edificio isolato tis.pdf

8/14/2019 edificio isolato tis.pdf

http://slidepdf.com/reader/full/edificio-isolato-tispdf 1/8

7 th

International Seminar on

Seismic Isolation, Passive Energy Dissipation and Active Control of Vibrations of Structures

Assisi, Italy, October 2-5, 2001

AN EXTENSIVE APPLICATION OF SEISMIC ISOLATION TO PRIVATEBUILDINGS IN ITALY

R. Marnetto

R&D Manager of T.I.S. S.p.A, Rome, Italy

www.tis.it

C. Castino

Technical Department of T.I.S. S.p.A., Rome, Italywww.tis.it

M. DolceUniversity of Basilicata, Potenza, Italy

[email protected]

ABSTRACT

A widespread application of seismic isolation to private houses in Calabria, a highly seismicregion of Southern Italy, is in progress. This initiative has two interesting aspects. It is, on one

hand, the most extensive application to private buildings in Italy (36 two-house buildings in

the first phase) and the choice of isolating is, on the other hand, only due to a decision of the

owners, without any public incentive.

The houses are all equal and are part of a project of urbanisation of a new area. Each build-

ing has only three stories. Under this condition the cost of isolation has, usually, a high

impact, higher than for bigger structures. Nevertheless, owners decided on purpose to invest

on a better protection of their houses, to get the no damage condition under the design earth-

quake. The present work shows the design approach, the choices made to optimise the

isolation system, the seismic response of the structure. A particular attention is devoted to thecomparison between the conventional and the base isolation solution, regarding both the

economical and the technological aspects.

1. INTRODUCTION

In Italy, though being a highly seismic country, the diffusion of innovative technologies for

seismic vibration control is progressing very slowly. The reasons can be identified in three

main complementary problems, which makes the process of technological renovation hard.

On one hand there is the prudence, perhaps excessive, of authorities in making the currentuse of seismic isolation easy, by preparing simple design tools (guidelines and norms), with-

out the needs of further submission, which often discourage for the time delay and the conse-

quent costs in the realisation process. On the other hand there is low competence and attitude

of designers with respect to a design process involving the application of new technologies for



the control of seismic response, besides the needed care for details (joints, systems, etc.) that

this implies. Last but not least there is the practically total ignorance of the final users with

respect to the protection potentials of the new technologies. This aspect makes the demand

from the base lacking, which impedes the activation of the virtuous process of diffusion of a product, when it meets the expectance of the market, and which can, alone, push for a fast

overcoming of the previous two points. Nevertheless, somewhat seems to move in the right direction. The joint action of industry, in

this specific case TIS SpA, and of the University, namely of Basilicata, seems to have

stimulated the interest of some designers after some short courses on the new technologies for

the seismic protection of constructions. The consequence is that ing. Panzarella has proposed

to his customer, the “cooperativa MIVAR”, the realisation of 36 two-house buildings, and

then, perhaps, of 19 more, protected with seismic isolation systems. The private customers,

without any economical incentive, but just being convinced of the overall benefit in

drastically reducing the seismic risk for events which are felt as surely happening in theirlifetime, decided to undergo greater construction costs.

The activity of transforming a conventional fixed base design in a seismic isolation design

is still in progress, being carried according to the current Italian Guidelines (C.S.LL.PP.,1998). This experience will be described in the present paper. The attempt of optimising and

specialising the seismic isolation conception and design to the specific requirements of the

Guidelines, but also to the peculiarities of the problem at hand, will be carefully examined.

Actually, as many of the designers of seismic isolated structure know, there is not one

universal solution for any problem, considering the different economical and behavioural

features of the different seismic isolation devices and systems. A great design effort has been

made to get the right compromise among the different architectural, economical and safetydesign inputs. It seems opportune that the problems met and the proposed solutions be shared

with other designers and experts.In the paper the main design problems and solution, the choice and optimisation of the

isolation system, which makes use of a new purposely developed isolation device, as well as a

comparison of the seismic response of the structural system with and without seismic

isolation will be presented and discussed.

2. DESIGN PROBLEMS

An architectural view of the building is shown in fig. 1. One of the main objective of the

design was to get an isolation system arrangement that would not change the architecturalfeatures of the building. This was achieved by

limiting the encumbrance of both the devicesand of all the structural arrangements needed.

In fig. 2 there is shown an axonometric view

of the structure and of the structural arrange-

ment of the isolation system on top of the first

story. The final version of the isolation

system is made of steel-teflon sliders allowingfor up to 200 mm displacements, one under

each column, and four energy dissipatingdevices, suitably located in order to avoid any

global torsional effect. As described in the Fig. 1 – Perspective view of the building.



following paragraph, the four devices are deputed to restore the initial configuration of the

structure, under low-medium intensity earthquake (DLS), and to dissipate energy, under high

intensity earthquakes (ULS). The position of the devices is such that they will transmit their

forces to the substructure, directly on the retaining walls or very strong columns, along the

perimeter of the building. The columns of the substructure, therefore, must only bear verticalloads, which will act eccentrically when the superstructure is displaced under an earthquake,

and the friction forces in the sliders. A light bar system connects the top of the columns of thesubstructure, in order to avoid any stress concentration in the internal columns, in case of

anomalous increased friction coefficient in any of the sliders.

3. SELECTION AND OPTIMIZATION OF THE ISOLATION SYSTEM

The optimisation of the isolation system starts from the choice of the most suitable system toachieve the performance and economic objectives, taking into account criteria and require-ments specified in the reference guidelines (C.S.LL.PP., 1998). The comparison has been

carried out on both the outcome of simplified design procedures and the results of simulationanalyses of single DOF systems. The SDOF systems have the estimated mass of the super-

structure, i.e. 590 t, while the isolation devices are modelled as combination of non linear ele-

ments simulating the actual non linear behaviour of devices. Three isolation systems have

been considered.

1) Sliders combined with high damping rubber bearings,

2) Sliders combined with elasto-plastic steel elements,

3) Sliders combined with hybrid rubber-steel elasto-plastic devices with low initial stiffness.

For the three systems similar design conditions have been assumed, relevant to anintermediate soil profile (type B), whose normalised spectrum in the range of period of

interest is given by (C.S.L.L.P.P., 1998):

Se(T) = sηηηηββββ0 (TC/T)k1 for T>TC

s accounts for the stratigraphic and geotechnical conditions of the site (= 1 for soil B),

ηηηη=3

7 2/ ( )+ ξ ≥≥≥≥ 0,7 accounts for equivalent viscous damping ratio ξ other than 5%, ββββ0 dynamic amplification factor of seismic response = 2.5 for soil B;

TC period separating the spectrum branches from each other (0,6 sec. for soil B);

k1 exponent used to define the descendant branch of the response spectrum = 1.

Being the building in a highly seismic site, the assumed PGA is equal to 0.35g for the

Fig. 2 – Axonometric view of the entire structure (left) and of the isolation system (right).



Ultimate Limit State (ULS) and 0.10g for the Damage Limit State (DLS).

As there are serious technological difficulties in realising a HDRB system, due to the lowweight of the building, this system as been considered as a combination of sliders and rubber

isolators (Braga et al. 2001). Assuming 2 secs. vibration period, 10% equivalent damping andno friction in the sliders, the following quantities are evaluated:

• Secant stiffness K iso = 5817 kN/m• Spectral acceleration at ULS Se = 0.22g

• Maximum displacement at ULS d2 = 0.22 m

• Maximum isolation force (base shear) at ULS Fiso = 1274Considering that, according to the guidelines, the base shear can be reduced by a behaviour

factor 1.1 and that the maximum displacement in the two directions must be vectoriallycombined, the following data are obtained:

• ULS base shear force S baseD = 1158 kN

• Design displacement of the isolation system dE = 0.31 mThe second isolation system, made of steel based energy dissipating devices, must fulfil the

no yielding condition of the Guidelines under DLS actions, i.e. for 0.1g PGA. Due to the high

initial stiffness of the devices, the initial period of the structural system is less than 0.6 sec.Therefore one obtains:

Spectral acceleration Se(T<0.6 sec., 5%, soil B, PGA=0.10g) = 0.25gAssuming that the base shear, due to the modal combination, is about 75% of the product of

the mass by the spectral acceleration, the design force of the isolation system, which is alsothe minimum allowable yielding force, can be estimated as:

Fyiso = 1106 kNTo evaluate approximately the base shear, friction must be taken into account, equal to 3%

of the weight, obtaining:S base = 1106 + 174 = 1280 kN at DLS

At the ULS the base shear force is increased by the hardening of the devices, which resultsin about 20%, and reduced by the behaviour factor equal to 1.1, to get:

S base = (1106*1.2 + 174) / 1.1 = 1364 kN at ULSSince the structure must fully keep its elastic state, the DLS verification is in general more

conditioning than the ULS verification. By comparing the forces, it also come out that thesecond isolation hypothesis is more onerous than the first one.

The third isolation system is made of sliders under each column, combined with four newdevices, patented by Dolce and Marnetto, and manufactured by TIS SpA, especially

conceived to fully satisfy the DLS and the ULS conditions of the Guidelines. It joins thedesirable properties of rubber isolators (low stiffness and good energy dissipation capacity)

for the DLS condition and of

the elasto-plastic devices (limit-ation of the force transmitted tothe superstructure). To combine

the two properties, the newdevice is made of two devices

arranged in series (i.e. one onthe other), as shown in fig. 3. In

this case the elasto-plastic de-vice is made of U-shaped steel

elements (Dolce et al. 1996). Infig. 4 there are shown the cyclic Fig. 3 – hybrid rubber-steel device.



behaviours of the two parts of the hybrid device and its overall behaviour.

Due to its general feature and the related possibilities of calibration of its components, thehybrid device can satisfy the following design objectives:

• DLS: reduce structural stresses by elongating the initial period,

• ULS: limit the maximum base shear, approximately within the value reached at the DLS.

Obviously these two conditions shall be balanced with the isolation displacements that can beaccepted for the two limit states.With the help of some design formulae, based on the equivalent linearization technique, the

optimisation of the devices, taking intoaccount also the geometrical constraint of the

structural system, leads to the followingapproximate design quantities:

T1 = 1.0 sec.F1 = 784 kN

S base = 784 + 174 = 958 kN at DLSd1 = d1a + d1b = 25.3 + 8.4 mm = 33,7 mm

S base = (784*1.2+174)/1.1 = 1115 kN at ULSd2 = d2a + d2b = 31 + 70 = 101 mm

If the shear strain of rubber is limited to100% for the maximum ULS displacement,its thickness shall be of the order of 30 mm.

Further reduction of the base shear can be,obviously, obtained by increasing the period

T1. The choice made is, however, justified bythe need to limit vertical force eccentricity, and then moments, on the substructure columns,

as well as the size of the sliders. Moreover, it is of little use to reduce stresses in thesuperstructure too much, due to the minimum code requirements on geometry and reinforce-

ment of structural elements, that, otherwise, governs the structural design.

4. NUMERICAL SIMULATIONS ON SDOF MODELS

To evaluate the performances of the three considered isolation systems, numerical non linear

dynamic analyses have been carried out on SDOF systems, whose force-displacementcharacteristics describe adequately their cyclic behaviour.

Rubber with

frictionT=2 sec.

Rubber

withoutfrictionT=2 sec.

Steel

with friction

Steel

withoutfriction

Hybrid

with friction

Hybrid

withoutfriction

Daverage 134 230 29 38 101 117

Ddev 17 16 6 9 23 15

Dm+dev 151 246 35 47 123 132

ddesign 189 325 53 70 185 198

Faverage 956 1305 1443 1327 1104 963

Sbase 869,5 1186,1 1311,4 1206,4 1004 875

Table 1 – Main results of the numerical simulation analyses.

F

dd1b

F1

d1d1a

d1=d1a+d1b

Fig. 4 – Cyclic behaviour of the hybrid devi-

ce, made of rubber (a) and steel (b) parts.



The proposed hybrid solution is the best compromise in terms of base shear and displace-

ment. The rubber solution is the most demanding in terms of displacement (325 mm), whilethe elasto-plastic steel solution is in terms of force. The hybrid solution gives about 200 mm

design displacement (117 mm mean value) and about 1000 kN base shear. In fig. 5 there arereported some diagrams relevant to the hybrid solution, considering or not friction of sliders.

5. SEISMIC PERFORMANCE OF THE ISOLATED BUILDING

No optimisation of the isolated structure has been made up to now. Therefore the geometry of

Total force - total displacement

-1500-1000

-500

0

50 0

1000

1500

-0,10 -0,05 0 ,00 0,05 0,10d (m)

F

( k N )

Total displaceme nt - time

-0,10

-0,05

0,00

0,05

0,10

0 5 10 15 20 25 30 35t (sec )

d ( m ) )

Total force - rubber displacement

-1500

-1000

-500

0

50 0

1000

1500

-0,10 -0,05 0,00 0,05 0,10d (m)

F

( k N )

Total force - steel displacement

-1500

-1000

-500

0

50 0

1000

1500

-0,10 -0,05 0,00 0,05 0,10d (m)

F

( k N )


-1500

-1000

-500

0

50 0

1000

1500

-0,10 -0,05 0 ,00 0,05 0,10d (m)

F

( k N )

Total displaceme nt - time

-0,10

-0,05

0,00

0,05

0,10

0 5 10 15 20 25 30 35t (sec )

d ( m ) )

Total force - rubber displacement

-1500

-1000

-500

0

50 0

1000

1500

-0,10 -0,05 0,00 0,05 0,10d (m)

F

( k N )

Total force - steel displacement

-1500

-1000

-500

0

500

1000

1500

-0,10 -0,05 0,00 0,05 0,10d (m)

F

( k N )

Fig. 5 – Behaviour of the hybrid isolation system considering (top) and not considering

bottom friction, under accelero ram 2.



the columns and beams are kept the same as they were initially designed under the fixed basehypothesis. The seismic behaviour of the structural system has been checked through a set of

non linear simulation analyses, carried out on a three-dimensional refined model, according to

the specifications of the Guidelines. The structural elements have been modelled with linearelements, while the isolation devices have been modelled with non linear elements, and thesliders as gap elements with friction. About 2% modal damping has been considered for the

second modes, i.e. those involving the structure movement.The analyses have been carried out by considering, at the same time, all the three compo-

nents of the ground motion. Also the original fixed base structure has been analysed with thesame model and accelerograms. Fig. 6 shows the diagrams of the response of the isolation

system without friction, in X direction. The similitude with the results relevant to the SDOF isvery strong, both in terms of displacement time history and of force-displacement curve of the

isolation system. Some reduction of the maximum force is to be ascribed to the slightlysmaller total isolated mass of the complete structural model, which results to be 550 t instead

of 590 t, and the consequent adjustments made to the isolation devices.All the three systems have been subjected to 4 artificial accelerograms, generated according

to the specifications given in the Guidelines. Two values of the friction coefficient have beentaken into consideration: 0% and 3%. The main results and the design values of the main

quantities, obtained according to the Guidelines, are shown in table 1. The boldface characteridentify the most unfavourable values that must be assumed for design. Average and standard

deviation values are referred to the four accelerograms. The design values have been obtainedaccording to the directions given in the guidelines. In particular the design displacements of

the two elasto-plastic isolation systems (solutions 2 and 3) have been evaluated considering a1.5 safety coefficient on the mean value plus one standard deviation. In order to show the

considerable safety gain obtainable with the isolation system, in fig. 7 there are reported the


-1500

-1000

-500

0

50 0

1000

1500

-0,10 -0,05 0,00 0,05 0,10

d (m)

F ( k N )

Total displacement - timei

-0,10

-0,05

0,00

0,05

0,10

0,00 10,00 20,00 30,00

t (sec)

d ( m )

Fig. 6 – Response of the isolation system (X-direction) in the detailed structural model

without friction in sliders.

1st story

-0,010

-0,008

-0,006

-0,004

-0,002

0,000

0,002

0,004

0,006

0,008

0,010

0 5 10 15 20 25 30 35time (sec)

d x ( m )

Fixed base

isolated with friction

2nd story

-0,010

-0,008

-0,006

-0,004

-0,002

0,000

0,002

0,004

0,006

0,008

0,010

0 5 10 15 20 25 30 35time (sec)

d x ( m )

Fixed base

isolated with friction

Fig. 7 – Interstory drift of the isolated and non isolated structures (X-direction).



interstory drift time histories at the first isolated story in X direction, for the unfavourable

condition of friction in the sliders. The same quantity is reported also for the fixed basestructure. The fixed base structure has been modelled as an elastic system, like the isolated

structure, considering, however, 5% modal damping for the first mode, instead of 2%, toaccount for the higher stresses. Therefore the values obtained for the fixed base structure

should be considered as an underestimate of the values in the real structure, which wouldundergo large inelastic deformations under the ULS design earthquake used in the analyses.

As can be seen, the maximum interstory drift in the isolated structure is of the order of 1.5-2mm, i.e. less than 0.1% of the story height, thus excluding any structural and non structural

damage under the ULS earthquake. Considering that the Guidelines limit for the drift is 0.1%under the DLS conditions, wide margins exist to still optimise the isolated structure.

The comparison of the response of structures with and without seismic isolation, also consi-dering the underestimation of the response of the fixed base structure implied by the elastic

model, emphasises the considerable safety improvement produced by seismic isolation, beingthe interstory drift in the isolated structure about four times smaller.

6. CONCLUSION

The seismic isolation solution for private houses considered in the present paper has beenoptimised with respect to the geometrical and architectural constraints, structural performan-

ces and costs. A new hybrid device has been proposed, which combines the low initialstiffness of rubber isolator with the force limiting capability of steel based elasto-plastic

devices. The solution resulting from the use of sliders and hybrid devices offers the bestcompromise among the different needs, namely the limitation of the base shear and of the

isolation displacement. Moreover it fulfil a strict requirement of the Italian Guidelines, that noyielding shall occur in steel-based device under Damage Limit State conditions.

The final non linear analyses on a refined structural model emphasise the high safety levelsattained with the isolation system. No damage to structural and non structural elements are

envisaged under the design (500 year return period) earthquake. This was a very convincingaspect for the designer and the private customers, that were prepared to carry some additional

costs to get a higher safety level for their house. However the exceptionally large safetymargins obtained with the isolation solution also suggests that some structural optimisation

can still be carried out to reduce the additional costs.

REFERENCES

Consiglio Superiore dei Lavori Pubblici, 1998 “Guidelines for the Design, Construction and

Check of Structures with Seismic Isolation”, Min. LL.PP., Rome.

Braga F., Laterza M., 2001, “Differenti tecniche di isolamento alla base: il sistema misto

scivolatori/isolatori elastomerici della struttura sperimentale di Rapolla”, Proc. 10th Italian

Conference on Earthquake Engineering, Potenza-Matera.

Dolce M., Filardi B., Marnetto R., Nigro D., 1996, “Experimental Tests and Applications of a

New Biaxial Elastoplastic Device for the Passive Control of Structures”. Proc. 4th W. Con-

gress on Joint Sealants and Bearing Systems for Concrete Structures, Sacramento, CA (USA).

edificio isolato tis.pdf

Documents