Filtern
Erscheinungsjahr
Dokumenttyp
- Vortrag (18)
- Beitrag zu einem Tagungsband (17)
- Zeitschriftenartikel (5)
- Beitrag zu einem Sammelband (4)
- Forschungsbericht (2)
Schlagworte
- POD (23)
- Reliability (18)
- NDT (7)
- Ultrasonic (7)
- Zuverlässigkeit (6)
- Railway (4)
- Ultrasonics (4)
- Zerstörungsfreie Prüfung (4)
- Human factors (3)
- Probability of detection (3)
Eingeladener Vortrag
- nein (18)
Currently at Deutsche Bahn (DB) ultrasonic inspections in maintenance procedures for wheelset axles with a bore hole are, to the greatest possible extent, carried out using automated ultrasonic inspection system. Although the acceptance levels are in accordance with DIN 27201 part 7, the testing results have shown in recent years that in the case of true indications, the effective defect sizes were far below the level of acceptance. Due to this experience it can be assumed that the automated ultrasonic inspection systems are testing substantially more sensitively than required. This increased sensitivity leads to an increase in false indications, generally resulting in the unnecessary demounting of wheelsets. In Research cooperation between Federal Institute for Materials Research and Testing (BAM) and DB the effective flaw detection sensitivity of existing automated ultrasonic inspection system will be determined and the true/false indication ratio optimised. Through systematic investigations with the Probability of Detection method on the existing automated ultrasonic inspection system the results could be directly applied to the optimisation of existing ultrasonic inspections of wheelset axles with a bore hole in maintenance procedures and the level of reliability can be considerably increased.
Ultrasonic phased array NDE has been applied to ensure the integrity of canisters for encapsulation of spent nuclear fuel. The performance of the NDE system is evaluated by the POD analysis. The POD analysis using a common method, â versus a, has been modified for the phased array ultrasonic inspections by two approaches: to take more influencing parameters into a, and to use a more sophisticated quantity as â. The POD with new a allows more detailed interpretation of POD for each parameter, and the new â gives more realistic POD. The methods are discussed and demonstrated with experimental data. In addition, an investigation of human factors is being planned and the plan is discussed.
The General Guidance in the Swedish regulations state that the safety assessment for a final repository for spent nuclear fuel should be 1 million years after closure. SKB developed the KBS-3 method, according to which the spent nuclear fuel is protected by three barriers. It is encapsulated in canisters with a diameter of 1 metre and a length of 5 metres. The canister consists of a cast iron insert surrounded by a 5 centimetre thick shell of copper. The canisters are disposed in the bedrock at a depth of about 500 meters surrounded by bentonite clay. In order to assess the safety over this extremely long period, an extensive quality control programme is applied to the canisters before deposit. In this programme, the use of
non-destructive testing (NDT) is vital. The safety assessment of the canister in turn places high demands on the coverage, detectability, and reliability of the applied NDT inspections of the canister parts, i.e. cast iron insert, copper base, tube and lid, and the copper friction stir welds (FSW). This paper presents the extensive full-scale inspection development programme that runs at the Canister Laboratory in Oskarshamn (Sweden). In order to fulfil the high demands, phased array ultrasonic inspection techniques are developed using practical trials aided by ultrasonic modelling. The techniques apply, for example, different frequencies, inspection angles, focus depths, and both longitudinal and shear waves. Increased inspection reliability of the FSW is achieved by applying digital X-ray technique using a 9 MeV linear accelerator and a line detector. To complete the coverage, complementary surface inspections methods, i.e. eddy current array, magnetic flux sensor techniques and magnetic particle inspection, are applied. The canister safety assessment was the driving force to include reliability studies during the NDT development. Initially, the technical reliability was considered, resulting in development of advanced POD models (probability of detection). In combination with human factors studies, these models were implemented as tools in the development of the NDT techniques. Human factors studies were also applied to improve the inspection procedures to be more user-friendly enabling reliable inspections.
The vast experience with the automated, ultrasonic system for the inspection of hollow railway axles used by Deutsche Bahn shows that much smaller flaws are detectable than required. This results in a number of false calls. False calls lead to unnecessary demounting and disassembling of wheelsets, which generates unnecessary additional costs. In order to adjust the sensitivity of the inspection system to reduce the number of false calls without compromising safety, the capability of the system to detect cracks needs to be comprehensively established. This capability can be quantified by using probability of detection (POD) curves for the system. The multi-parameter POD model makes it possible to include several factors that influence the crack detection in the analysis. The analysis presented in this paper shows that crack position, orientation, depth extension, and shape as well as the geometry of the axle all have influence on the ultrasonic response amplitude. For future work, calculation of the POD using multi-parameter POD model with these parameters is planned.
It is not always only the size of the flaw that determines the severity of the flaw for the structure. In such cases, it is important to express the capability of the non-destructive testing (NDT) system to detect a flaw with respect to exactly those parameters that determine flaw severity. The multi-parameter reliability model presented in this article shows a way of calculating and expressing the probability of detection (POD) as a function of different influencing parameters, using numerically simulated NDT system responses and experimentally measured responses. A successful application of the model is demonstrated on the data from a transmit-receive longitudinal (TRL) ultrasonic inspection of a cast iron component. The POD of the surface-breaking semi-elliptical crack-like flaw is expressed as a function of its depth and length. In a direct comparison with the conventional signal response analysis, where the POD is expressed as a function of only the flaw size, the method provides a more comprehensive estimation of the reliability of the NDT system.
This report describes the methodology of the reliability investigation performed on the ultrasonic
phased array NDT system, developed by SKB in collaboration with Posiva, for inspection of the
canisters for permanent storage of nuclear spent fuel.
The canister is composed of a cast iron insert surrounded by a copper shell. The shell is composed
of the tube and the lid/base which are welded to the tube after the fuel has been place, in the
tube. The manufacturing process of the canister parts and the welding process are described.
Possible defects, which might arise in the canister components during the manufacturing
or in the weld during the welding, are indentified.
The number of real defects in manufactured components have been limited. Therefore the reliability
of the NDT system has been determined using a number of test objects with artifical defects.
The reliability analysis is based on the signal response analysis. The conventional signal response
analysis is adopted and further developed before applied on the modern ultrasonic phased-array
NDT system. The concept of multi-parameter a, where the response of the NDT system is dependent
on more than just one parameter, is introduced. The weakness of use of the peak signal response
in the analysis is demonstrated and integration of the amplitudes in the C-scan is proposed as an
alternative. The calculation of the volume POD, when the part is inspected with more configurations,
is also presented. The reliability analysis is supported by the ultrasonic simulation based on the
point source synthesis method.
When pushed to the limits of their detection capability, NDE systems do not produce consistent hit/miss
indications. Their capability of detecting small defects is therefore expressed in terms of POD. An adequate
NDE system is required to ensure the structural integrity. In conventional signal response analysis, the POD is
expressed as a function of the defect size, and its adequacy for the inspection task is tested against the maximum
allowable defect size which will not undermine the structural integrity. Analyses of modern structures show that
other parameters, beside the defect size, can both significantly influence the POD and determine the severity of
the defect for the structure. Within the multi-parameter reliability analysis, the POD is expressed as a function of
those influencing parameters. When determining the adequacy of the NDE system, the capability of detecting a
defect has to be expressed and tested against the critical value of exactly that parameter that determines defects
severity for the structure. Failing to do so can lead to a rejection of the healthy, or acceptance of the bad part.
The principle is demonstrated on the example of the Transmit-Receive Longitudinal (TRL) ultrasonic inspection
of the iron cast component for semi-elliptical surface defects.