Dataset from dynamic shake-table testing of five full-scale single leaf and cavity URM walls subjected to out-of-plane two-way bending

This paper provides information related to the sensor measurements obtained from five different unreinforced masonry (URM) walls subjected to incremental dynamic shake-table tests at EUCENTRE, Pavia, Italy. This information has been made available to assist in the development and calibration of analytical and numerical models intended to simulate the out-of-plane (OOP) two-way bending response of URM walls. For further interpretation of the sensor recordings, and for a detailed discussion on the observed seismic performance of the specimens, the reader is referred to the article entitled “Experimental Response of URM Single Leaf and Cavity Walls in Out-Of-Plane Two-Way Bending Generated by Seismic Excitation” [1]. Videos documenting the failure of each specimen are also available on YouTube [2].


Data
Data corresponding to incremental dynamic testing of five full-scale URM walls is provided. Each specimen was densely instrumented with various sensors measuring accelerations and displacements throughout the testing sequence. The locations of these sensors and their operating status throughout the testing sequence (Table 4 of reference article [1]) is provided in Tables 1e5 and Figs. 2e6. Figs. 2e6 also provide information on the mass distribution assumed to compute the provided inertial force associated with each specimen. The recorded data is organized into folders with each folder corresponding to a single specimen. Each folder contains several files with a single file containing the data of all instruments recording in a particular test. The name of each file provides information about which test of the testing sequence it contains data of. Within each .txt file, the first column corresponds to a time vector whereas all other columns correspond to instrument readings. All data acquired was filtered from frequencies higher than 50 Hz. All recordings of accelerations and forces are provided in units of g and kN, respectively while displacements are given in mm.

Experimental design, materials, and methods
This article presents the experimental data obtained from incremental dynamic testing of five fullscale URM walls subjected to two-way bending OOP seismic excitation. Four full scale URM walls Specifications  Value of the data The data provides detailed information about the dynamic response of URM walls in two-way bending. It may serve as a benchmark for the development as well as calibration of numerical models to simulate the response of URM in the out-ofplane direction (e.g. Refs. [4e6]). The data can also be used to validate simplified analytical methods to assess the response of URM in the out-of-plane direction.
The data may serve to evaluate the effectiveness of the test setup.
(single leaf as well as cavity) were also previously tested by the authors under one-way bending excitation [3]. The data acquired from such tests may represent a benchmark for the development as well as calibration of numerical models to simulate the response of URM in the out-of-plane direction (e.g. Refs. [4e6]). This experimental data can also provide valuable insights on the comparison between results observed for walls with that of local two-way-bending failures observed in full-scale buildings (e.g. Tomassetti et al. [7]). The five tested walls represent the first full-scale URM walls tested up to collapse that have been reported in literature. Three of these specimens were constructed in calcium silicate masonry (CS), one in clay (CL) masonry and another one was a cavity wall consisting of an inner leaf in CS and an outer leaf in CL masonry connected to each other by metal ties. All dynamic tests were carried out at the uniaxial shake table of EUCENTRE Pavia. The main input motions used in this part of the campaign corresponded to second floor accelerograms recorded either from a building prototype tested by Graziotti et al. [8] or from a calibrated numerical model of the tested building [5]. Low amplitude random excitations (RN) were used in between test runs to identify the dynamic properties of the specimens A summary of the naming adopted and boundary conditions associated with the tested specimens can be observed in Fig. 1. Characteristics of the employed input motions and their sequence along with the employed scaling factors are summarised in Table 4 of the reference article [1]. Every specimen was densely instrumented with sensors that recorded the dynamic response at various locations. The instrumentation adopted for each specimen consisted of accelerometers, potentiometers, wire potentiometers and a 3D optical acquisition system (used for all specimens except CS-005-RR). The location of all the instrumentation adopted for each specimen was decided based on the boundary conditions envisaged and correspondingly expected deformed shapes. Accelerometers were installed on the OOP panel of the specimen in order to record acceleration-time histories. Additional accelerometers were also installed at the specimen foundation, top beam, rigid frame and the return walls. Potentiometers were used to measure relative displacements associated with various locations of the specimen. Wire potentiometers attached to the rigid frame in several locations were used to record horizontal displacements relative to the shake table. Potentiometers were also adopted to record the relative displacements between the main panel and the return walls. All data acquired was filtered from frequencies higher than 50 Hz. Accelerations and forces are provided in units of g and kN, respectively; displacements are given in mm. For each specimen, a folder is created named as the specimen: the folder containing data from all the tests corresponding to the second specimen is named as "CS-000-RF". This folder contains .txt files for each test named as "TestT#" where "T#" refer to the same quantity provided in the testing sequence included in the reference article [1] (Table 4). Within each .txt file, the first column corresponds to a time vector whereas all other columns correspond to instrument readings. The instrument recordings contained in different columns for each specimen as well as coordinates of their exact location are provided in Tables 1e5. Figs. 2e6 shows graphically the employed instruments for each specimen. In these tables and figures, Acc.: refers to accelerometer, WP: refers to wire potentiometer, Pot.: refers to potentiometer and Opt./Marker: refers to optical acquisition. Please note that moving towards higher intensities of shaking WP measurements were replaced with those obtained from a 3D optical (Opt./ Marker) acquisition system.   Tables 1e5 indicate also the mass associated with each accelerometer for the calculation of the inertial force of the OOP panel (provided in the .txt files). This associated lumped mass distribution changed throughout the testing sequence with the development of cracks and the adopted distribution throughout the testing sequence can also be found in Figs. 2e6. More details about how the inertial force was calculated can be found in the reference article [1]. It is worth noticing as in the case of CS-000-RF specimen during the last stages of testing (Test 19e22), due to lower number of instruments recording, a linear acceleration amplification was assumed along its height and half of the relative acceleration recorded by accelerometer 12 (marked in grey in Fig. 3) was assigned to the centre of the cracked panel. This was done in order to not overestimate the inertial force associated with the specimen. Additionally, the column "Offline" in Tables 1e5 indicates test numbers (Table 4 of reference article [1]) when a particular instrument was not recording.
As an illustrative example, with reference to Table 2, column 3 of the file "Test6.txt" in the folder "CS-000-RF" corresponds to recordings of the 'Foundation Acc.' when specimen CS-000-RF was subjected to FEQ2-DS3 scaled at 50% i.e. T#6 in Table 4 of the reference article [1].

Acknowledgments
This paper describes an activity that is part of the ''Study of the vulnerability of masonry buildings in Groningen" project at the EUCENTRE, undertaken within the framework of the research program for hazard and risk of induced seismicity in Groningen sponsored by the Nederlandse Aardolie Maatschappij BV. The authors would like to thank all the parties involved in this project, namely EUCENTRE and University of Pavia (DICAr) laboratories that performed the tests, Arup and TU Delft. The valuable

Transparency document
Transparency document associated with this article can be found in the online version at https:// doi.org/10.1016/j.dib.2019.103854.