Select your language

Goals:
• Import, registration and integration of multimodal and multiscale non-destructive testing data
• Extraction and Visualisation of relevant material properties and defect structures. Besides conventional segmentation techniques, machine learning techniques (Deep Convolutional Neural Networks) will be developed, trained and tested. From the generated multimodal, multiscale NDT data, properties of pores, cracks, delaminations, impurities, disbonds, or fiber bundles and fiber waviness as well as volume-porosity and defect density will be quantified and categorised.
• Development of Virtual Reality (VR) widgets for comparative and ensemble visualisation concerning design optimisation and NDT- data analysis. Visualisation techniques for user friendly analysis for the examination and aggregation of the big number of extracted features in the defined space over time and in higher dimensions for the quanititative visualisation of the data will be developed.
• For defect monitoring and trend monitoring, comparing visualisation techniques will be implemented, in order to allow a detailed analysis for direct comparision of a smaller amount of datasets on the one hand and an analysis of complete ensembles of datasets on the other hand.

Description of the content:
The main goal in this work package is to import all the different NDT-modalities in one software and then automatically interpret and visualise specific features and values using appropriate data analysis techniques. The focus in this WP will be on the integration of new techniques which are rarely used in the industrial NDT methods.

Methods:
These new techniques will be implemented in open_iA (https://github.com/3dct/open_iA), an open source tool from the FHOOE. open_iA is an open source program for the visual analysis and processing of volume data, with the focus on industrial computed tomography. open_iA handles various volume data formats as well as different surface data formats. It allows the slice-by-slice navigation in the 2D views and 3D navigation and intersecting planes in the 3D view. open_iA can easily be expanded and it is the central development platform of the research group for computed tomography.

 

 

Funding

This project is funded by the following FFG-programm: TAKE-OFF Das österreichische Luftfahrtprogramm.
Project numberr: 874540
Titel: Platform for predictive evaluation of aerospace components with multimodal, multiscale inspection.