible, the wires were winded spire against spire on four overlapped layers (fig. 13) of 34, 31, 28 and 25 spires respectively, that supplied a total prestressing force of ~91 O kN. The first three layers were painted

Figure 12. Winding machine testing.

Figure 13. Drum bel! detail.

with a wax film (fig. 15) to f ill the capillary voids between the spires and to prevent the crevice corrosion phenomena.

Each layer was anchored to the Tower through inde­pendent stainless steel clamps while stainless steel pins were properly installed to contrast the deviation forces at deviation points (fig. 16) . The wire belt was secured to the Tower with five narrow equispaced stainless

Figure 14. Marble decorations.

Figure 15. Wax painting.

Figure 16. Clamp and pino

plates. The job was made during the first sixteen days of January 200 I under difficult weather conditions.

The circling of the Loggia base was realised, in a similar way but working in a very cramped space (fig. 17), with a belt of ~ I 00 mm only (fig. 18) formed by three layers of stainless steel wires (24 + 21 + 18 spires respectively) that supply a total prestressing force of ~480 kN. The marble surface of the Tower was carefully protected from the direct contact of the steel chain with a wooden belt wrapped by a rubber band (fig. 19).

Figure 17 . Loggia base circl ing.

Figure 18. Loggia belt detai l.

Figure 19. Marble protection.

776

5 THE CATINO BOTTOM RESTORrNG AND ITS CONNECTION TO THE TOWER (2001 Scientiflc Responsibles: Pro! G. Macchi, Pro! L. Sanpaolesi)

The first solution designed by the Committee for the Tower stabilization could be summarised very shortly in the construction of a concrete ring beam (fig. 20) under the Catino surface blocked to the Tower founda­tion through internai bonded tendons. Ten deep ground anchors, of 1000 kN capacity each one, anchored on the North side to the ring beam allowed to apply a defined, adjustable and monitored force to the Tower on its overslope side, to reduce and to control its lean. The project is known as "10 anchors project" too.

North n' 10 deep ground anchors

Figure 20. Calina concrete ring beam - Plan.

7 ~ South side

~~

When the first five segments of the foundation ring in the North edge of the Tower were made, dur­ing the ground freezing for the construction of the first segment on the South (underslope) side, the lean of the Tower increased suddenly (September 1995). The prestressing force in the lead-holder beam was immediately increased, according to the design speci­fications, and the counterweight increased up to 900 t. That unexpected and worrying behaviour, due to an unknown accidental connection between the Catino and the Tower, caused the prompt stop of the works and successively the abandon ofthat project.

In the year 200 I, when the stabilizing operation according to the underexcavation method was ended, the Committee decided to restore the bottom of the Catino and to connect it to the Tower foundation through a concrete ring that includes the five seg­ments already realised, pressed against the founda­tion by means of post-tensioning internai replaceable unbonded tendons.

The intervention had not to remove the Cafino foun­dations on the South side but only its upper layer (~140 mm thick) that was already cracked, had to leave visible the half height of the first step at least, had to provide the correct waterproofing and drainage ofthe water in the Catino area, had to utilise only stainless reinforcing bars.

The design, developed by theAuthor in co-operation with G. Nicolini of ALGA group and the essen­tial contribution of G. Macchi, L. Sanpaolesi and R. 8artelletti, led to the realisation of a thin ring slab, 140 mm thick only, concreted on the South side of the Catino encasing n036 0.62" monostrand tendons positioned on two layers.

The connection ofthe 36 monostrand tendons com­ing from the South side to the 4 upper 12T ISS tendons coming through the five concrete segments from the North side (fig. 21), was realised with two spe­cial trapezoidal steel anchorages 390 mm thick only

6

anchoring stainless steel bars

North side ~

n'18+18 O .62" mon~~

Figure 21. Calino concrete ring slab - Longitudinal section.

777

Figure 22. Intermediate steel anchorage.

Figure 23. Tendons installation.

(fig. 22), resisting the 6000 kN stated ringing force, on which the tendons were anchored.

The moderate verticallifting force (120 kN) caused by lhe misaligrunent of the centrelines of the North and South orders of tendons was balanced by the anchorages mass, by their friction against the Tower foundation, by their connection to the Catino perime­trai wall.

To assure the maximum possible anticorrosion protection, galvanised, waxed and sheathed 15,7 mm

Figure 24. Tendons stressing.

Figure 25. Steel anchorage wax filling.

super strand was used. The unbonded monostrand tendons were furthermore individually encased in additional HDPE continuous tubes, sealed at their extremity to the steel anchorages (fig. 23), to allow their replacement if necessary. The steel anchor­ages were protected from the corrosion with a thick (~1200 I-Lmm) fi lm of epoxy-polyamminic paint that provides their electric insulation too. The ring slab and the five concrete segments were properly anchored to the Catino and Tower foundations with stainless steel bars. The stressing operation was performed through equilibrated, symmetrical and incrementai steps according to a detailed stressing program involv­ing the constant Monument monitoring. Jacks and deviators designed on purpose (fig. 24) were used to operate in the small available anchorages space.

After a convenient control period, the steel anchor­ages were inspected and filled (fig. 25), with a proper anti-corrosion, hydrophobic, insoluble and climatic

778

variation resistant wax, in order to impede any water infiltration.

The realisation does not preclude the possibility to install and connect ten deep ground anchors to the five concrete sectors on the North side, if necessary.

6 CONCLUSIONS

The described realisations show how a prestressing technique, conceived for the post-tensioning ofindus­trial buildings or structures, can be utilised for the strengthening ofMonuments ofhigh architectural and historical value, if properly developed in details and suited to that particular use.

779

REFERENCES

Lodigiani, A. 1993. Prestressing lhe Leaning Tower of Pisa. VSL NEWS, /993 nO 2

Macchi, G. 1993. 11 viaggio di ritomo dalla Torre di Pisa. ANA r KH, n° 4 - December /993

Macchi, G. 1996. Leaning Tower of Pisa. Structural prob­lems. Historische Bauwerke, Internationale Tagung des SFB 315, HeI[ 14/1996, University of Kar/sruhe

Macchi , G. 200 I. La torre di Pisa : ritomo ai futuro. Problemi strutturali e consolidamento. Restauro, n°. 158, 200/: 45- 94

Macchi , G. , Marioni , A. & Lodigiani , A. 1993. Precom­pressione monotrefolo non aderente ad a lta durabili -tà. GiornateAICAP'93, Pisa 3- 5 Giugno: 563-568