[2C8] Tracking composite plies using simulated realistic ultrasonic fields
N Pilashev
University of Bristol, UK
The detection, characterisation and measurement of out-of-plane wrinkles in high-performance composite laminates is critical to retaining the design strength. Whilst it is often possible to detect wrinkles in carefully acquired B-scan data and even measure the wrinkle severity, this is highly skilled and time-consuming. The process has benefited considerably from recent developments using analytic-signal analysis at the ply resonance frequency to track the particular instantaneous phase that occurs at each resin layer between plies[1]. Most of the modelling work done on ultrasonic analytic-signal ply tracking in composite laminates to date utilises 1D analytical models and thus assumes plane-wave propagation at normal incidence to the plies.
The aim of this paper is to extend this study to include realistic ultrasonic fields from planar or focused transducers. This is achieved through the use of a finite-element model, which allows for the behaviour of the analytic signal to be analysed within a diffraction-limited ultrasonic field. Taking into consideration the non-normal portion of the wave propagating in a transversely isotropic medium such as a composite ply could be key to explaining some discrepancies that occur in experimental cases of ply tracking. Thus, a better understanding of how the analytic signal, and the instantaneous phase and frequency in particular, of a given reflection perturbed by its relative position within a realistic ultrasonic field can help in correcting the depth locations of the resin layers. This will also be crucial for distinguishing between changes in resin-layer thickness and porosity at resin layers in a method currently being developed for 3D distribution mapping of porosity. These two produce similar instantaneous-amplitude responses but different instantaneous-phase responses, which are invertible at the resin-layer depth, so it is crucial to know that depth precisely. The centre frequency and pulse-echo bandwidth of the input pulse are chosen to be close to the ply resonance frequency so that the instantaneous phase becomes locked onto the resin layers. Using this property, a plane-wave field was initially utilised in FE simulations to derive the plane-wave time-of-flight between the internal resin layers of an eleven-ply composite component. This agreed with the analytical model used previously[1]. The results of this model were compared with those obtained from an FE model of a more realistic ultrasonic field with a specified F-number and focal depth in the material in order to examine how the local field position affects the properties of the analytic signal close to the actual resin-layer time-of-flight. The focal point of the transducer was then varied to investigate its influence on the instantaneous phase and consequently on the predicted time-of-flight between resin layers for the same composite component.
Reference 1. R A Smith, L J Nelson, M J Mienczakowski and P D Wilcox, ‘Ultrasonic tracking of ply drops in composite laminates’, Proceedings of the Review of Quantitative NDE, Minneapolis, July 2015. In: AIP Conference Proceedings, Vol 1706, Article No 050006, 2016. DOI: 10.1063/1.4940505
The aim of this paper is to extend this study to include realistic ultrasonic fields from planar or focused transducers. This is achieved through the use of a finite-element model, which allows for the behaviour of the analytic signal to be analysed within a diffraction-limited ultrasonic field. Taking into consideration the non-normal portion of the wave propagating in a transversely isotropic medium such as a composite ply could be key to explaining some discrepancies that occur in experimental cases of ply tracking. Thus, a better understanding of how the analytic signal, and the instantaneous phase and frequency in particular, of a given reflection perturbed by its relative position within a realistic ultrasonic field can help in correcting the depth locations of the resin layers. This will also be crucial for distinguishing between changes in resin-layer thickness and porosity at resin layers in a method currently being developed for 3D distribution mapping of porosity. These two produce similar instantaneous-amplitude responses but different instantaneous-phase responses, which are invertible at the resin-layer depth, so it is crucial to know that depth precisely. The centre frequency and pulse-echo bandwidth of the input pulse are chosen to be close to the ply resonance frequency so that the instantaneous phase becomes locked onto the resin layers. Using this property, a plane-wave field was initially utilised in FE simulations to derive the plane-wave time-of-flight between the internal resin layers of an eleven-ply composite component. This agreed with the analytical model used previously[1]. The results of this model were compared with those obtained from an FE model of a more realistic ultrasonic field with a specified F-number and focal depth in the material in order to examine how the local field position affects the properties of the analytic signal close to the actual resin-layer time-of-flight. The focal point of the transducer was then varied to investigate its influence on the instantaneous phase and consequently on the predicted time-of-flight between resin layers for the same composite component.
Reference 1. R A Smith, L J Nelson, M J Mienczakowski and P D Wilcox, ‘Ultrasonic tracking of ply drops in composite laminates’, Proceedings of the Review of Quantitative NDE, Minneapolis, July 2015. In: AIP Conference Proceedings, Vol 1706, Article No 050006, 2016. DOI: 10.1063/1.4940505
