Dataset of real-time hybrid simulation testing of multiple shape memory alloy based structural control devices

Description This dataset contains data from experimental testing of two shape memory alloy (SMA) wire based dampers, including training data from the SMA wires, sinusoidal cyclic loading tests of the dampers and real-time hybrid simulation (RTHS) testing of the dampers (see 'Load_Protocol_v1.0.0.pdf'). The experiments were carried out using the RTHS setup at the RWTHDynLab of the Chair of Structural Analysis and Dynamics (LBB) at RWTH Aachen University. The experimental campaign was developed to investigate the performance of the designed large-scale shape memory alloy wire dampers (SMAWDs), focusing on the interaction between SMAWDs and the surrounding structure using RTHS, as well as to validate the damping efficiency and reusability of SMA-based control devices under seismic loading. Two large-scale SMAWD prototypes were tested using RTHS, with the dampers as the experimental substructure. The test structure represented both two-dimensional and three-dimensional steel frame structures subjected to multiple earthquake excitations with different spectral properties and magnitudes. The main objectives of the test campaign were: Evaluation of dampers' dynamic characteristics. Investigation of the influence of SMA dampers on the dynamic response of the structures. Investigation of different damper configurations and their effectiveness under various loading conditions. Investigation of the reusability of the dampers after strong earthquakes events. Test procedure The SMA wires used for the SMAWDs were manufactured by HMETech, Germany. They have a diameter of 3 mm and consist of nitinol with approximately 55.7 % nickel. The austenite finish temperature is 10 °C. No additional heat treatment was performed. All tests in this data set are performed with 800 mm wires from the same batch. Before using them in the damper, the wires are subjected to cyclic tensile loading to stabilize the hysteresis and reduce the occuring residual deformation during the RTHS. In the present study, 20 consecutive cycles to a maximum strain of 5 % with a strain rate of 0.5 %/s were carried out. For these tests, an INSTRON tensile testing machine was used. The INSTRON specifications are: load capacity: +/- 160 kN stroke: +/- 250 mm working frequency range: 0 to 20 Hz sampling frequency: 5 Hz. The SMAWDs are constructed in the internal laboratory. All steel parts consist of S355 cold drawn steel. The wires are installed with achor grips and wedges manufactured by Paul Maschinenfabrik, Germany. The damper design and realized prototype is shown in ShapeMemoryAlloyWireDamper_v1.0.0.pdf. The dampers are tested with a different number of wires, different loading rates and amplitudes. Sinusoidal tests are performed with two MTS actuators.The actuators' specifications are: 201.45 MTS actuator dynamic stroke: +/- 250 mm load capacity: 443 kN working frequency range: 1 to 50 Hz MTS 500 kN 551.23F-01 load cell included in the actuator sampling frequency: 512 Hz. Following the identification of the damper properties, RTHS tests were performed in both two-dimensional and three-dimensional structures. The surrounding structure was simulated as the numerical substructure in the finite element analysis program OpenSees while the SMAWDs were tested as the experimental substructure in the laboratory. The resulting displacements and restoring forces at the interface of numerical and experimental substructures were calculated, measured and synchronized in real-time, thus the experimental obtained force feedback of the SMAWDs was included in the next time step of the numerical simulation in OpenSees.The RTHS specifications are: simulation PC: OpenSees and OpenFresco for hybrid finite element simulation real-time target PC: Matlab® and Simulink for predictor-corrector and delay-compensation algorithms controller PC: actuator control 2x 201.45 MTS actuator 3x SCRAMNet cards (one in each PC) to minimize communication delay. The two-dimensional structure is a steel moment resisting frame structure with three bays and three stories. The story height is 2000 mm and each bay has a width of 2400 mm. The structure is fixed with pinned supports on ground level. All columns and beams are realized as HEA 200 and IPE 160 steel profiles, respectively. The strong axis of each profile is used. In each node, an additional point mass of 1 ton is added. Two damper configurations were tested: in Config. 1, only one damper with eight wires is installed in the first story middle bay of the structure. in Config. 2, two dampers with four wires each are installed in the first and second story in the middle bay of the structure. The three-dimensional structure is an extension of the two-dimensional structure with two bays of depth. The dimensions and sections correspond to the two-dimensional structure. The weak axis is in y-direction. On ground level, the structure is pinned with fixed rotation in the weak direction. One and two tons point masses are added to each node, leading to an asymmetric weight distribution. Two damper configurations were tested: in Config. 1, two dampers are installed in the first story in the main (x) direction in the middle bay, one in the front and one in the back of the structure. in Config. 2, one damper is installed in the first story in the x- and y-direction each. The damper configurations are shown in 2D_Structure_v1.0.0.pdf and 3D_Structure_v1.0.0.pdf, respectively.Three different earthquakes are applied as ground acceleration in the numerical structures: Northridge (1994), Kreta (2021) and Kobe (1995). The earthquakes are applied both in x- and y-direction. For the two-dimensional testing, only the stronger (x) direction of each earthquake is used. They are scaled to 0.15 g and 0.30 g peak ground acceleration, while maintaining the original ratio of x- and y-direction, and filtered to a passband of 1 to 20 Hz. The total test procedure is reported in Load_Protocol_v1.0.0.pdf. Files ShapeMemoryAlloyWireDamper_v1.0.0.pdf damper design and photo of realized prototype. Data_v1.0.0.zip contains all data files according to the load protocols. the experimental data is provided as .csv files for each load protocol. contains the unscaled earthquake ground accelerations used for RTHS. Load_Protocol_v1.0.0.pdf file listing the test procedure. 2D_Structure_v1.0.0.pdf illustrating the numerical two-dimensional substructure for RTHS. 3D_Structure_v1.0.0.pdf illustrating the numerical three-dimensional substructure for RTHS. SMA_Wire_Time_Histories_v1.0.0.pdf including plots of the measurement data of the tensile loading of the SMA wires. SMAWD_Time_Histories_v1.0.0.pdf including plots of the measurement data of the cyclic loading and RTHS testing of the SMAWDs. File format of the datasets The data is stored in .csv files, where each file contains the following columns: 01_SMA_Wires Time: time in (s) from the start of testing (time step: 0.2 s) Strain: strain in (%) of the SMA wires. Stress: stress in (MPa) of the SMA wires. These data files can easily be uploaded using the pandas library in Python. For example by: import pandas as pd df = pd.read_csv('Wire_01.csv') Time = df["Time"] Strain = df["Strain"] Stress = df["Stress"] 02_SMAWD Time: time in (s) from the start of testing time step for sinusoidal loading: 1/512 s time step for RTHS: 10/1024 s Actuator1_Disp: displacement signal in (mm) of the first actuator / SMAWD Actuator1_Force: force signal in (kN) of the first actuator / SMAWD Actuator2_Disp: displacement signal in (mm) of the second actuator / SMAWD Actuator2_Force: force signal in (kN) of the second actuator / SMAWD These data files can easily be uploaded using the pandas library in Python. For example by: import pandas as pd df = pd.read_csv('07_4wires_sin20mm_0p5Hz.csv') Time = df["Time"] Actuator1_Disp = df["Actuator1_Disp"] Actuator1_Force = df["Actuator1_Force"]Actuator2_Disp = df["Actuator2_Disp"]Actuator2_Force = df["Actuator2_Force"] Important: The displacement and force signal of the second actuator is only included in test series with two SMAWDs (see LoadProtocol_v1.0.0.pdf). 03_Earthquakes Time: time in (s) from the start of testing (time step: 10/1024 s) Acceleration: acceleration in (g) with the original ratio of x and x excitation for each earthquake. These data files can easily be uploaded using the pandas library in Python. For example by: import pandas as pd df = pd.read_csv('Northridge1994X.csv') Time = df["Time"] Acceleration = df["Acceleration"] Contact: Please send your enquiries regarding the real-time hybrid simulation set-up to [e-mail]. Further information can be found on our website. Usage/License The data is licensed under CC BY-SA 4.0. If you have used our data and are publishing your work, we ask you to please reference this database by its DOI. Acknowledgement We gratefully acknowledge the funding of the RWTHDynLab including the RTHS setup by the German Research Foundation (DFG) under grant numbers INST 222/1161-1 FUGG and INST 222/1403-1 FUGG.