Dataset collected using a control strategy on a triaxial platform
Description
This dataset presents experimental results from a multiaxial vibration test campaign conducted on a triaxial electrodynamic shaker platform. The aim was to assess how different acceleration-control strategies influence the dynamic response of the platform and cross-axis effects. Four closed-loop strategies were investigated: (1) single-point control using one triaxial accelerometer, (2) control using three single-axis accelerometers, one per axis, (3) distributed control using three triaxial accelerometers, and (4) multi-point response with rectangular control through three accelerometers. Acceleration measurements were collected during sweep-sine and random vibration tests over the 5-2000 Hz frequency range. All data were acquired on the LMN triaxial vibration test bench, which enables coupled or uncoupled excitation along three orthogonal axes. Signal generation, closed-loop control, and data acquisition were performed using Spectral Dynamics Jaguar hardware and software. The instrumentation consisted of five triaxial accelerometers and three single-axis accelerometers mounted on the platform. Sweep-sine tests were conducted at 1.5 g with a sweep rate of 1 octave per minute, whereas random vibration tests were performed at 1 g-RMS for 5 min. The dataset includes raw sweep-sine spectra (.frq), frequency response functions (.frf), random-vibration power spectral density files (.psd), and MATLAB post-processed figure files (.fig).
Files
Steps to reproduce
The dataset was obtained during an experimental campaign carried out on a triaxial air-cooled electrodynamic shaker system composed of three orthogonally mounted shakers (TIRA/Kokusai/Spectral Dynamics) installed at the Mechanical Laboratory of Normandie (LMN), INSA Rouen Normandie. The platform was controlled using a Spectral Dynamics Jaguar ACP acquisition and control unit, which was used to configure the excitation signals, perform closed-loop control, acquire the response measurements, and compute the frequency-domain outputs recorded in the dataset. A magnesium cube table was mounted on the platform and used as the fixture interface. Eight accelerometers were installed on the test bench: five triaxial accelerometers (PT1 to PT5) mounted on the top surface of the cube and three single-axis accelerometers (PT6X, PT6Y and PT6Z) mounted on the side surfaces. This instrumentation was used both for response measurements and for the implementation of the four control strategies investigated in the study: (S1) single-point control using one triaxial accelerometer, (S2) control using three single-axis accelerometers, one per axis, (S3) distributed control using three triaxial accelerometers, and (S4) multi-point rectangular control using three accelerometers per axis. For all tests, the excitation frequency range was set from 5 to 2000 Hz. Sweep-sine tests were performed at an acceleration level of 1.5 g with a sweep rate of 1 octave per minute. Random vibration tests were performed using a white-noise PSD profile at 1 g RMS for 5 min. In random vibration tests, the Jaguar controller was configured through a PSD matrix including auto-PSDs and cross-PSDs; relative phase and coherence between excitation axes were defined in the controller settings. In swept-sine tests, relative phase between the input excitations was also specified in the Jaguar software. During each run, the Jaguar system recorded frequency-domain quantities directly. The dataset was then organised into raw sweep-sine spectra (.frq), frequency response functions (.frf), random-vibration power spectral density files (.psd), and MATLAB post-processed figure files (.fig). Raw sweep-sine files contain frequency, amplitude, and phase for the 18 acquisition channels. FRF files contain frequency together with magnitude and phase values, and PSD files contain frequency and PSD values for the 18 channels. To reproduce the dataset generation procedure, a user would need a triaxial shaker platform capable of coupled or uncoupled excitation over 5-2000 Hz, a multichannel controller equivalent to the Spectral Dynamics Jaguar ACP, and an instrumentation layout comparable to the one described above. The same control-channel selections, excitation profiles, frequency range, acceleration levels, and test durations should be used for each of the four strategies reported in Table 3 of the article.
Institutions
- Institut National des Sciences Appliquées Rouen NormandieNormandy, Saint-Étienne-du-Rouvray