Nuclear, automotive, oil & gas... Non-destructive testing (NDT) is used in many industrial fields to inspect parts of various types (geometry and materials). While the signals acquired enable 2D images to be reconstructed in real time, the 3D equivalent is much more complex, requiring a very large number of signals and computing power that is impossible to embed in current inspection devices.
For the first time, CEA-List has met a scientific and technical challenge: providing access to real-time, three-dimensional part reconstruction.
The researchers developed new algorithms, based on 3D reconstruction methods in the Fourier domain, to reduce the number of computational operations required to a minimum. The result is a prototype 4D (real-time 3D) imager.
Promises kept: the imager was successfully tested and validated on artificial defects (cracks, porosities and shrinkage) produced by additive manufacturing in steel blocks on the Additive Factory Hub (AFH) platform. It even achieved unprecedented performance in visualizing porosities as small as 0.6 mm in diameter, locating them in the volume with an accuracy of the order of a tenth of a millimeter.
CEA-List researchers are now focusing on increasing the performance and reducing the computing time of these new imaging methods, without any loss of quality.
Visit the site: