UTFacultiesETDepartmentsMS3Research ChairsDynamics Based MaintenanceProjectsFinished MSc ProjectsSietze Bruinsma - Development of a Diagnostic Tool Based on Continuous Scanning LDV Output Spectra

Sietze Bruinsma - Development of a Diagnostic Tool Based on Continuous Scanning LDV Output Spectra

(December 2017 – October 2018)

Link to report in repository: http://purl.utwente.nl/essays/76777

SUMMARY

The industry is looking for better diagnostic tools to use for structural health monitoring in operating conditions. A method based on Continuous Scanning Laser Doppler Vibrometry (CSLDV) is proposed to detect damage in a cantilever beam. The CSLDV technique measures the vibration response of a component remotely regarding spatial and temporal information. This fast acquisition technique can be implemented to rapidly detect unhealthy dynamic behaviour in components as a result of stiffness degradation. Current methods using CSLDV require reconstruction of the Operational Deflection Shape (ODS) of the component, as in Figure 1a, the new method devised in this research does not.

Figure 1: ODS and spectrum for Pristine (blue) and Damage (red) cases

 The prismatic, aluminium beam is modelled, multiple damage types and severities are applied. The CSLDV time signal is simulated for every frequency from the resulting vibration responses. From these time signals, characteristic frequency spectra are calculated, seen in figure 1b. The damage detection method proposed in this thesis aims at identifying the deviation in sideband amplitudes as a result of damage, numerical results in Figure 2a. The proposed damage indicator called the Relative Amplitude of the Sidebands to the Total Amplitude Reference (RASTAR) should determine if damage has occurred in the beam. This RASTAR method is then experimentally validated, Figure 2b.

KEY RESULTS

·       A more severe damage resulted in a higher value for the damage indicator.

·       The indicator seems to be most sensitive to the damage at the anti-resonances of the beam.

·       RASTAR was able to detect 1x1x30 mm damage in a 400x10x30 mm beam.

Figure 2: Key results