Critical Argument regarding the European TOFD Draft standard (TOFD). P. Kreier, Innotest, Eschlikon (CH); G. Brekow, BAM, Berlin. Abstract: The ultrasonic testing method time of flight diffraction (TOFD) was applied at selected weld specimens of 15 to 60 mm wall thickness consisting 1mm to 5 mm spark-eroded applied artificial notches. The TOFD results are compared with the detail results of the pulse echo method. The TOFD technique is able to detect defects within the middle section of the wall, but the results for near to surface defects are restricted. That means that weld defects which need to be recorded according to the established standard (e.g. HP 5/3), can not be detected reliably and in conformity to the new European draft standard. (DGZfP proceeding).

We will present the complete lecture in a future issue. Here is just one example the Specimen P5 for which the TOFD technique was not able to measure crack size. Mr. Kreier was concerned that the promised economic benefits of TOFD can not be realized; it will be necessary to use Meanderscanning for the TOFD technique also - resulting in increased test time.

Mr. Hädrich (DIN Berlin) was more in favor of the TOFD draft standard: "Part 6 of the draft standard has not been published yet; changes must be made. That means the 3 year notification deadline has not started yet. During this time changes could be worked out, or the standard could be completely rejected - so everything is still open ".

Jan Verkooijen of AEA Sonomatic BV said that he too had some reservations about the current draft norm, and that it still needs to be optimized. On the defects that were not found during the trials performed, he said that "There is no Non Destructive Examination technique that will find everything in all circumstances". Furthermore he said that certainly by optimization of parameter settings it should be possible to minimize the deficiencies that are found during Dr. Kreier 's examinations. In The Netherlands, a new project is being set up, to optimize the norms for TOFD by validation and qualification in comparison with other techniques. All experts that wish to co-oporate in doing this are invited to take part in this.

Mr. Hecht Archivebild

During the discussion Mr. A. Hecht, BASF Aktiengesellschaft, Ludwigshafen presented some recent crack sizing results performed with ultrasonic methods on a thick-walled pressure vessel weld. A manually applied ultrasonic in-service inspection revealed significant indications within the weld root. Subsequently, a p-scan inspection was applied for accurate documentation of the suspected zones. The echo amplitudes from the root section ranged as high as FBH (flat bottom hole) 3 mm + 20 dB.

Service companies performed further defect analyses ("sizing") at
a) specific zones with the SAFT- method and
b) the total weld was tested using the TOFD method

The SAFT method determined a maximum defect size of 37 mm, however the TOFD method showed a maximum defect size of only 5 mm. Based on the SAFT results the pressure vessel was taken out of service.
The SAFT results were verified by destructive metallographic cut-outs, see graphic. The detected defects were surface breaking cracks within the weld-material at the vessel's inner surface. (The four test zones for the TOFD results are illustrated in the graphic as 3 mm cracks, while the service company determined a crack size of less than 3 mm.) Based on these results it is uncertain if TOFD is a reliable method for detection of cracks or sharp grooves at the inner wall of vessels or piping.

Further studies will be necessary to learn more about the limitations of TOFD's application, as well as how to exploit any possible advantages (BASF Press Release).

Jan Verkooijen was confronted with the findings of Dr. Hecht for the first time during the conference, and wants to receive full insight in the results. He also showed results of the well known Dutch Thin Plate project by the Dutch Welding Institute (NIL) and KINT (the Dutch society for Quality surveillance, Inspection and Non Destructive Testing). During the Project, which costed nearly 0.5 M$, some 250 defects where examined with different NDT techniques and the results verified afterwards by destructive examination. From the project it was shown quite clearly, that only taking into account the defect detection Pulse Echo Meanderscanning was best, which was never denied. In the discussion about the norm however, it should not be a confrontation between mechanized Meanderscanning in Pulse Echo mode and TOFD, but it is more important to see the differences compared to Radiography and Manual Ultrasonic testing, because that is what the Norm is about: Pre-Service inspection. He repeated his invitation to the experts to work together internationally in the already mentioned new project.

Mr. Hecht restates a big drawback of TOFD - the uncertainty of defect detection in the weld rood. Prof. Langenberg used a model to show that a small crack (very dangerous for a part's fitness) cannot actually generate a diffraction. He is very worried that in about 3 Years the HP 5/3 standard will be replaced by the TC 54 and it could be possible that the inspector will be free to decide whether to use the ultrasonic TOFD technique or the mechanized ultrasonic pulse echo method.

Mr. Kreier commented : If TOFD were applied with a necessarily high sensibility, then the Falls Call Rate would be so high that practically every second weld would have to be rejected. At another point he warned against applying techniques fro economic reasons alone. This has already happened in the US in the testing of nuclear power plants. If we all start doing that now we will have problems in the next decade and beyond.

listen now or download wuesten.ra 230kb Download the Player

Mr. Wüstenberg: "Thanks for that statement. I think that is a very important issue. I tried to raise awareness of that issue at the beginning. Please Mr. Verkooijen and other colleagues, try to understand, if we seem frightened, it's not because of ignorance, but because our experience has shown us that this method is not able to do reliable detection for these, contrary to the propaganda. But if we want to ensure reliable (certain) probability of detection then we must qualify a testing method that will work under 'Worst Case' conditions. We have that, for instance with the HP 5/3 in class C by the stated parameter settings and it would be correct to make sure that is stated very clearly for the TOFD 'Worst Case' conditions, as Mr. Hecht suggested. Then I consider that TOFD of course, a useful addition for many inspection tasks, but as it stands now, TOFD's future is very limited, if applications and proposed solutions are promoted that are contrary to the laws of physics. It is a risk, exactly as was explained by Mr. Kreier. For that reason it's also interesting to us. Mr. Verkooijen, perhaps in working groups, perhaps in conjunction with the working group for the EN draft standard, we can combine forces to to work out the fundamentals and develop a consensus within the European institutions. Because we should avoid the usual pitfalls; that is a dangerous situation for us."

Mr. Wüstenberg thanks Mr. Verkooijen, that even in the critical atmosphere here he has explained some things about the TOFD method. "I consider, that the discussion will proceed to investigate the position of such technical improvements and new methods, but adhering to the concepts of safety and responsible strategy. But that has yet to happen and discussions like this one introduced by Mr. Hecht is step in the right direction.

Contacts :
AEA Sonomatic B.V. Oosterhout NL, Jan Verkooijen Jan.Verkooijen@aea-technology.nl
BAM Berlin D, Prof Dr. Ing. Hermann Wuestenberg Hermann.Wuestenberg@bam.de
BAM Berlin D, Dr. Ing. Gerhard Brekow Gerhard.Brekow@bam.de
BASF AG Ludwigshafen D, Dr. Andreas Hecht andreas.hecht@ze.basf-ag.de
Innotest AG Eschlikon CH, Dr. Peter Kreider innotest@compuserve.com

For more information concerning TOFD see: 09/97 UTonline Journal: Focus on TOFD