Real-time quality monitoring for friction stir welding AA2219-T845 aerospace aluminium alloy via model-based spectral analysis
- Baraka A. ,
- Panoutsos G. and
- Cater S.
- Baraka A. ,
- Panoutsos G. and
- Cater S.
Repository
Description
An investigation is presented into a systematic model based, real time process monitoring framework to analyse the performance and behaviour of the friction stir welding (FSW) process and into the framework's ability to predict weld quality thresholds for different levels of process conditions in welding of AA 2219-T845 aluminium alloy plates (thickness 76 mm). The modelling dataset was obtained from experimental FSW trials on AA 2219-T845 using a 38 mm TriFlat weld tool with different levels of tool rotational speed and travel speed in order to determine the welding process envelope. Principles of human like information granulation in granular computing and computational intelligence were applied to build a data driven model to facilitate real time forecasting of quantitative markers of weld quality. Part quality thresholds were extracted from the frequency spectra of axial and transverse feedback forces. Linguistic based feedback was gained via an Interval Type-2 radial basis function model to describe process dynamic behaviour.
11th International Symposium, 17-19 May 2016, Session 10B: Monitoring-Inspection, Paper 01
Related Records
