**4. Energy dissipation devices installed at university of L'Aquila buildings**

Among several interventions, designed with the intent of increasing the dissipative capacity of the structure through seismic protection elements, the case of the Edifice A of the Engineer‐ ing Campus has been here selected as case study. The peculiarity of this intervention should be searched on the idea of enhancing the control performance through the dissipative con‐ nection of adjacent structures. Indeed, the last two decades increasing attention on the mitigation of seismic or wind induced vibrations in adjacent structures through their "smart"


**Table 3.** Examples of structural monitoring systems installed at L'Aquila

exploitation of confined material in the elasto-plastic regime during the earthquake. Only in the case of the Edifice A of the Engineering Faculty Building forty-three (43) nonlinear viscous fluid dampers of three different types have been installed looking for the increase of dissipation

Before the 2009 L'Aquila earthquake a strong network of seismic accelerometers were func‐ tioning close to the epicenter, mostly managed by the Italian Institute of Geophysics and Volcanology (INGV) [15], while very few structure were equipped by a permanent structural monitoring managed by Department of Civil Protection (DPC) [16] (see, also, Table 3). In particular, the response of the Pizzoli Town Hall during the main shock has been recorded and analyzed by DPC, giving special insights on the potentiality of these systems for immediate evaluation of the damaged occurred during an earthquake. The large amount of installed, temporally or permanently, devices of different type (accelerometers, smart wireless devices, displacement and velocity transducers, inclinometers, etc) reach a number of around three hundred (300) evidencing a large impact of this technology in the post-earthquake emergency phase, especially during the earthquake swarms. In particular several monitoring systems have been installed in the emergency phase, during the construction of temporary scaffolding, in order to verify the efficacy of the added structural system especially in the case of monu‐ mental building (see for example [17]). Because of this scope, in many cases, the permanent monitoring has worked only for a limited number of months (in the Table 3, the period is not always precisely known to the author and sometimes it should be considered indicative). In other cases, the monitoring system is permanently installed on the structure and it can be used also to determine the change that will occur in the structural behavior during the reconstruc‐

In several cases, the structural monitoring system uses only accelerometers, starting from very few measures (three channels in the minor case) to larger number of devices with different characteristics and sensitivity. Instead more complex monitoring systems are used in complex monumental churches and buildings where accelerometers are joined with crackmeters,

**4. Energy dissipation devices installed at university of L'Aquila buildings**

Among several interventions, designed with the intent of increasing the dissipative capacity of the structure through seismic protection elements, the case of the Edifice A of the Engineer‐ ing Campus has been here selected as case study. The peculiarity of this intervention should be searched on the idea of enhancing the control performance through the dissipative con‐ nection of adjacent structures. Indeed, the last two decades increasing attention on the mitigation of seismic or wind induced vibrations in adjacent structures through their "smart"

through the relative velocity of adjacent sub-structures.

212 Engineering Seismology, Geotechnical and Structural Earthquake Engineering

inclinometers, and temperature measurement devices, etc.

tion phase [8,9].

**3. Structural monitoring systems installed at L'Aquila**

coupling has been examined. Several studies have been devoted to optimize the dynamic performance of slender structures, such as skyscrapers or tall buildings, introducing dissipa‐ tion systems acting on the relative motion and aiming to reduce the maximum displacements at the higher floors. Different applications of similar concepts have been applied in the retrofitting of existing adjacent structures. The placements of viscous-type coupling devices into seismic joints have been proposed to dissipate energy and to avoid hammering phenom‐ ena [18-21]. In all cases "smart" coupling between adjacent structures has been exploited using passive, semi-active, and active control systems with different features and performances.

Focusing the attention on the passive coupling of adjacent structures, different modelling approaches have been used. The synthetic description of the main problem features through a pair of simple oscillators interconnected by means of a springs and dashpot in series or parallel fashion has been proposed by many authors [22-25]. The use of a simple oscillator pair has been pursued by the research group of L'Aquila both for the proposal of a new design method [26-28] and the use of it at the preliminary stage of the design of the more complex system installed at the Edifice A of the Engineering Faculty [6]. In the following the entire process has been summarized.

where **u** is the displacement vector, **M** and **K** are the mass and stiffness matrices, **s** and **r** are

Advanced Applications in the Field of Structural Control and Health Monitoring After the 2009 L'Aquila Earthquake

1 0 1 0 1 1 , , ,, <sup>0</sup> <sup>0</sup> 1 1 *u*

Different rheological models of the coupling damper are introduced to define the constitutive law *u*(*u*, *u***˙**) , relating the control force to the displacement/velocity vector. Adopting a state-

where the state matrix A, the allocation control vector b the external input vector h are,

Constitutive models describing with increasing complexity the damper behaviour can be formulated joining, in different combination schemes, simple elements: a linear spring with elastic constant *K*, and a linear dashpot with viscous constant *C*. Introducing the dimensionless

> 1 1 1 1 , <sup>2</sup> *K C M M*

the *KV* and the *Ma* model correspond to the alternative parallel or series combination of the

It is worth noting that the *Ma* model entails an increment of the model dimension due to the damper internal dynamics, described by a supplementary half degree-of-freedom. It can be demonstrated that in the *KV* case, the design coupling parameters can be chosen according to

w

 g

<sup>1</sup> , , - é ùìü ï ï ïï ì ü = == ê ú í ý íý ê ú - -ï ï ï ï - ë ûîþ î þ **<sup>1</sup> 0I 0 <sup>0</sup> A bh**

2

spring and the dashpot, respectively. Consequently, the constitutive law reads

w

h

*η* )

**•** *KV* model *u* =*η*(*u*<sup>2</sup> −*u*1) + 2*γ*(*u*˙ <sup>2</sup> −*u*˙ 1)

**•** *Ma* model *<sup>u</sup>* =2*γ*(*u*˙ <sup>2</sup> <sup>−</sup>*u*˙ <sup>1</sup> <sup>−</sup> *<sup>u</sup>*˙

the following equations

 *u* é ù é ù ì ü ï ï ïï ï ï ìü ì ü - = = = == ê ú ê ú í ý íý í ý ë û ê ú ë û ï ï î þ ïï ï ï îþ î þ

2

rb

space representation, with the use of the state vector *x* ={*u* T, *u***˙** T}<sup>T</sup>

1

**M K u rs** (4)

*<sup>g</sup>* **x Ax b h** & = ++ *u u*&& (5)

**MK 0 Ms <sup>r</sup>** (6)

= = (7)

the equation (3) can be

http://dx.doi.org/10.5772/55438

215

2

the position vectors of the control and external forces

r

rewritten as

respectively

parameters

**Figure 1.** Passive control of adjacent structures: (a) two-dofs model, (b) damper models.

#### **4.1. Simple model of two coupled oscillators for preliminary design**

Consider two simple linear oscillators with mass *M <sup>j</sup>* and stiffness *K <sup>j</sup>* , (j=1,2), coupled by a passive damper (Figure 1a). Denoting *U <sup>1</sup>* and *U <sup>2</sup>* the relative horizontal displacements and *F* the mutual force applied by the coupling damper, the dynamic response of the two-degreesof-freedom (*dofs*) system to a synchronous horizontal ground displacement *U <sup>g</sup>*, is governed by the equations

$$\begin{aligned} M\_1 \ddot{\mathcal{U}}\_1 + K\_1 \mathcal{U}\_1 - F &= -M\_1 \ddot{\mathcal{U}}\_{\mathcal{g}} \\ M\_2 \ddot{\mathcal{U}}\_2 + K\_2 \mathcal{U}\_2 + F &= -M\_2 \ddot{\mathcal{U}}\_{\mathcal{g}} \end{aligned} \tag{1}$$

where dot indicates derivative with respect to time t. Denoting L a convenient reference length, and the following dimensionless variables and parameters can be introduced

$$\alpha\_j = \frac{\mathcal{U}\_j}{\mathcal{L}}, \quad u\_g = \frac{\mathcal{U}\_g}{\mathcal{L}}, \quad \alpha\_j^2 = \frac{\mathcal{K}\_j}{\mathcal{M}\_j}, \quad \beta = \frac{\alpha\_2}{\alpha\_1}, \quad \rho = \frac{\mathcal{M}\_2}{\mathcal{M}\_1}, \quad u = \frac{F}{\alpha\_1^2 M\_1 L}, \quad \tau = \alpha\_1 \tau \tag{2}$$

where the dimensionless force u is understood as the control variable, and the relevant parameters *ρ* and *β* stand for the mass and frequency ratio between the two uncoupled oscillators, respectively. The equations of motion can be rewritten in the synthetic form

$$\mathbf{M}\ddot{\mathbf{u}} + \mathbf{K}\mathbf{u} + s\boldsymbol{\omega}\left(\mathbf{u}, \dot{\mathbf{u}}\right) = -\mathbf{M}\mathbf{r}\ddot{\boldsymbol{\imath}}\_{\mathcal{g}}\tag{3}$$

where **u** is the displacement vector, **M** and **K** are the mass and stiffness matrices, **s** and **r** are the position vectors of the control and external forces

$$\mathbf{M} = \begin{bmatrix} 1 & 0 \\ 0 & \rho \end{bmatrix}', \quad \mathbf{K} = \begin{bmatrix} 1 & 0 \\ 0 & \rho \beta^2 \end{bmatrix}', \quad \mathbf{u} = \begin{Bmatrix} u\_1 \\ u\_2 \end{Bmatrix}', \quad \mathbf{r} = \begin{Bmatrix} 1 \\ 1 \end{Bmatrix}', \quad \mathbf{s} = \begin{Bmatrix} -1 \\ 1 \end{Bmatrix}' \tag{4}$$

Different rheological models of the coupling damper are introduced to define the constitutive law *u*(*u*, *u***˙**) , relating the control force to the displacement/velocity vector. Adopting a statespace representation, with the use of the state vector *x* ={*u* T, *u***˙** T}<sup>T</sup> the equation (3) can be rewritten as

$$
\dot{\mathbf{x}} = \mathbf{A}\mathbf{x} + \mathbf{b}u + \mathbf{h}\ddot{u}\_{\mathcal{g}} \tag{5}
$$

where the state matrix A, the allocation control vector b the external input vector h are, respectively

$$\mathbf{A} = \begin{bmatrix} \mathbf{0} & \mathbf{I} \\ -\mathbf{M}^{-1}\mathbf{K} & \mathbf{0} \end{bmatrix}' \quad \mathbf{b} = \begin{Bmatrix} \mathbf{0} \\ -\mathbf{M}^{-1}\mathbf{s} \end{Bmatrix}, \quad \mathbf{h} = \begin{Bmatrix} \mathbf{0} \\ -\mathbf{r} \end{Bmatrix} \tag{6}$$

Constitutive models describing with increasing complexity the damper behaviour can be formulated joining, in different combination schemes, simple elements: a linear spring with elastic constant *K*, and a linear dashpot with viscous constant *C*. Introducing the dimensionless parameters

$$\eta = \frac{\text{K}}{\text{o}\_1^2 \text{M}\_1}, \quad \eta = \frac{\text{C}}{2\text{o}\_1 \text{M}\_1} \tag{7}$$

the *KV* and the *Ma* model correspond to the alternative parallel or series combination of the spring and the dashpot, respectively. Consequently, the constitutive law reads

$$\text{• } KV \text{ model } \mathfrak{u} = \mathfrak{h} (\mathfrak{u}\_2 - \mathfrak{u}\_1) + 2\gamma (\dot{\mathfrak{u}}\_2 - \dot{\mathfrak{u}}\_1)$$

**•** *Ma* model *<sup>u</sup>* =2*γ*(*u*˙ <sup>2</sup> <sup>−</sup>*u*˙ <sup>1</sup> <sup>−</sup> *<sup>u</sup>*˙ *η* )

system installed at the Edifice A of the Engineering Faculty [6]. In the following the entire

*U*g

passive damper (Figure 1a). Denoting *U <sup>1</sup>* and *U <sup>2</sup>* the relative horizontal displacements and *F* the mutual force applied by the coupling damper, the dynamic response of the two-degreesof-freedom (*dofs*) system to a synchronous horizontal ground displacement *U <sup>g</sup>*, is governed

where dot indicates derivative with respect to time t. Denoting L a convenient reference length,

*g g*

1 1 1 1

= = = = = = <sup>=</sup> (2)

&& && (1)

w

( ) *<sup>g</sup>* **Mu Ku u,u Mr** && + + =- *su* & &&*u* (3)

2 1

t wt

11 11 1 22 22 2

2 2 2

, , ,, , ,

br

where the dimensionless force u is understood as the control variable, and the relevant parameters *ρ* and *β* stand for the mass and frequency ratio between the two uncoupled oscillators, respectively. The equations of motion can be rewritten in the synthetic form

w

*LLM M M L* w

*MU KU F MU MU KU F MU* + - =- + + =- && &&

and the following dimensionless variables and parameters can be introduced

*UU K M F*

*j*

*u u u*

**Figure 1.** Passive control of adjacent structures: (a) two-dofs model, (b) damper models.

**4.1. Simple model of two coupled oscillators for preliminary design**

Consider two simple linear oscillators with mass *M <sup>j</sup>* and stiffness *K <sup>j</sup>*

*K C, α*

*C, α*

*KELVIN-VOIGT*

*MAXWELL*

, (j=1,2), coupled by a

process has been summarized.

by the equations

*M*<sup>1</sup> *M*<sup>2</sup>

*DAMPER*

*K*<sup>1</sup> *K*<sup>2</sup>

*jg j*

w

*jg j*

*U*<sup>1</sup> *U*<sup>2</sup>

214 Engineering Seismology, Geotechnical and Structural Earthquake Engineering

(a) (b) *<sup>K</sup>*

It is worth noting that the *Ma* model entails an increment of the model dimension due to the damper internal dynamics, described by a supplementary half degree-of-freedom. It can be demonstrated that in the *KV* case, the design coupling parameters can be chosen according to the following equations

$$\eta\_c = \frac{\rho (1 - \rho^2)(1 - \rho^2 \beta^2)}{(1 + \rho)(1 + \rho \beta^2)}; \quad \gamma\_c = \frac{\rho}{1 + \rho} \left( 1 + \eta\_c + \beta^2 + \frac{\eta\_c}{\rho} - 2 \left( \beta^2 \left( 1 + \eta\_c \right) + \frac{\eta\_c}{\rho} \right)^{1/2} \right)^{1/2} \tag{8}$$

failed under the combined effects of the unexpected cyclic axial loads and the repeated impacts of the bricks falling down from above. In the progression of the damage the failure of the connection played a great role facilitating the augment of both relative displacements between the two structures (facade and three-dimensional frame) and absolute displacements and

Advanced Applications in the Field of Structural Control and Health Monitoring After the 2009 L'Aquila Earthquake

http://dx.doi.org/10.5772/55438

217

**Figure 2.** Edifice A: a) plan view at the main entrance level 0, b) facade view c) section A-A

**Figure 3.** Damages caused by 2009 earthquake to the Edifice A: a) internal view of the main facade, b) internal parti‐

tioning walls c) heavy bricks fallen down inside the building from the facade.

acceleration on the facade.

in order to assure for the coupled system specific features with respect to the base excitation [28].

Similar characteristics have been found in the *Ma* case for which only numerical analysis have been performed to determine the design coupling parameters *η <sup>c</sup>*, *γ <sup>c</sup>*.

#### **4.2. Seismic protection of Edifice A through nonlinear viscous dampers**

During the seismic event of 6th April 2009, the edifices of the Engineering faculty have suffered particularly for seismic induced large structural displacements and accelerations which have brought them out of order due to the failure of non-structural elements [4], the breakage of wiring and piping systems and the destruction of furniture and machineries. In particular, among the three recently-built buildings of the campus, erected in the early 90's, the so-called "Edifice A" presents the most critical damage scenario, which needs a significant rehabilitating intervention.

Edifice A is a four-story building with the resistant structure made of reinforced concrete frames, sitting on a sloping site. Several seismic joints divide the structure into seven inde‐ pendent substructures (Figure 2); some of them are structurally featured by a frame-shear-wall interactive system. In the substructures, the walls are widely used to reinforce and to stiffen the acute corners, the rounded staircases close to the elevator cores and the lower floors. The plan is characterized by asymmetry, with uneven distribution of stiffness and vertical irregu‐ larities, and double- or triple-height rooms. The amphitheater facing the main entrance, on the north-west side, is sustained by an independent structure. Concrete slabs are used to realize all the horizontal planes including the roof.

The most evident damages in the Edifice A of Engineering Faculty were found to be localized in the main facade, which has lost large portions of the veneer masonry, made of heavy splitface bricks (Figure 3), laying bare the underlying reinforced concrete structure, remained practically undamaged. All the results collected during the early inspections confirmed that the structure underwent an excessive displacement and acceleration level, surely incompatible with the resistance of many non-structural elements. The massive inward cascade of heavy bricks and sharp glass, fallen down from the facade and the wall of the internal stairs, has realized an unpleasant dangerous scenario [1,4].

Aiming to understand the structural reasons for this inadequate behavior, it should be considered that the design concept follows the idea to have the planar structure sustaining the principal facade rigidly coupled with the three-dimensional frame of the building behind.

Horizontal steel tubes, functioning as interconnecting rods at different floor levels, ensured the coupling between the two substructures (Figure 3c). The bolted anchorages at the rod ends were probably under-dimensioned for the exceptional seismic action, since many of them failed under the combined effects of the unexpected cyclic axial loads and the repeated impacts of the bricks falling down from above. In the progression of the damage the failure of the connection played a great role facilitating the augment of both relative displacements between the two structures (facade and three-dimensional frame) and absolute displacements and acceleration on the facade.

( )

 h  h

> r

(8)

*c c*

 b

1/2 1/2 2 22

rr

in order to assure for the coupled system specific features with respect to the base excitation

Similar characteristics have been found in the *Ma* case for which only numerical analysis have

During the seismic event of 6th April 2009, the edifices of the Engineering faculty have suffered particularly for seismic induced large structural displacements and accelerations which have brought them out of order due to the failure of non-structural elements [4], the breakage of wiring and piping systems and the destruction of furniture and machineries. In particular, among the three recently-built buildings of the campus, erected in the early 90's, the so-called "Edifice A" presents the most critical damage scenario, which needs a significant rehabilitating

Edifice A is a four-story building with the resistant structure made of reinforced concrete frames, sitting on a sloping site. Several seismic joints divide the structure into seven inde‐ pendent substructures (Figure 2); some of them are structurally featured by a frame-shear-wall interactive system. In the substructures, the walls are widely used to reinforce and to stiffen the acute corners, the rounded staircases close to the elevator cores and the lower floors. The plan is characterized by asymmetry, with uneven distribution of stiffness and vertical irregu‐ larities, and double- or triple-height rooms. The amphitheater facing the main entrance, on the north-west side, is sustained by an independent structure. Concrete slabs are used to realize

The most evident damages in the Edifice A of Engineering Faculty were found to be localized in the main facade, which has lost large portions of the veneer masonry, made of heavy splitface bricks (Figure 3), laying bare the underlying reinforced concrete structure, remained practically undamaged. All the results collected during the early inspections confirmed that the structure underwent an excessive displacement and acceleration level, surely incompatible with the resistance of many non-structural elements. The massive inward cascade of heavy bricks and sharp glass, fallen down from the facade and the wall of the internal stairs, has

Aiming to understand the structural reasons for this inadequate behavior, it should be considered that the design concept follows the idea to have the planar structure sustaining the principal facade rigidly coupled with the three-dimensional frame of the building behind. Horizontal steel tubes, functioning as interconnecting rods at different floor levels, ensured the coupling between the two substructures (Figure 3c). The bolted anchorages at the rod ends were probably under-dimensioned for the exceptional seismic action, since many of them

 hb

æ ö - - æ ö <sup>=</sup> = ++ + - + + ç ÷ ç ÷ + + <sup>+</sup> è ø è ø

(1 )(1 ); 1 21

*c c c c*

been performed to determine the design coupling parameters *η <sup>c</sup>*, *γ <sup>c</sup>*.

**4.2. Seismic protection of Edifice A through nonlinear viscous dampers**

g

216 Engineering Seismology, Geotechnical and Structural Earthquake Engineering

 r

2

(1 )(1 ) 1

 rb

 rb

all the horizontal planes including the roof.

realized an unpleasant dangerous scenario [1,4].

r b

r

h

[28].

intervention.

2 2

h

**Figure 2.** Edifice A: a) plan view at the main entrance level 0, b) facade view c) section A-A

**Figure 3.** Damages caused by 2009 earthquake to the Edifice A: a) internal view of the main facade, b) internal parti‐ tioning walls c) heavy bricks fallen down inside the building from the facade.

exploring new geometric arrangements and technical solutions. After several discussions taking into account comparative criteria (including structural performance, aesthetic outcome and economic aspects), dissipative steel bars, embedding viscous dampers and arranged in a stiff K-shaped configuration, reproducing a planar truss structure, have been selected to restore the facade-structure connection. The leading idea is to realize a dissipative coupling between two adjacent structures with different stiffness, that is the stiff principal threedimensional structure and the flexible planar frame sustaining the facade. Here, a complete analysis on the benefits reached by the K-shaped dissipative configuration is performed by means of direct time-integration of the nonlinear motion equation numerically obtained through a classical finite element approach in both the case of rigid or dissipative intercon‐ nection, for which the nonlinear constitutive relation is fully supported by experimental evidence of the assumed coefficients in the analysis [5]. Seven different acceleration time histories, with different time records (35s-70s) and described through 200 samples per second, have been used to describe the base motion, with spectrum characteristics compatible with the site (Figures 4) [4]. The numerical simulations carried out looking at the complete dynamic structural response (see for example Figure 5) have used as first starting value, the stiffness and viscous coefficient design parameters of the *Ma* linear model obtained through the method

Advanced Applications in the Field of Structural Control and Health Monitoring After the 2009 L'Aquila Earthquake

http://dx.doi.org/10.5772/55438

219

However, the final assessment of the viscous coefficient *c* characterizing mainly the nonlinear viscous dampers (Figure 5d) has been determined from a multistep iterative process, which has allowed the selection of its optimal values [6]. The fractional exponent α has been consid‐ ered with fixed in the design process (α=0.15), because the manufacturer has assigned it. Selection criteria including both the minimization of the displacements/accelerations at the

The analyses show a good performance of the dissipative coupling if both the adjacent structures are subject to significant absolute and relative displacements, as verified in sub‐ structures A3 and A4. Differently, when the natural frequencies of the coupled structures are appreciably different, as occurred in the stiffer substructures A1, A2 and A6, despite the dissipation is potentially maximized; low displacements are associated to lower dissipated

To reduce the displacements in the longitudinal direction on substructures A3 and A4, a proper coupling to the adjacent A2 and A6 substructures has been designed. The frequency difference in the dominant longitudinal modes of substructures A2 and A3, A4 and A6 has permitted to enhance the efficiency of the dampers in reducing the inter-storey drift in the higher floors. The dampers were installed at the third and fourth floor on the elevator tube-section (A2 and A6), or the frame (A3 and A4). The definition of three synthetic performance indices (*J* <sup>i</sup>

ratios between the structural performance of each original undamaged and retrofitted substructures, in terms of peak displacements (*J* 1), accelerations (*J* 2), and based shear forces (*J* 3), allow to clearly emphasize the achieved enhancement in the seismic behavior. Moreover,

an additional index (*J* 4) represents the average of previous indices.

), the

highest floor, and the reduction of the base section shear stresses have been used.

mentioned above for the preliminary design.

energy.

**Figure 4.** Seven natural earthquake realizations with an average spectrum compatible with the design one used for the evaluation of the seismic protection performances.

Moreover, the overall dynamical phenomenon was probably emphasized by the different mechanical properties of the coupled substructures.

A deep knowledge of the structures has permitted to design an optimized retrofitting inter‐ vention, able to satisfy high performance criteria defined in the current Italian National Code [29]. Before the retrofitting interventions, the most vulnerable aspects of the original design have been detected through the comparison with the limitations imposed by the newest national design code. Finite element models for each independent substructure were used, based on the previously obtained information, to verify both the operational limit state and ultimate limit state requests in terms of inter-storey drifts and ultimate strength of each element, respectively. These analyses put into evidence excessive deformation levels of the higher floors, while the other substructures have resulted lower flexible, due to the stiffening presence of fully-height shear walls. The other substructures satisfy the maximum inter-story drift requirements at operational limit state [4]. The effectiveness of the connection between the principal three-dimensional structure and the planar frame sustaining the facade has been recognized as the critical issue to be addressed for the enhancement of the seismic performance.

The limited efficacy of the original metallic tubes, which rigidly couple the facade with the main structure, evidences, also through the occurred damage, large absolute facade displace‐ ments and accelerations with high frequency content. This occurrence has suggested consid‐ ering and comparing different alternatives in reconstructing the damaged coupling elements, exploring new geometric arrangements and technical solutions. After several discussions taking into account comparative criteria (including structural performance, aesthetic outcome and economic aspects), dissipative steel bars, embedding viscous dampers and arranged in a stiff K-shaped configuration, reproducing a planar truss structure, have been selected to restore the facade-structure connection. The leading idea is to realize a dissipative coupling between two adjacent structures with different stiffness, that is the stiff principal threedimensional structure and the flexible planar frame sustaining the facade. Here, a complete analysis on the benefits reached by the K-shaped dissipative configuration is performed by means of direct time-integration of the nonlinear motion equation numerically obtained through a classical finite element approach in both the case of rigid or dissipative intercon‐ nection, for which the nonlinear constitutive relation is fully supported by experimental evidence of the assumed coefficients in the analysis [5]. Seven different acceleration time histories, with different time records (35s-70s) and described through 200 samples per second, have been used to describe the base motion, with spectrum characteristics compatible with the site (Figures 4) [4]. The numerical simulations carried out looking at the complete dynamic structural response (see for example Figure 5) have used as first starting value, the stiffness and viscous coefficient design parameters of the *Ma* linear model obtained through the method mentioned above for the preliminary design.

However, the final assessment of the viscous coefficient *c* characterizing mainly the nonlinear viscous dampers (Figure 5d) has been determined from a multistep iterative process, which has allowed the selection of its optimal values [6]. The fractional exponent α has been consid‐ ered with fixed in the design process (α=0.15), because the manufacturer has assigned it. Selection criteria including both the minimization of the displacements/accelerations at the highest floor, and the reduction of the base section shear stresses have been used.

0 1 2 3 **T [sec]** 4

0 10 20 30 40 50 60

**[g]** 000600-x

0 10 20 30 40 50 60

**[g]** 006947-x

**T [sec]**

000600-y

**T [sec]**

006947-y

0

**Figure 4.** Seven natural earthquake realizations with an average spectrum compatible with the design one used for

Moreover, the overall dynamical phenomenon was probably emphasized by the different

A deep knowledge of the structures has permitted to design an optimized retrofitting inter‐ vention, able to satisfy high performance criteria defined in the current Italian National Code [29]. Before the retrofitting interventions, the most vulnerable aspects of the original design have been detected through the comparison with the limitations imposed by the newest national design code. Finite element models for each independent substructure were used, based on the previously obtained information, to verify both the operational limit state and ultimate limit state requests in terms of inter-storey drifts and ultimate strength of each element, respectively. These analyses put into evidence excessive deformation levels of the higher floors, while the other substructures have resulted lower flexible, due to the stiffening presence of fully-height shear walls. The other substructures satisfy the maximum inter-story drift requirements at operational limit state [4]. The effectiveness of the connection between the principal three-dimensional structure and the planar frame sustaining the facade has been recognized as the critical issue to be addressed for the enhancement of the seismic performance.

The limited efficacy of the original metallic tubes, which rigidly couple the facade with the main structure, evidences, also through the occurred damage, large absolute facade displace‐ ments and accelerations with high frequency content. This occurrence has suggested consid‐ ering and comparing different alternatives in reconstructing the damaged coupling elements,

**T [sec]**

000181-y

**T [sec]**

**T [sec]**

005794-y

000292-y

**T [sec]**

000369-y

**T [sec]**

001246-y






0 5 10 15 20 25 30 35 40

mechanical properties of the coupled substructures.

the evaluation of the seismic protection performances.

**[g]** 000181-x

0 5 10 15 20 25 30 35 40 45

0 10 20 30 40 50 60 70

**[g]** 005794-x

**[g]** 000292-x

0 5 10 15 20 25

**[g]** 000369-x

0 10 20 30 40 50 60 70

**[g]** 001246-x

218 Engineering Seismology, Geotechnical and Structural Earthquake Engineering

0.5

1

1.5

2

2.5 **a [g]**



> The analyses show a good performance of the dissipative coupling if both the adjacent structures are subject to significant absolute and relative displacements, as verified in sub‐ structures A3 and A4. Differently, when the natural frequencies of the coupled structures are appreciably different, as occurred in the stiffer substructures A1, A2 and A6, despite the dissipation is potentially maximized; low displacements are associated to lower dissipated energy.

> To reduce the displacements in the longitudinal direction on substructures A3 and A4, a proper coupling to the adjacent A2 and A6 substructures has been designed. The frequency difference in the dominant longitudinal modes of substructures A2 and A3, A4 and A6 has permitted to enhance the efficiency of the dampers in reducing the inter-storey drift in the higher floors. The dampers were installed at the third and fourth floor on the elevator tube-section (A2 and A6), or the frame (A3 and A4). The definition of three synthetic performance indices (*J* <sup>i</sup> ), the ratios between the structural performance of each original undamaged and retrofitted substructures, in terms of peak displacements (*J* 1), accelerations (*J* 2), and based shear forces (*J* 3), allow to clearly emphasize the achieved enhancement in the seismic behavior. Moreover, an additional index (*J* 4) represents the average of previous indices.

) (a) (b

**OTP 20/ 180 Constitutive Law Test**

F

http://dx.doi.org/10.5772/55438

221

FLB

FUB

Fexp

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 **Velocity [m/ s]**

*Nonlinear viscous devices*

**Force [kN]**

Advanced Applications in the Field of Structural Control and Health Monitoring After the 2009 L'Aquila Earthquake

**Figure 6.** Design force-velocity relations for the three types of nonlinear viscous dampers; a) DV1 b) DV2 c) DV3 d)

*Longitudinal nonlinear Rigid connections*

*Transversal nonlinear viscous devices*

DV1

**Figure 7.** Three-dimensional sketch drawings reporting fluid viscous dampers locations in the structural retrofitting of

DV3

(c) (d)

*viscous devices*

experimental data from the characterization tests [7].

DV2

the Edifice A of the Engineering Faculty of University of L'Aquila.

**Figure 5.** Numerical simulations: a) finite element model of substructure A3, b) c) numerically simulated dissipative cycles for the nonlinear viscous devices, d) nonlinear *Ma* model, e) experimental dissipative cycles.


**Table 4.** Performance indices evaluated for the adopted solution.

The rewarding enhancement of seismic structural behavior of substructure A3 and A4, are demonstrated in Table 4, evidencing the effect of viscous coupling in the principal directions, monitored on the top floor. The designed retrofitting reduces substantially the maximum peak displacement (see *J* 1 in Table 4), particularly in transversal direction, out of plane of the coupled facade frame. The stiffer substructure A2 and substructure A6 contribute, through the viscous coupling, to reduce the maximum displacement in the longitudinal direction. Similar beneficial effects are registered in the peak acceleration reduction, both in transversal and longitudinal direction (see *J* 2 in Table 4) while the transverse shear force at the base of vertical resistant elements appears not significantly reduced by the viscous coupling (*J* 3 in Table 4). Figures 5b and c show selected dissipative cycles evaluated during the numerical simulation for a given different base excitation within the seven cases. One of the simulated behavior for a selected device has been also reproduced (Figure 5e) during the campaign tests for the mechanical characterization of the installed devices confirming the expected performances [7].

The use of the performance indexes have permitted to determine an optimized solution which take into account the possibility of having a limited number of different type of dampers, for production reasons. Figures 6a, b and c show the designed constitutive relations for the three selected dampers in the final solution furnished to the manufacturer. Figure 6d shows the results obtained during the test campaign [7].

Advanced Applications in the Field of Structural Control and Health Monitoring After the 2009 L'Aquila Earthquake http://dx.doi.org/10.5772/55438 221

(a)

(b)

)

(c)

cycles for the nonlinear viscous devices, d) nonlinear *Ma* model, e) experimental dissipative cycles.

**Table 4.** Performance indices evaluated for the adopted solution.

220 Engineering Seismology, Geotechnical and Structural Earthquake Engineering

results obtained during the test campaign [7].

**Figure 5.** Numerical simulations: a) finite element model of substructure A3, b) c) numerically simulated dissipative

Longitudinal 0.74 0.85 0.73 0.78 Transversal 0.58 0.81 0.96 0.78

The rewarding enhancement of seismic structural behavior of substructure A3 and A4, are demonstrated in Table 4, evidencing the effect of viscous coupling in the principal directions, monitored on the top floor. The designed retrofitting reduces substantially the maximum peak displacement (see *J* 1 in Table 4), particularly in transversal direction, out of plane of the coupled facade frame. The stiffer substructure A2 and substructure A6 contribute, through the viscous coupling, to reduce the maximum displacement in the longitudinal direction. Similar beneficial effects are registered in the peak acceleration reduction, both in transversal and longitudinal direction (see *J* 2 in Table 4) while the transverse shear force at the base of vertical resistant elements appears not significantly reduced by the viscous coupling (*J* 3 in Table 4). Figures 5b and c show selected dissipative cycles evaluated during the numerical simulation for a given different base excitation within the seven cases. One of the simulated behavior for a selected device has been also reproduced (Figure 5e) during the campaign tests for the mechanical

characterization of the installed devices confirming the expected performances [7].

The use of the performance indexes have permitted to determine an optimized solution which take into account the possibility of having a limited number of different type of dampers, for production reasons. Figures 6a, b and c show the designed constitutive relations for the three selected dampers in the final solution furnished to the manufacturer. Figure 6d shows the

*J*<sup>1</sup> *J*<sup>2</sup> *J*<sup>3</sup> *J*<sup>4</sup>

(e)

)

(d)

**K C** *f f*

Nonlinear viscous Maxwell model 

÷ ÷ ø ö ç ç è <sup>æ</sup> -= *<sup>k</sup> <sup>F</sup> xCF* & & ÷ ø <sup>ö</sup> <sup>ç</sup> è <sup>æ</sup> -= *<sup>k</sup> <sup>f</sup> xcf* & 1

**Figure 6.** Design force-velocity relations for the three types of nonlinear viscous dampers; a) DV1 b) DV2 c) DV3 d) experimental data from the characterization tests [7].

**Figure 7.** Three-dimensional sketch drawings reporting fluid viscous dampers locations in the structural retrofitting of the Edifice A of the Engineering Faculty of University of L'Aquila.

Figure 7 clarifies the location of the 43 devices: 18 DV1 (Type I ); 17 DV2 (Type II) ; 8 DV3 (Type III) (as also reported in Table 2). In particular, in the transversal direction are working 18 DV1 devices in the higher positions (in the A3 substructure: 10 DV1, 4 horizontal and 6 oblique; in the A4 substructure 8 DV1, 2 horizontal and 6 oblique) and 17 DV2 in the lower positions and along the alignment of the slabs P1 (see Figure 2a) (in the A3 substructure: 4 DV2, 2 horizontal and 2 oblique; in the A4 substructure: 1 DV2 oblique; in contrast with the P1 slabs 12 DV2, 4 horizontal and 2 oblique) while 8 DV3 devices are working in the longitudinal direction positioned between A3-A2 and A4-A5.

failure at the non-structural elements especially through the connection of both the recon‐ structed and the remained brick cladding with the reinforced concrete structures to avoid local failure due to the overturning of wall portion. The partition walls inside the building have been completely substituted with plasterboard fixed to aluminum profiles well anchored to the structural elements. Even if in the other two buildings (A and C) it was not necessary the use of seismic devices for structural protection, the approach followed in the work done in the Edifice A through the direct action of a non profit organization, have been extended to the

**Figure 9.** Reconstruction at Edifice A: a) damage scenario involving the facade, b) reconstruction of the facade, c)

Advanced Applications in the Field of Structural Control and Health Monitoring After the 2009 L'Aquila Earthquake

http://dx.doi.org/10.5772/55438

223

close view of the installed device, d) dissipative truss structures.

**5. Structural health monitoring research activities at university of L'Aquila**

A group of researchers of CERFIS (www.cerfis.it) with complementary skills is conducting a wide plan of activities in the field of dynamic testing under environmental loading and structural health monitoring for a series of buildings, with strategic or historical value, at L'Aquila. In the following a synthetic description of the most challenging findings is reported.

In order to achieve adequate level of confidence on the structural dynamic behaviour of the studied buildings a schedule of consequent activities are currently performed: (*i*) on-site dynamic testing under environmental actions with standard equipments [5,9,11,30]; (*ii*) finite element modelling based on exhaustive survey and material testing; (*iii*) definition of SHM-WSN sensor features; (*iv*) laboratory dynamic testing on 1:3 scaled frame in order to validate

other cases making realizable the return to the campus in the 2013 spring semester.

**Figure 8.** Nonlinear viscous damper placement: a) transversal; b) longitudinal.

Figure 8 shows a transversal and a longitudinal section where the protective devices are installed. In particular in Figure 8 it can be noted that a pair of DV3 devices is positioned at each of the two last level working in contrast between the substructures A3 and A2 thanks to the presence of a relevant seismic joint (depth= 20cm).

Figure 9 summarizes some relevant information such as: the large damage scenario appearing in the morning of April 7, 2009 immediately after the earthquake at the main facade of the Edifice A (Figure 9a); the facade completely rebuilt in a picture taken during the reconstruction (September 2011); of the same period two pictures presenting a close view of a DV1 horizontal device in the P1 zone (Figure 9c) and the four alignments of the dissipative trusses that following the perspective belongs the first one to the substructure A3 followed by two alignments in the P1 zone and completed by the last alignment which is the first one for the sub-structure A4 (Figure 9d). It can be noticed that in the last alignment due to the presence to the light stairs coming from the under floor the horizontal device is missed, this occurrence justifies the even total number of installed devices.

Together with the main structural seismic protection, here illustrated, the rehabilitation of the Edifice A has been conducted through the use of several technological applications to avoid Advanced Applications in the Field of Structural Control and Health Monitoring After the 2009 L'Aquila Earthquake http://dx.doi.org/10.5772/55438 223

Figure 7 clarifies the location of the 43 devices: 18 DV1 (Type I ); 17 DV2 (Type II) ; 8 DV3 (Type III) (as also reported in Table 2). In particular, in the transversal direction are working 18 DV1 devices in the higher positions (in the A3 substructure: 10 DV1, 4 horizontal and 6 oblique; in the A4 substructure 8 DV1, 2 horizontal and 6 oblique) and 17 DV2 in the lower positions and along the alignment of the slabs P1 (see Figure 2a) (in the A3 substructure: 4 DV2, 2 horizontal and 2 oblique; in the A4 substructure: 1 DV2 oblique; in contrast with the P1 slabs 12 DV2, 4 horizontal and 2 oblique) while 8 DV3 devices are working in the longitudinal direction

(a) (b)

Figure 8 shows a transversal and a longitudinal section where the protective devices are installed. In particular in Figure 8 it can be noted that a pair of DV3 devices is positioned at each of the two last level working in contrast between the substructures A3 and A2 thanks to

Figure 9 summarizes some relevant information such as: the large damage scenario appearing in the morning of April 7, 2009 immediately after the earthquake at the main facade of the Edifice A (Figure 9a); the facade completely rebuilt in a picture taken during the reconstruction (September 2011); of the same period two pictures presenting a close view of a DV1 horizontal device in the P1 zone (Figure 9c) and the four alignments of the dissipative trusses that following the perspective belongs the first one to the substructure A3 followed by two alignments in the P1 zone and completed by the last alignment which is the first one for the sub-structure A4 (Figure 9d). It can be noticed that in the last alignment due to the presence to the light stairs coming from the under floor the horizontal device is missed, this occurrence

Together with the main structural seismic protection, here illustrated, the rehabilitation of the Edifice A has been conducted through the use of several technological applications to avoid

**Figure 8.** Nonlinear viscous damper placement: a) transversal; b) longitudinal.

the presence of a relevant seismic joint (depth= 20cm).

justifies the even total number of installed devices.

positioned between A3-A2 and A4-A5.

222 Engineering Seismology, Geotechnical and Structural Earthquake Engineering

**Figure 9.** Reconstruction at Edifice A: a) damage scenario involving the facade, b) reconstruction of the facade, c) close view of the installed device, d) dissipative truss structures.

failure at the non-structural elements especially through the connection of both the recon‐ structed and the remained brick cladding with the reinforced concrete structures to avoid local failure due to the overturning of wall portion. The partition walls inside the building have been completely substituted with plasterboard fixed to aluminum profiles well anchored to the structural elements. Even if in the other two buildings (A and C) it was not necessary the use of seismic devices for structural protection, the approach followed in the work done in the Edifice A through the direct action of a non profit organization, have been extended to the other cases making realizable the return to the campus in the 2013 spring semester.
