1. Chalmers University of Technology Advanced NDT Mathematical modelling of the ultrasonic phased array technique within the project Quantification of the reliability of flaw detection (NDT) using probability of detection (POD) based on synthetic data Håkan Wirdelius Advanced NDT /SCeNDT Chalmers

2. Chalmers University of Technology Advanced NDT NDE capabilities Probability of detection curve (POD) Flaw and object variables NDE variables (method, equipment, procedure and process) Calibration Acceptance/decision variables (procedure) Human factor Reliability Capability (POD) Reproducibility (calibration) Repeatability (NDE process control in application)

3. Chalmers University of Technology Advanced NDT When is Quantitative NDE required? Structural integrity – assumes that flaws of an assumed design size will not propagate into size that will induce failures in service => dependent on detection and removal of flaws larger than this size New fatigue and fracture/damage tolerance criteria New applications or use of a design – life extension of an in-service component or introduction of damage tolerance in life-cycle management (also new components) Defect depth (mm) Stress amplitude (MPa)

4. Chalmers University of Technology Advanced NDT Mathematical modeling? Test pieces and artificial defects (static) Volumetric defects: cheap and available parameters can be well specified Crack like defects (notches): ? Real cracks: expensive, time demanding and few available parameters can be specified Mathematical modeling (dynamic) Cost-effective Time saving Individual parameters can be altered Enables Worst Case studies

5. Chalmers University of Technology Advanced NDT Mathematical modelling of UT The simSUNDT program (freeware) consists of a Windows ® -based pre- processor and postprocessor UTDefect used as a mathematical kernel Model of ultrasonic backscattering due to grain growth in a welded region The result can be read by a number of commercial analysis software Reduce number of test pieces and synthetic defects simSUNDT

6. Chalmers University of Technology Advanced NDT The simSUNDT software

7. Chalmers University of Technology Advanced NDT Dimensions of used test specimen 250 mm 50 mm 230 mm 0°5°10°15°20° 130 mm 50 mm 40 mm 20 mm 10 mm20 mm SDH  2 mm WP 2Experimental verification of the model:

8. Chalmers University of Technology Advanced NDT B-scan comparison, SDH, 45° S probe Direct S-wave: 0 dB 38.5 mm 33  s 67  s 0 mm

9. Chalmers University of Technology Advanced NDT B-scan comparison, SDH, 45° P probe Direct P-wave: 0 dB 60 mm 15.08  s 33.92  s 0 mm

10. Chalmers University of Technology Advanced NDT B-scan comparison, 0° backwall tilt, 45° P probe 0 mm 11.64  s 60 mm 39.68  s P-wave corner echo: 2.1 dB P- and S-wave corner echo: 4.9 dB S-wave corner echo: 5.1 dB Strip like-Single Rectangular crack Strip like-Subdiv. 7 elements

11. Chalmers University of Technology Advanced NDT WP 1Development of simulation model: Develop a mathematical model of a phased array probe and incorporate it into the simSUNDT software. The correlation between prescribed boundary conditions and resulting wave propagation within the material will be validated. WP 2Experimental verification of the model: Experimentally verify and proper validate the developed model for a very specific case. The case has been chosen in order to ensure reliability when it comes to relevant parameters such as defect depths, sizes and characteristics. Quantification of the reliability of flaw detection for non-destructive techniques using probability of detection (POD) based on synthetic data

12. Chalmers University of Technology Advanced NDT WP 1Development of simulation model: E.A. Ginzel1 and D. Stewart2, PHOTO-ELASTIC VISUALISATION OF PHASED ARRAY ULTRASONIC PULSES IN SOLIDS, Proc WCNDT 2004 Quantification of the reliability of flaw detection for non-destructive techniques using probability of detection (POD) based on synthetic data

13. Chalmers University of Technology Advanced NDT WP 1Development of simulation model: Quantification of the reliability of flaw detection for non-destructive techniques using probability of detection (POD) based on synthetic data Physical dimensions of the phased array probe p e g H A A = length in steering direction (active) H = width (passive plane) e = length of element p = pitch (centre to centre distance) g = spacing between elements N = number of elements

14. Chalmers University of Technology Advanced NDT WP 1Development of simulation model: Quantification of the reliability of flaw detection for non-destructive techniques using probability of detection (POD) based on synthetic data  j  0  j  p(r) 00 A0A0 A1A1 A2A2 A -1 A -2 x Figur 2

15. Chalmers University of Technology Advanced NDT Dimensions of used test specimen WP 2Experimental verification of the model: 20 4060 10 80 300 5 Side-Drilled Holes 6 Flat- Bottom Holes all  = 2.4 mm 20 10 35 5 Depth 20, 40 and 60 Depth 10, 30 and 50 60

16. Chalmers University of Technology Advanced NDT WP 3Procedure for synthetic data based POD: Develop a technique that uses mathematical models in order to develop a data base that will be used as the basis for a POD curve. The procedure shall be for a specific NDT task based on experiences made in WP1 and 2. WP 4Experimental validation of the synthetic POD: Experimentally validate the above evolved POD curve. This in order to optimize the inspection procedure and also to make it as a convention tool in the development of POD characteristics. Quantification of the reliability of flaw detection for non-destructive techniques using probability of detection (POD) based on synthetic data Indicators of NDT capability: a 50, a 90 and a 90/95 (95% confidence)