NEW APPROACH TO ACOUSTIC EMISSION TESTING METALLIC PRESSURE VESSELS

Size: px

Start display at page:

Download "NEW APPROACH TO ACOUSTIC EMISSION TESTING METALLIC PRESSURE VESSELS"

Laurence Chapman
5 years ago
Views:

1 NEW APPROACH TO ACOUSTIC EMISSION TESTING OF METALLIC PRESSURE VESSELS 11th European Pressure Equipment Conference Munich ANVIXED sarl copyright

2 Aim of the presentation: tti Review the basic principles i of AE examination List the main factors impacting the quality of an acoustic emission test Observe to what extent they are taken into account in conventional AET practice Offer a new approach which corrects the effects of theses factors and improve quality of AE tests on metallic pressure vessels ANVIXED sarl copyright

3 1st part: Basis review of Acoustic Emission Testing ANVIXED sarl copyright

4 Definition of Acoustic emission phenomenon energy liberation in the form of transient elastic waves, resulting from microdisplacements inside a material under stress A E source load ANVIXED sarl copyright

5 Sources of acoustic emission crack growth local plastic deformation corrosion phase changes inclusion fracture stress concentration are main generating mechanisms But only when evolving with applied load ANVIXED sarl copyright

6 Detection principle preamp sensor Spherical wave AE event Applied load ANVIXED sarl copyright

7 Hit (burst emission case) ANVIXED sarl copyright

8 AE testing installation at on a pressure e vessel ANVIXED sarl copyright

9 Typical pressure diagram Pressure increase segment Pressure hold segment 2 nd sequence: Completing test up to hydrotest pressure End of 1t 1st sequence: Loading up to design pressure ANVIXED sarl copyright

10 Detection and location mesh CALCULATION OF ACOUSTIC EVENT LOCATION sensor C acoustic event located at X,Y sensor A wave propagation, velocity = V sensor B hit detected by sensor A t0 hit detected by sensor B t1 hit detected by sensor C t2 X,Y = f(t0, t1, t2, V, location of sensors A, B and C) ANVIXED sarl copyright

11 Planar (hyperbolic) location principle AE Event hyperbolas ANVIXED sarl copyright

12 Planar location of Located Event (LE) on a sphere ANVIXED sarl copyright

13 Acoustic Emission signals (1/2) signals emitted by the structure during testing are physical events clusters of physical events are concentrations or sources dues to defect evolution signals detected by sensors are hits a physical event releases waves which reach one or several sensors: one or several hits belong to an acoustic event ANVIXED sarl copyright

14 Acoustic Emission signals (2/2) the processing system transforms hits into detected events (Ev) events for which at less 3 sensors are reached are «located events» (LE) the main characteristics of an AE event, determined on the 1st hit, are: arrival time amplitude energy location All data recorded during the test is stored in the primary file (.PRI) ANVIXED sarl copyright

15 2nd part: main factors impacting results & quality of AE tests ANVIXED sarl copyright

16 Main factors affecting AE test results & grading (1/6) Group Goup1-Vessel 1 - geometry 2 - material 3 - metal volume 4 - residual stress distribution Group 2 - Process 1 - fluid 2 - maximum previous applied pressure 3 - in-service i conditions i ANVIXED sarl copyright

17 Main factors affecting AE test results & grading (2/6) Group 3- Sensors array 1 - distance between sensors 2 - type of mesh Group 4- Sensors type 1 - sensor sensitivity 2 - sensor bandwidth ANVIXED sarl copyright

18 Main factors affecting AE test results & grading (3/6) Group 5- Acquisition settings 1 - acquisition threshold 2 - other channel parameters 3 - front-end filter set-up Group 6- Loading cycle 1 - delta P above max. previous pressure 2 - pressure rate 3 - hold periods duration ANVIXED sarl copyright

19 Main factors affecting AE test results & grading (4/6) Group 7- Data processing 1 - primary file qualification 2 - filtering 3 - event builder adjustments 4 - data analysis method Group 8- Classification protocol 1 - activity based criteria 2 - intensity based criteria 3 - Kaiser effect criterion 4 - Felicity ratio criterion ANVIXED sarl copyright

20 Main factors affecting AE test results & grading (5/6) Developping item 7.4: data analysis method 1 - Hits analysis signature 2 - Application of an amplitude evaluation threshold 3 - Mixing i acoustic analysis with classification mechanism 4 - Zonal location limitations ANVIXED sarl copyright

21 Main factors affecting AE test results & grading (6/6) Developping item 7.4: data analysis method 5 - Correlation between zonal and planar locations 6 - Correction of activities versus acquisition threshold 7 - Correction of amplitudes versus distance ANVIXED sarl copyright

22 How are these factors managed in current practices? There is a lot of factors conditioning conditioning performances of AE tests: numeric results, acoustic signature, and classification What is more, acquisition & processing settings can be specific to a group of channels / sensors Within the framework of this presentation, we will only browse a few points ANVIXED sarl copyright

23 3rd part: conventional approach and its limitations ANVIXED sarl copyright

24 Item 6: AE test pressure cycle (1/2) pressure of the AE test must be in usual practice 5or 10 % over the maximum historic pressure PMA PMA is the max pressure value applied to the vessel during a specified period of time (generally 12 months) overpressure from PMA aims to provide a sufficient test sensitivity but this usual practice don t guaranteea a convenient stress value ANVIXED sarl copyright

25 Item 6: AE test pressure cycle (2/2) Example from CERN helium tanks 250 m3 / DP=26 bar hydrotest carried out in fab plant up to 39 bar a month later, after installation on site: pneumatic AE examination AE test pressure to apply according to the rule: 43 bar! maximum pressure of the AE test: 26 bar results: satisfactory AE test sensitivity ANVIXED sarl copyright

26 Item 7.3: Event builder parameters influence FHCDT (ms) 1,6 2,5 3 3, DT1X max (ms) DTNX Max (ms) 16 1,6 16 1,6 16 1,6 16 1,6 16 1,6 16 1,6 16 1,6 cumul SIGS Σ SIGS cumul Ev1 cumul Ev (Ev1+) cumul Ev2 Σ Ev cumul Ev cumul Ev3 cumul Ev cumul LE Σ LE ANVIXED sarl copyright

27 Item 7.4.1: Hits or events analysis? (1/2) ANVIXED sarl copyright

28 Item 7.4.1: Hits or events analysis? (2/2) Hits are «raw» signals, and: hits number is depending on sensors array and number of sensors the hits population includes non significant data as mechanic noise, electromagnetic interference, The hit population cannot describe the global acoustic signature of the vessel The hit population has to be assembled by the «event builder» to create the events population ANVIXED sarl copyright

29 Item 7.4.2: Analysis with amplitude thresholds (1/2) ANVIXED sarl copyright

30 Item 7.4.2: Analysis with amplitude thresholds (2/2) Green dark curve: no filter > true acoustic behavior Green clear curve: amplitude filter (threshold 50 db) > light distorsion of activity evolution versus time Red curve: amplitude filter (threshold 65 db) > strong distorsion of activity evolution versus time ANVIXED sarl copyright

31 Item 7.4.4: Zonal locationboundaries (1/2) the boundary is a line drawn in the middle of the areas between zones thus sensors located on the junction line between two parts of the vessel don t permit to pick up independantly AE activity it from these parts ANVIXED sarl copyright

32 Item 7.4.4: Zonal location boundaries (2/2) ANVIXED sarl copyright

33 Item 8: Classificationcriterias criterias (1/7) ASTM E 569: applicable to planar location Activity: trend Intensity: relative value Directly applicable Not provider or code dependant Easy to implement ANVIXED sarl copyright

34 Item 8: Classificationcriterias criterias (2/7) ASME V Art 12: applicable to zonal location Activity: value beyond a limit in a given period value beyond a limit for hits with Amp> Ao trend Intensity: trend Not directly applicable Code dependant Not easy to implement ANVIXED sarl copyright

35 Item 8: Classificationcriterias criterias (3/7) MONPAC: applicable tozonal location Activity: no direct criterion Intensity: Relative energy value of a pack of 200 events: H(t) Absolute value of the 10 hits with max energy: Sav Directly applicable Easy to implement ANVIXED sarl copyright

36 Item 8: Classificationcriterias criterias (4/7) EN 14584: applicable to planar location Activity: activity / amplitude mixed criterias trend Intensity: trend Not directly applicable Provider dependant Not easy to apply ANVIXED sarl copyright

37 Item 8: Classificationcriterias criterias (5/7) EN 15495: applicable to zonal location Activity: activity / amplitude mixed criterias Intensity: trend Not directly applicable Provider dependant Not easy to implement ANVIXED sarl copyright

38 Item 8: Classification criterias (6/7) GBP AFIAP: applicable to zonal location Activity: activity / amplitude mixed criterias Kaiser effect / Felicity ratio trend Intensity: trend applicable to planar location (sources) Activity: value above a given value Intensity: no requirements Directly applicable Not easy to implement ANVIXED sarl copyright

39 Item 8: Classificationcriterias criterias (7/7): discussion each protocol is dependant on numerous factors influencing the measurement of acoustic activity and energy three of them are dependant on external requirements Under these conditions, it is difficult to think they can conclude to the same diagnostic! ANVIXED sarl copyright

40 4th part: new approach to pressure vessels AE examination ANVIXED sarl copyright

41 What does new approach really offer? requirements for AE tests providers directions to follow by practitioners principles for specifications, procedures and reporting tools for supervisors of AE tests new concepts for standards improvements ANVIXED sarl copyright

42 General AE data processing (1/2) step 1 - Qualified tests applied to the primary data file (PRI) step 2 - Qualified population extraction from the file (PRI) step 3 - Optimized events building step 4 - Events population analysis according to the different levels of acoustic signature ANVIXED sarl copyright

43 General AE data processing (2/2) step 5 - Acoustic behavior analysis joining zonal and planar location methods (acoustic emission domain) step 6 - Acoustic signatures interpretation (physical domain) step 7 - Application of classification criterias step 8 - Follow-up ND T > according to damage grading step 9 - Reporting All the steps are clearly identified and separately yprocessed ANVIXED sarl copyright

44 Taking into account the main influence factors: reference pressure reference pressure is more realistic than previous maximum operating pressure (MOP) the reference pressure (Pref) is the pressure beyond which the the vessel is considered as overstressed il may be lower than MOP if the vessel has been submitted to various stress situations: temperature stress, lifting, i shipping, i on-site installation i stresses, in-service cycling stress the value is determined by the AET provider and the vessel user, considering the former history of the vessel ANVIXED sarl copyright

45 Taking into account the main influence factors: overloading versus pressure uncertainty pressure uncertainty t is generally around 2 % for an overstress uncertainty limited to 5 %: over pressure uncertainty is 40 %! for an overstress uncertainty limited to 10 %: overpressure uncertainty is 20 % Overloading must reach at less 20 % of the reference pressure ANVIXED sarl copyright

46 Taking into account the main influence factors: Kaiser test & Felicity ratio significant the Kaiser effect test is a major test which is at the basis of AE application to non destructive examination To be valid the Kaiser test demands a sufficient pressure decrease, not compatible with a small overpressure span Relevant to the Kaiser effect is the Felicity ratio which h can be easily and accurately computed ANVIXED sarl copyright

47 Taking into account the main influence factors: overloading span Activities N are dependant on the overpressure span (Pspan) applied to the vessel during the acoustic emission test To compare the activities, one must introduce a factor to «standardise» the pressure span effect Then the activities have to be expressed as the ratio of N to Pspan: K= N / Pspan ANVIXED sarl copyright

48 Taking into account the main influence factors: sensor sensitivity a factor of sensor sensitivity has to be introduced, to do so, a reference sensitivity must be defined then, the sensor sensitivity is expressed as a value in db 6dB for example means that the sensor is 2 times more sensible than the reference one ANVIXED sarl copyright

49 Taking into account the main influence factors: Amplitude / distance correction this point only applies to planar location of AE sources this correction allows to take into account the «mesh density» and the attenuation curve in the metal up to now, only EN considers this correction ANVIXED sarl copyright

50 Taking in account the main influence factors: Acquisition threshold the acquisition threshold disturbs measurements: value of AE activity cutting of low energy hits distorsion on shape parameters of hits false location of events the amplitude distribution of signals curves allow to correct the effect of acquisition threshold on: AE activities measurements AE energy measurements ANVIXED sarl copyright

51 Taking in account the main influence factors: Acquisition threshold ANVIXED sarl copyright

52 Taking in account the main influence factors: Acquisition threshold ANVIXED sarl copyright

53 Taking in account the main influence factors: Acquisition threshold ANVIXED sarl copyright

54 Taking in account the main influence factors: Acquisition threshold ANVIXED sarl copyright

55 Taking in account the main influence factors: Acquisition threshold Measured values: amplitude average value: standard deviation: activity: 43 db 10 db 80 Ev Corrected values: amplitude average value: standard deviation (inchanged): activity: 40 db 10 db 100 Ev ANVIXED sarl copyright

56 Amplitude distribution based analysis: Source type 1 ANVIXED sarl copyright

57 Amplitude distribution based analysis: Source type 1 Is analysis significant? Corrected average amplitude value? Corrected activity value? ANVIXED sarl copyright

58 Amplitude distribution based analysis: Source type 2 ANVIXED sarl copyright

59 Consequences: is AET a sampling method? the numerous factors that minuse activity measurement lead to wonder if AE testing is a sampling method taking only a part of indications In fact, the cutting of a part of the event population does not comply with rules of sampling the detection of indications resulting of discontinuities is in general not affected ; only the weak amplitude events are cancelled ANVIXED sarl copyright

60 Consequences: Is AET a qualitative or quantitative method? the numerous factors that minuse activity measurement lead to wonder if AE testing is a sampling method taking only a part of indications In fact, the cutting of a part of the event population does not comply with rules of sampling the detection of indications resulting of discontinuities is in general not affected ; only the weak amplitude events are cancelled ANVIXED sarl copyright

61 Ways to a new grading protocol use of «normalized» activities & amplitudes use of energy as intensity measurement instead if amplitudes criteria application to events and not to hits energy computation with «normalized» values enrichment of sources signatures with zonal activities contribution ANVIXED sarl copyright

62 Conclusion AET is a powerfull tool applied to integrity evaluation of pressure vessels, but new approach implementation will allows: to compare the tests to improve the efficiency and universality through an advanced grading mechanism to go further in standardisation ANVIXED sarl copyright

63 Thanks for your kind attention If you have any question We are also pleased to inform you of the writing of our booklet: «Contribution i to acoustic emission i testing of metallic pressure vessels» (issued in october 2015) Ordering information on request from ANVIXED ANVIXED sarl copyright

JOHANN CATTY CETIM, 52 Avenue Félix Louat, Senlis Cedex, France. What is the effect of operating conditions on the result of the testing?

JOHANN CATTY CETIM, 52 Avenue Félix Louat, Senlis Cedex, France. What is the effect of operating conditions on the result of the testing? ACOUSTIC EMISSION TESTING - DEFINING A NEW STANDARD OF ACOUSTIC EMISSION TESTING FOR PRESSURE VESSELS Part 2: Performance analysis of different configurations of real case testing and recommendations for