Application Report 11/15/5471 Company: Kienle Spiess

Transcription

1 11/15/5471 Company: Kienle Spiess Birger Niehaus, 12/07/2015 Radioscopic (DR) Inspection and Computer Tomography (CT) of Squirrel-Cage Motors Technology with Passion

2 System description MU 2000-D CT X-ray tube: Y.TU 160-D06 Focal spot size: 0.4 mm (Al material) (EN12543) 1.0 mm (Cu material) X-ray settings: 160 kv / 4.3 ma (Al) 160 kv / 9.3 ma (Cu) Diaphragm with filter: 0.5 mm (DR) 1.5 mm (CT) Focus detector distance: 900 mm Magnification: 1.9 to 2.9 X-ray detector: Y.Panel HDR Settings: 200 µm pixel size 67 (Al) / 200 (Cu) ms/frame 50 (Al) / 100 (Cu) % sensitivity CT- scan time: 7 min 50 sec 2

3 Photo of test parts and setup The upper image shows the parts and the lower one the fixing on the turntable of the MU DR: Demonstration of defect sizes that can be seen in Al as well as in Cu front side areas. CT: Demonstration of anomalies that can be found in Al front side areas. Wire Image Quality Indicators (IQI) acc. to standard EN ISO (EN462-1) are used to evaluate the sizes of smallest anomalies, which can be detected. 3

4 DR: X-ray images Original image of Part Al1HU Pos. 1 with IQI: W10ALEN. Integration time i = 5 s Magnification m 2.1 Diameters of Aluminum Wire- IQI Wire No W 10 W 11 W 12 W 13 W 14 W 15 W 16 Ø [µm] A lot of anomalies can be seen. But due to the high complexity of the part the gray value scaling in the 8 bit image is very smooth and small details cannot clearly be detected. 4

5 DR: X-ray images Original image of Part Al1HU Pos. 1 with IQI: W10ALEN, filtered by Bandpass 1. W 13 Integration time i = 5 s Magnification m 2.1 Diameters of Aluminum Wire- IQI Wire No W 10 W 11 W 12 W 13 W 14 W 15 W 16 Ø [µm] By using the YXLON filter Bandpass 1 the gray value gradient was taken out of the images and the inner structures became more precise. As the Al region is very spongy the wires are hard to define in the material. 5

6 DR: X-ray images W 10 W 15 To show which wire the system is able to detect in this wall thickness an Image with IQI was subtracted by an Image without IQI. Now W 15 (Ø = 125 µm) can be named. 6

7 DR: X-ray images HDR (High Dynamic Radioscopy) filtered image of Part Al1HU Pos. 1 in combination with motion detection and averaging. In this mode the live X-ray images will be processed. When the image is low-noise enough it can just be saved so that no additional image enhancement process is needed. Averaging time i 5 s The HDR- image is even a bit sharper than the integrated and Bandpass 1 filtered image. 7

8 DR: X-ray images Original image of Part Al2HU Pos. 1 with IQI: W13ALEN, filtered by Bandpass 1. W 15 Integration time i = 5 s Magnification m 2.8 Diameters of Aluminum Wire- IQI Wire No W 13 W 14 W 15 W 16 W 17 W 18 W 19 Ø [µm] In the image W 15 is partly visible. In the Image with IQI subtracted by Image without IQI Wire W16 can clearly be seen. 8

9 DR: X-ray images Original image of Part CU1HU Pos. 1 with IQI: W10CUEN, filtered by Bandpass 1. W 13 Integration time i = 10 s Magnification m 1.9 Wire W 13 is visible. 9

10 DR: X-ray images HDR filtered images of Part CU1HU Pos. 2 (upper image) and Pos. 4 (lower image). Measurement of anomalies with the image enhancement system Image Averaging time i 10 s Magnification m 1.9 A lot of anomalies can be detected in both images. In the upper image a pore is measured with the magic wand tool to 0.82 mm². In Pos. 4 the length of an anomaly is measured to mm and the width to 5.84 mm. 10

11 DR: X-ray images Part CU2HU : Bandpass 1 filtered image of Pos. 1 (upper image) and HDR filtered image of Pos. 4 (lower image). Integration /Averaging time i 10 s Magnification m 2.0 Again, a lot of anomalies can be detected in both images. 11

12 DR: Screenshot of user interface Image 3500 The user interface of the image enhancement system Image 3500 shows the HDR filtered image of Part CU1HU Pos. 4 including the length and width measurement. 12

13 CT images 3D image of the AL ring area of part Al1HU. High resolution scan Number of projections: 1024 Even for CT of the Al ring areas a 160 kv system is sufficient. 13

14 CT images All four Scenes (top-down, right-left, frontback, volume) of part Al1HU. High resolution scan Number of projections: 1024 A lot of anomalies can be detected in every projection. 14

15 CT images Part Al1HU : defect detection in all four scenes with an indicator set to an anomaly. High resolution scan Number of projections: 1024 A lot of pores up to a volume of even more than 250 mm³ are detected. 15

16 CT images Part Al2HU : All four Scenes with an indicator set to an anomaly. High resolution scan Number of projections: 1024 Again, beside the marked defect a lot of other anomalies can be detected in every projection. 16

17 CT: Screenshot of VG Studio user interface The user interface of the Volume Graphic (VG) Studio shows all four Scenes of part Al1HU with an indicator set to a defect on the left and the tools tree on the right side. 17

18 Screenshot of Windows Media Player Slice movies top- down were created from parts Al1HU and Al2HU to demonstrate the distributions of anomalies in the slices. After copying the videos on the hard disk of your PC you can open it via a normal video player as e.g. Windows Media Player. 18

19 Summary and interpretation DR: A 160 kv system in combination with the 16 bit flat panel detector Y.Panel HDR and the settings as described above is sufficient even for the front side areas of the Cu- parts. Depending on the material, the wall thicknesses and the structure of the parts wires with a diameter down to 100 µm can be detected. If there are larger Cu- parts than Cu 1 HU a 225kV system is necessary. CT: Even for CT of the front side Al ring areas a 160 kv system is sufficient. A more detailed analysis of the Al parts is possible with 3D, but it takes more time. For the front side areas of the Cu- parts at least 450kV are needed. All X-ray images, the slice movies, the report and the Image viewer are documented in the corresponding data packet. To get an impression of the enormous gray value range of the detector it is recommended to open the 16 bit images via the Image viewer. Clicking on the left mouse button, keeping the mouse button pressed and dragging a square can do gray value scaling from white to black for this rectangle. 19