Pre Conference Workshop on Advanced Digital imaging

Transcription

1 Pre Conference Workshop on Advanced Digital imaging Uwe Zscherpel, BAM Federal Institute for Materials Research and Testing, Division 8.3 Radiological Methods, Berlin, Germany 1

2 Content: 1. Measurement principle: profile measurement in image data 2. Length calibration and correct edge positions 3. Wall thickness measurement in projection radiography 4. Wall thickness measurement in penetration direction 5. Measurement of flaw dimensions in weld inspection 6. Conclusions Software: Isee! available from 2

3 1. Measurement principle: profiles in image data Three problems: a) Images have pixels, must be calibrated in length (mm) b) Radiographic images are unsharp, correct edge position? c) Monitor displays only 256 gray values, but image could contain more (e.g. 16 bit = gray values) Demo with Isee! 3

4 3. Wall thickness measurement in projection radiography classical tangential set-up of TRT: typical exposure: radiation source, Ir 192 profile line insulation pipe line film, imaging plate, or flat panel 4

5 Praktical application in the field: BASF Projection Radiography Example : distribution station film source on-site exposure, Ir 192 radiograph 5

6 Projection Radiography visual evaluation of wall thickness error sources: w - wrong density range - inaccurate edge detection (unsharpness) - wrong magnification factor 6

7 Projection Radiography PC based wall thickness measurement input exposure parameters wall thickness result table with measurement regions tangential wall thickness algorithm wall thickness deviation in measurement region 7

8 synchronous display of positions in profile and image! 8

9 Projection Radiography PC generated measurement protocol 9

10 Demonstration with Isee! and stepped pipe Ir 192 exposure CR image pipe OD 148mm steps with 2 10 mm wall thickness 10

11 Limits of Tangential Radiography (TRT): L max W L max = 2W sqrt(d a /W 1) 300 kv 200 kv Energy: L max 100 kv : 10 mm 200 kv : 30 mm 300 kv : 40 mm 400 kv : 50 mm Se-75 : 60 mm Ir-192 : 75 mm Co-60 : 120 mm 11

12 2006: Validation of wtscope, CR-Tower and Ir-192 As measurement method according to ISO BAM BASF GE S&IT 12

13 2010: BAM certification of GE S&IT Rhythm WallThickness Modul versus wtscope: 13

14 2 Standard drafts available : pren 16407: See contributions of S. Burch at the WCNDT conference 14

15 4. Wall thickness measurement in penetration direction Second pipe inspection technique: 2. Corrosion assessment from local intensity (film density) changes (Double Wall Technique, DWT) 15

16 local corrosion: wall thickness differences in penetration direction (DWT) source pipe erosion w- w detector intensity change, I attenuation law: I w = I 0 e -µ eff w µ eff - effective attenuation coefficient relative measurement: 1. Calibration: 2. Measurement: 2 step algorithm µ eff = ln ( I REF / I IQI ) / w IQI w = ln ( I REF / I MEAS ) / µ eff 16

17 DWT for local corrosion measurement At first calibration signal at measurement point signal at reference point 1. Calculation of effective attenuation coefficient µ eff known hole depth known pipe wall thickness w at reference point Result: µ eff = f(w) dep. on radiation energy and wall thickness differences 17

18 Demonstration with Isee! and stepped pipe Penetrameter function Ir 192 exposure CR image pipe OD 148mm steps with 2 10 mm wall thickness 18

19 Comparison Film / CR DWT with Ir 192: 0,06 Result of wtscope validation for CR Tower by BAM / BASF / GE S&IT effective attenuation coefficient [1/mm] 0,05 0,04 0,03 0,02 0,01 µ eff CR Film reason: scatter sensitivity! pipe wall thickness [mm] Accuracy in practice: not better than 10 % of wall thickness! 19

20 Basis: ISO Steel ISO Al 5. Measurement of flaw dimensions in welds (formerly EN 12517) NDT of welds acceptance levels for radiographic testing Part 1: Steel, Ni, Ti & alloys Part 2: Al & alloys 20

21 Extension of Isee! for flaw measurement and evaluation Interactive measurement of pore diameters Modular: Visualization of indications, interactive verification and correction 21

22 8 mm steel Clustered porosity: Evaluation of area 28mm² + 78 mm² / 1500 mm² = 7% Acceptance level 2, quality level C of EN ISO

23 Interface to Isee! via ASCII file (*.ini format) Open for external automatic flaw detection and evaluation! 23

24 PhD 2006 O. Alekseychuk, BAM : Automated crack detection, visualization by ISee! (rectangle and crack length) 24

25 2008 Moduls for automatic flaw detection Automatic Defect Detection and Identification Processor (ADDIP), collection of special image processing algorithms for analysis of industrial radiographs ADDIP developed as programming environment for easy testing of algorithms for detection and classification of flaw indications in radiographic images TU Szczecin, Poland exposure detection 25

26 6. Conclusions 1. Measurement of dimensions only on raw data and not on the displayed monitor image (LUT clipping!) 2. Suitable calibration necessary (absolute dimension for TRT, µ eff for DWT) 3. Corrosion measurement methods validated as RT measurment method acc. to ISO (measurement value AND measurement uncertainty) 4. Standards for corrosion inspection in preparation at CEN: pren part 1 (TRT) and part 2 (DWT), will go to ISO? 5. Interactive measurement of flaw dimensions in ISee! possible, evaluation according to ISO supported 6. No general automatic flaw detection algorithms available, special algorithms (porosities, elongated cracks) developed and tested, simple interface to ISee! 26

27 The End 27