1. Structural Dynamics & Vibration Control Lab., KAIST 1 Structural Vibration Control Using Semiactive Tuned Mass Damper Han-Rok Ji, Graduate Student, KAIST, Korea Yeong-Jong Moon, Ph. D. Candidate, KAIST, Korea Chun-Ho Kim, Professor, Joongbu University, Korea In-Won Lee, Professor, KAIST, Korea The Eighteenth KKCNN Symposium on Civil Engineering

2. Structural Dynamics & Vibration Control Lab., KAIST 2 Introduction Semiactive Tuned Mass Damper Numerical Analysis Conclusions CONTENTS

3. Structural Dynamics & Vibration Control Lab., KAIST 3 Introduction Tuned Mass Damper ― widely used mechanical damping device ― Simple and efficient vibration control system ― No external power, energy dissipation, inherent reliability ― Restricted performance resulted from the fixed parameters Semiactive Tuned Mass Damper ― Alternative device of conventional TMD ― Improved control performance with stability of TMD ― High robustness and adaptability

4. Structural Dynamics & Vibration Control Lab., KAIST 4 Objective Analytical study on semiactive TMD using MR damper for mitigating the vibration of structures Application of various semiactive control algorithms to MR damper Robustness analysis for the semiactive TMD system

5. Structural Dynamics & Vibration Control Lab., KAIST 5 Semiactive Tuned Mass Damper m1m1 k1k1 c1c1 m2m2 c(t) x1x1 x2x2 k2k2 m1m1 ― Equation of Motion (1) SDOF system with semiactive TMD – Controllable damping device is installed in the place of passive dashpot. – Produce the additional control effect to the primary structure.

6. Structural Dynamics & Vibration Control Lab., KAIST 6 c0c0 c1c1 k1k1 k0k0 c1c1 c0c0 k0k0 k1k1 Modified Bouc-Wen Model Bouc-Wen ― modified Bouc-Wen model (Spencer et al., 1997) (2) Dynamic model of MR damper

7. Structural Dynamics & Vibration Control Lab., KAIST 7 Semiactive Control Algorithms ― on-off velocity based groundhook control ― on-off displacement based groundhook control ― clipped optimal algorithm ― maximum energy dissipation algorithm

8. Structural Dynamics & Vibration Control Lab., KAIST 8 On-off velocity based groundhook control (Koo et al. 2003) ― Based on velocity of primary system (v 1 ) and TMD (v 2 ) (3) On-off displacement based groundhook control (Koo et al. 2003) (4) ― Based on velocity of primary system (v 1 ) and TMD (v 2 ) displacement of primary system (x 1 )

9. Structural Dynamics & Vibration Control Lab., KAIST 9 ― linear optimal controller and clipped algorithm F c : desired damper force by optimal controller F d : measured damper force Clipped optimal algorithm (Dyke et al, 1996) (5) Maximum energy dissipation algorithm (Jansen and Dyke, 2000) (6) ― Control voltage is determined so that the structure dissipates the maximum energy F d : measured damper force

10. Structural Dynamics & Vibration Control Lab., KAIST 10 Three-story shear building MR damper m TMD = 150 kg, k TMD = 36,401 N/m Input earthquake excitations ― amplitude scaled El Centro, Hachinohe earthquakes Numerical Analysis

11. Structural Dynamics & Vibration Control Lab., KAIST 11 Value c oa 21.0 N  sec/cm aa 140 N/cm c ob 3.50 N  sec/cm  V bb 695 N/cm  V koko 46.9 N/cm  363 cm -2 c 1a 283 N  sec/cm  363 cm -2 c 1b 2.95 N  sec/cm  V A301 k1k1 5.00 N/cmn2 xoxo 14.3 cm  190 sec -1 Parameters of MR damper (Spencer et al., 1997) c0c0 c1c1 k1k1 k0k0 c1c1 c0c0 k0k0 k1k1 Modified Bouc-Wen model Bouc-Wen ― maximum damper force : 1,500 N ― minimum voltage : 0 V ― maximum voltage : 2.25 V

12. Structural Dynamics & Vibration Control Lab., KAIST 12 Response of building model TMD passive off passive on on-off DBG on-off VBG clipped optimal MEDA Scaled El Centro (PGA 0.10g) J1J1 0.380.390.500.350.390.360.39 J2J2 0.37 0.520.330.340.320.34 J3J3 0.450.470.500.44 0.430.44 Scaled Hachinohe (PGA 0.08g) J1J1 0.350.360.510.350.400.360.40 J2J2 0.35 0.490.320.390.340.39 J3J3 0.380.410.470.360.370.350.37 J 1 : normalized peak floor displacement J 2 : normalized peak interstory drift J 3 : normalized peak acceleration

13. Structural Dynamics & Vibration Control Lab., KAIST 13 ― El Centro earthquake ― Hachinohe earthquake ― The efficiency of semiactive TMD is slightly better than that of TMD. ― Passive on mode has the worst performance. Normalized value ― Evaluation criteria under two earthquakes

14. Structural Dynamics & Vibration Control Lab., KAIST 14 Robustness Analysis Response with stiffness matrix perturbation : amount of perturbation (-15%, -10%, -5%, +5%, +10% and +15%) ― Perturbed stiffness matrix (7) ― Real structures can have structural uncertainties in many reasons. ― Control performance of TMD is restricted considerably due to off-tuning effect. — Stiffness perturbation is considered to verify the robustness of the semiactive TMD

15. Structural Dynamics & Vibration Control Lab., KAIST 15 Time (sec) Interstory drift (cm) Acceleration (m/sec 2 ) Time history with +15% stiffness perturbation under Hachinohe earthquake ― The maximum and RMS values with semiactive TMD are reduced compared with that of conventional TMD.

16. Structural Dynamics & Vibration Control Lab., KAIST 16 Normalized peak drift (J 2 ) Normalized peak acceleration (J 3 ) ― Overall performance of semiactive TMD is better than that of TMD. ― Efficient algorithm : on-off DBG control for interstory drift clipped optimal control for acceleration ― Evaluation criteria under El Centro earthquake

17. Structural Dynamics & Vibration Control Lab., KAIST 17 Normalized peak drift (J 2 ) Normalized peak acceleration (J 3 ) ― Semiactive TMD is superior to conventional TMD. ― On-off DBG and clipped optimal algorithm have sufficient robustness. ― Evaluation criteria under Hachinohe earthquake

18. Structural Dynamics & Vibration Control Lab., KAIST 18 ― Various semiactive control algorithms are adopted and the performance of each algorithm is evaluated. ― Semiactive TMD system shows slightly better performance than conventional TMD system. ― Analytical study on semiactive TMD with MR damper is performed. Conclusions

19. Structural Dynamics & Vibration Control Lab., KAIST 19 ― Sufficient robustness is obtained under the structural perturbation with semiactive TMD. ― The on-off displacement based groundhook theory and clipped optimal algorithm is appropriate algorithm for semiactive TMD system.