SEISMIC REHABILITATION OF A REINFORCED CONCRETE FRAMED STRUCTURE BY THE USE OF TUNED MASS DAMPERS AND VISCOUS DAMPERS

Tulei Elena, Cretu Dan, Ghindea Cristian ABSTRACT The paper presents two modern rehabilitation methods of a five storey reinforced concrete framed structure built in Bucharest in 1965. The first method consists in placing tuned mass dampers on the roof floor. In the second method, viscous dampers are attached to steel braces placed on the diagonal of the panels from the lateral spans of the building. The structural response obtained when these two methods are applied, is compared with the structural response obtained in the traditional strengthening solution, which consists in building reinforced concrete walls that work together with the framed structure by means of chemical anchors. The study shows that each dissipative device has a certain destination and has to be designed according to the structural system, soil conditions, earthquake type and the required performance level. Keywords: structural control, tuned mass damper, viscous damper 1. EXISTING STRUCTURE DESCRIPTION A 5 storey reinforced concrete framed structure is analyzed before and after seismic rehabilitation. The plan of the third floor and the transverse cross section are shown in Figure 1. The structure has 9 equal bays of 4.7 m in longitudinal direction and 3 bays of 6.175, 3.4 and 7.175 m in transverse direction. The five storeys are 3.8 m high each. The frames are regular and placed on orthogonal directions. The columns have rectangular cross sections with different dimensions on the height of the building. The slab is 12 cm thick on the lateral spans and 10 cm thick on the middle span. The structure is made of concrete C12/15 with reinforcing steel Fe360 (OB37) and was designed according with the Romanian seismic code of the time, P13-63.
1-1

12

12

12

+15.13

65

50

20 30

60

12 50

65 12

65 12

315

330

380

255 55

395

272.5 55

30

445

25

445

25

445

25

445

25

445

25

445

25

443.5

25 772.5

12

10

12

+11.33

60

582.5 65

30

300 50

30 65

682.5

315

12 40

53

60

60

65

20

20

53

60 40 20 60 40 60 60 23 42 70 70
D

65

703

683

20 315

20 330

20 380 315

445

945 35 360 35 255

25
+ 7.53

60 12

582.5

30

300 10 50

30

682.5 12

65

30

30

20

12 20 10 65 10

300 1745

10 65

315

330

380

300

300

12

10

12

+ 3.73

60 65

577.5

35

300 50

35

677.5 65

315

65

35

220

35

65

20

30

20

30

12 65

12 65
- 0.07

315

330

380

60 12

572.5

40

300 10

40

672.5 12

30 672.5 603 583

65

50

315

330

380

10

30 20

60

60

35

435 470

35

435 470

35

435 470

35

435 470

35

435 470 4258.5

35

435 470

35

435 470

35

435 470

35

435 470

28.5
672.5 300
B C

10

772.5

Fig. 1. The plan of the third floor and the transverse cross section

10

10

40 12

70 - 3.87

562.5

40

300

40

662.5

315

65

315

40

40

The structure strengthening is required due to its high deformability and to the small resistance capacity of the structural members. The presence of masonry partition walls prevented major damages in the structure, but these walls are seriously damaged, having fracture lines at 45o. The analyses according to the present Romanian codes show that the structure presents a high level of vulnerability. In order to evaluate the safety of the building in accordance with the present Romanian codes for new structures, linear static, nonlinear static (pushover) and nonlinear time history analyses have been done. The structure being regular in plan and elevation, only the transverse frame behaviour has been analyzed. Linear static analysis showed that node and storey mechanisms may develop. Due to large displacements and lack of ductility, brittle failure of the concrete members may occur. Node and storey mechanisms can also be identified in the plastic mechanism shown in Figure 2, obtained by nonlinear static analysis. Figure 3 shows the relation between the displacements at the roof level and the base shear force, in the case of mean material strength values. The nonlinear static analysis showed the exceeding of the interstorey drift and resistance capacities, both in service limit state (SLS) and ultimate limit state (ULS). The nonlinear time history analysis showed the limited resistance capacities of the structural members, due to the premature forming of the global plastic mechanism and the large deformability that exceeds the deformation capacities. Raighley structural damping has been considered, the same in all vibration modes, C = a0 M + a1K The coefficients a0 and a1 have been established considering the first two vibration modes with concrete structure damping ratio 1 = 2 = 5% .

Lateral forces + X
700

600

SLS B1

ULS C D

500

Base Shear Force [kN]

400

B
300 200

100

A
-0.05

0 0 0.05 0.1 0.15 0.2 0.25 0.3

Roof displacement [m]

Fig. 2 Pushover analysis: plastic mechanism for design strength values

Fig. 3 Pushover analysis: base shear force roof displacement curve for mean strength values

2. ANALYSES RESULTS AFTER TUNED MASS DAMPER IMPLEMENTATION The classic tuned mass damper (TMD), also known as Dynamic Vibration Absorber, consists of a mass supported by springs. It is usually installed on the roof of the building, in order to take over some of the earthquake energy input, thus reducing the seismic response of the existing structure. The TMD dynamic characteristics have to be calibrated until its oscillations are out of phase with the structure oscillations during the earthquake excitation. The optimal TMD mass (m) and stiffness (k) are determined through numerical simulations, until the lateral displacements of the structure at the roof level are minimum.

The TMD that consist of a mass supported by linear elastic springs without damping is known as the Hooke type TMD (Fig. 4, a). In order to increase the tuning band of the dominant frequencies structure - seismic action, dampers can be added to the classic TMD, thus obtaining the Maxwell type TMD (Fig. 4, b), with the following characteristics: ca = 2ma a , a = 2 TTMD , TTMD T1 TMD stiffness may be chosen ka = 1000caa in order to obtain pure damping. For = 1 , the dampers behave linear elastically. Figure 4, c shows a Zener type TMD, with linear elastic springs and nonlinear dampers.
F = ku

k
a. Hooke type TMD

F ka ca

k F ka ca F

& ) sign (u &) F = k a uk + ca (u

u = uk + uc 0 .2 1
b. Maxwell type TMD

c. Zener type TMD (a + b in parallel)

Fig. 4 TMD types used in the analyses 2.1 Existing structure with TMD The additional mass installed on the roof is calculated only for the movement in the building transverse direction. The TMD calibration has been done for the north-south component of the accelerogram recorded at INCERC-Bucharest during the severe earthquake from 4th March 1977, with PGA = 0.211g (INCERC) and three generated artificial accelerograms (ACC1, ACC2, ACC3). The elastic response spectra for 5% viscous damping for the four accelerograms, in comparison with the spectra from the codes P13-63 and P100-2006 are shown in Figure 5. The following situations have been compared: existing structure without TMD; structure with Hooke type TMD; structure with Zener type TMD, having a = 15% and = 0.4 . Time history analyses have been performed for the accelerogram INCERC, the artificial &&g (t ) = ao sin t , with ao = 1 m/sec2, accelerogram ACC1 and the harmonic accelerogram u

= 2 rad/sec (T T1 = 1.059 sec), 0 t 5 sec. From the analysed cases, the most efficient TMD is the Hooke type TMD, with m = 0.015M and a = s (TMD tuned with the structure). Figure 6 shows the time evolution of the base shear force comparatively for three situations: existing structure without TMD, structure with Hooke type TMD, structure with Zener type TMD having ca = 12 . The Hooke type TMD reduces the lateral displacements and the base shear force, but only after the principal seismic attack (the first peak in the response history). The Zener type TMD reduces the structural vibration amplitude after the principal shock with about 37%, but increases the base shear force. In practice, the TMD needs some time in order to get tuned with the excitation and the structure vibrations.

3.5

2.5

1.5

ACC1 ACC2 ACC4 INCERC-77-NS P100-2006 P13-63

Se/ag
1 0.5 0 0 1 2 3 4 5 6 7

T [sec]

Fig. 5 Normalized absolute acceleration elastic spectra

2500

2000

1500

1000

Fb [kN]

500

0 0 -500 2 4 6 8 10 12 14 16

-1000

-1500

-2000

Time [sec] w/o TMD w TMD w TMD CD=12

Fig. 6 Base shear force time evolution for the accelerogram INCERC, m = 0.015M
2.2. Combined strengthening solution, with reinforced concrete walls and TMD

A natural strengthening solution is to replace the masonry walls with reinforced concrete walls (Fig. 7, a). In order to ensure that the existing structure works together with these r. c. walls, chemical anchors are used. The TMD effect on the dynamic response of this new structure is analyzed (Fig. 7, b). The structural wall is made of reinforced concrete of class C22/25 with E = 30000 N/mm2, and is 0.15 m thick. The classical strengthening solution drastically reduces the interstory drift.

a. without TMD (T1 = 0.27 sec)

b. with TMD

Fig. 7 Strengthened structure

The new structure, considered in three situations (without TMD, with Hooke type TMD and Zener type TMD) has been analysed under the seismic actions represented by the accelerograms INCERC and ACC1. Figure 8 shows the time evolution of the base shear force in the case of the recorded accelerogram INCERC. As in the case of the existing structure without classical strengthening, the TMD has insignificant effects on the new structure. The displacements and the base shear force are slightly and randomly reduced.
Strengthened structure
1500

1000

500

Base Shear Force [kN]

0 0 2 4 6 8 10 12 14

-500

-1000

-1500

Time [sec] w/o TMD 1.5 % M w TMD w/o C 1.5 % M w TMD w C

Fig. 8 Base shear force time evolution for accelerogram INCERC

2.3 Initial structure with TMD, subjected to harmonic excitation

When the initial structure (without reinforced concrete walls) is subjected to harmonic excitation, the Hooke type TMD reduces the maximum lateral roof displacement with about 60% and the base shear force with about 51%. The Zener type TMD leads to smaller reductions, of about 40% both for roof displacements and base shear force. Figure 9 shows the time evolution of the base shear force. The advantage of the Zener type TMD with respect to the Hooke TMD is that it reduces more rapidly the vibrations after the ceasing of the exciting action (for t > 5 sec).

3000

2000

Base Shear Force [kN]

1000

0 0 2 4 6 8 10 12 14

-1000

-2000

-3000

Time [sec]
w/o TMD 1.5 % M w TMD w/o C 1.5 % M w TMD w C

Fig. 9 Base shear force time evolution for sinusoidal accelerogram

3. STRUCTURE REHABILITATION WITH VISCOUS DAMPERS

The strengthening method consists in introducing into the structure devices capable to dissipate the energy induced by the earthquake through deformation. The main passive control devices are: yielding metal devices, friction devices, viscous dampers and visco-elastic dampers. A force proportional with the displacement is developed in the first two devices, which dissipate the energy through hysteretic cycles of loadingunloading. In the last two devices, the force is proportional with their deformation velocity. These devices have an insignificant effect on the dynamic characteristics of the structure (vibration periods), reducing the structural response at seismic actions strictly by transforming the kinetic energy associated to the masses into heat energy. Unlike the base isolation devices, which act in series with the structure, absorbing the energy and filtering the motion before transmitting it to the structure, internal dampers act in parallel with the structure. The induced energy is absorbed by the system structure - dampers and dissipated according to the characteristics of its components. In order to obtain optimal performance, the dampers have to be tuned to the structure, which is difficult, especially because not all buildings behave better by increasing their damping. The use of modern control procedures depends on the structure type, site conditions, and frequency content of the earthquake. The most efficient damper, but the most expensive too, is the damper with viscous fluid. & (t ) , The force developed in a viscous damper is proportional with the deformation velocity u
& sgn(u &) Fd = c u

Figure 10 shows the existing structure with linear viscous dampers ( = 1 ) installed on metallic braces of steel OL37, making Maxwell type devices. Two models have been done. In the first model, the cross section area of the braces is 0.01 m2, and the damper coefficient is c = 1 kNsec/mm. In the second model, the cross section area of the braces is 0.02 m2 and the damper coefficient is c = 10 kNsec/mm.

Fig. 10 Structure with viscous dampers The time evolution of the displacements at the roof level and the base shear force is comparatively shown in Figures 11 and 12 for the initial structure and the structure with internal dampers. A reduction of the lateral displacements and of the base shear force is obtained by the dampers. But the reduction of the internal forces is not substantial and the structure collapses due to the exceeding of the resistance capacities of the transverse frame members. There are cases where the base shear force and the axial forces in columns increase, although the drift is reduced by the structural damping increasing.

Initial structure 6.00E-01

C=10000 kNs/m

C=1000 kNs/m
1000

Initial structure

C=10000 kNs/m

C=1000 kNs/m

800

5.00E-01
600

4.00E-01

400 Bae Shear Force [kN]

Displacement [m]

3.00E-01

200

0 10 -200 11 12 13 14 15 16

2.00E-01

1.00E-01

-400

-600

0.00E+00 10 11 12 13 14 15 16 17
-800

-1.00E-01 Time [sec]

-1000 Time [sec]

Fig. 11 Displacement evolution

4. CONCLUSIONS

Fig. 12 Base shear force evolution

Seismic rehabilitation of buildings that do not satisfy the safety conditions imposed by the present seismic code can be done by different strengthening solutions. The traditional solution consists in: controlling the plastic hinge occurrence, increasing the deformation capacity of the structural members and the limitation of the lateral displacements. Modern solutions use devices for the active, semi active or passive control of the structural response at seismic actions. Passive control consists either in structure base isolation, either in introduction in the structure of dissipative devices, like tuned mass dampers or viscous dampers. The decision for the most appropriate rehabilitation solution has to be taken in accordance with the structural dynamic characteristics, the structure deformation and resistance capacities, and the ground motion type, the earthquake frequency content and the site conditions. The examples show that the passive control devices like TMD and viscous dampers are not capable to improve the performance of the structure analysed in the paper. The classic strengthening solution is the most appropriate to ensure the safety of the reinforced concrete frame structure. The studies done on this building show the TMD inefficiency in reduction the displacements and forces produced by seismic actions. These devices may oscillate in phase with the structure and increase its response. On the other hand, the TMD requires that the structure behaves elastically, which is a condition quite difficult to be satisfied. The TMD efficiency under harmonic excitations tuned to the fundamental oscillation of the building recommends this device for reducing the oscillations under wind actions. When viscous dampers are installed on metallic braces, the base shear force may increase and affect the structure foundations, and the axial force in columns may be modified, reducing their resistance capacity. Analyses done on simple models may prevent possible failure in the tendency of introducing modern rehabilitation solutions.

REABILITAREA SEISMIC A UNEI STRUCTURI N CADRE DIN BETON ARMAT FOLOSIND AMORTIZORI CU MAS ACODAT I AMORTIZORI VSCOI

Lucrarea prezint dou metode moderne de reabilitare a unei cldiri cu 5 niveluri cu structur n cadre din beton armat, construit n Bucureti n 1965. Prima metod const n amplasarea amortizorilor cu mas acordat pe planeul de acoperi al cldirii. n a doua metod, amortizori vscoi sunt ataai la contravntuiri metalice dispuse pe diagonala panourilor din deschiderile laterale ale cldirii. Rspunsul structural obinut n aceste dou metode de consolidare este comparat cu rspunsul obinut n soluia de consolidare tradiional, prin care se construiesc perei de beton armat care conlucreaz cu structura prin intermediul ancorelor chimice. Studiul arat c fiecare dispozitiv disipativ are o anumit destinaie i trebuie s fie proiectat n funcie de sistemul structural, condiiile de teren, tipul de cutremur i nivelul de performan cerut.
REFERENCES

1. Soong T.T., Dargush G.F., Passive Energy Dissipation Systems in Structural Engineering, John Willey&Sons, 1997 2. Bozorgnia Y., Bertero V., Earthquake Engineering: from Engineering Seismology to Performance-Based Engineering, CRC Press LLC, 2004 3. Chopra A. K., Dynamics of Structures, Theory and Applications to Earthquake Engineering, Pearson Prentice Hall International, 2007 4. Paz M., Leigh W., Structural Dynamics-Theory and Computation, Springer Science, 2004 5. Kelly T. E., In Structure Damping and Energy Dissipation Design Guidelines, Holmes Consulting Group LTD, New Zeeland, 2001, www.holmesgroup.com
AUTHORS INFORMATION Dr. Dan Creu (cretud@utcb.ro) is the Head of the Strength of Materials Department from the Technical University of Civil Engineering, Bucharest (Romania). His research interests are in reinforced concrete and steel structures. He is member of IABSE (International Association for Bridge and Structural Engineering) and AICPS (Asociaia Inginerilor Constructori Proiectani de Structuri).

Dr. Elena Tulei (elena@utcb.ro) is lecturer at the Strength of Materials Department from the Technical University of Civil Engineering, Bucharest (Romania). Her research interests are in seismic behaviour of multistorey steel and reinforced concrete structures under seismic actions. She is member of ARIS (Asociaia Romn de Inginerie Seismic). Cristian Ghindea (ghindea@utcb.ro) is assistant professor at the Strength of Materials Department from Technical University of Civil Engineering Bucharest (Romania). As a PhD student in Civil Engineering, since 2002, his research interest is seismic protection of structures with special devices. He is member of ARIS (Asociaia Roman de Inginerie Seismic).