Reliability of Routine Radiographic Film Evaluation - an Extended ROC Study of the Human Factor
- Many radiographic exposures and film evaluations are made daily in a typical industrial NDT laboratory. But questions remain regarding the precise probability of detecting discontinuities, including the reliability of each individual inspector or laboratory. Although most experienced inspectors of the laboratories evaluate each radiograph, the actual quantitative reliability of each of these inspectors remains unknown. In order to estimate the performance of representative inspectors and laboratories involved in NDE, volunteer laboratories from Croatia and Hungary conducted a series of round-robin inspections using X-ray films of welds. While the RRT is still in progress, the authors try to summarize the results of the second round and discuss the evaluation method. The results of the first round had been reported already. The focus is here to present the new results in terms of additional inspectors and a new more field oriented evaluation scheme. So far the results of a total ofMany radiographic exposures and film evaluations are made daily in a typical industrial NDT laboratory. But questions remain regarding the precise probability of detecting discontinuities, including the reliability of each individual inspector or laboratory. Although most experienced inspectors of the laboratories evaluate each radiograph, the actual quantitative reliability of each of these inspectors remains unknown. In order to estimate the performance of representative inspectors and laboratories involved in NDE, volunteer laboratories from Croatia and Hungary conducted a series of round-robin inspections using X-ray films of welds. While the RRT is still in progress, the authors try to summarize the results of the second round and discuss the evaluation method. The results of the first round had been reported already. The focus is here to present the new results in terms of additional inspectors and a new more field oriented evaluation scheme. So far the results of a total of 35 inspectors of the different laboratories were evaluated. Those inspectors were selected who have done usually the work of radiographic film interpretation. They have from a half up to 35 years professional experience. For better overview we created four groups: maximum 4 year-experience, from 5 to 12 year-experience, from 15 to 25 year- experience and from 26 to 35 year-experience. Discussing the circumstances of the RRT with the participants, many of them expressed, they could not exactly follow the prescription of the test. According to their professional experience the small discontinuities can be omitted because of the strictest acceptance level accept them. So the most experienced evaluators did not write the small gases and slags into the list, which yield a wrong reliability of detection. As a consequence we should accomplish two types of evaluation: taking into consideration of all discontinuity and only the bigger once they are over the acceptance level. In addition we analyzed the true values in a critical review. These developments were made and the new results of the evaluation of the testers were more precise and reliable. The corresponding results are evaluated and presented in a ROC diagram where the p(TP) is plotted against the p(FP). The p(TP) means the probability of a true positive indication (correct defect finding) and p(FP) means the probability of a false positive (false alarm) indication. From the mean value of all inspectors maximum operation points we can estimate a representative overall reliability in terms of a correct indication rate of about 68 % (64 % new) and a false call rate of about 17 % (8 % new). This is a reasonable result within the strict 1 cm score.…
Autor*innen: | F. Fücsök, Christina Müller, Martina Rosenthal |
---|---|
Dokumenttyp: | Beitrag zu einem Tagungsband |
Veröffentlichungsform: | Graue Literatur |
Sprache: | Englisch |
Titel des übergeordneten Werkes (Englisch): | Proceedings of the 8th ECNDT |
Jahr der Erstveröffentlichung: | 2002 |
Herausgeber (Institution): | Asociación Española de Ensayos no Destructivos |
Verlag: | European Federation for Non-Destructive Testing |
Verlagsort: | Barcelona |
Veranstaltung: | 8th ECNDT |
Veranstaltungsort: | Barcelona, Spain |
Beginndatum der Veranstaltung: | 2002-06-17 |
Enddatum der Veranstaltung: | 2002-06-21 |
URL: | http://www.ndt.net/article/ecndt02/429/429.htm |
Bemerkung: | Geburtsname von Rosenthal, Martina: Scharmach, M. - Birth name of Rosenthal, Martina: Scharmach, M. |
Verfügbarkeit des Dokuments: | Physisches Exemplar in der Bibliothek der BAM vorhanden ("Hardcopy Access") |
Bibliotheksstandort: | Sonderstandort: Publica-Schrank |
Datum der Freischaltung: | 19.02.2016 |
Referierte Publikation: | Nein |