Filtern
Erscheinungsjahr
Dokumenttyp
- Vortrag (43)
- Beitrag zu einem Tagungsband (27)
- Zeitschriftenartikel (20)
- Posterpräsentation (8)
- Beitrag zu einem Sammelband (3)
- Monografie (1)
Sprache
- Englisch (102) (entfernen)
Schlagworte
- Degradation (12)
- Residual strength (10)
- Slow burst test (10)
- Regulations (9)
- Acceptance rate (8)
- Composite (8)
- Hydrogen (8)
- Sample testing (8)
- Composite cylinders (7)
- Probabilistic (7)
Organisationseinheit der BAM
- 3 Gefahrgutumschließungen; Energiespeicher (34)
- 3.0 Abteilungsleitung und andere (18)
- 3.5 Sicherheit von Gasspeichern (16)
- 5 Werkstofftechnik (3)
- 5.5 Materialmodellierung (3)
- 8 Zerstörungsfreie Prüfung (3)
- 8.6 Faseroptische Sensorik (2)
- 2 Prozess- und Anlagensicherheit (1)
- 2.1 Sicherheit von Energieträgern (1)
- 3.3 Sicherheit von Transportbehältern (1)
Eingeladener Vortrag
- nein (43)
Implementation and current situation of introducing the Directives 97/23/EC in the German legislaton
(2001)
The objective of this work is to find a method that describes the degree of damage from an impact experiment. This experiment was performed on Composite Pressure Vessels (CPV) in order to find the correlation of impact damage to the residual burst pressure. Computed Tomography (CT) approach was used to capture the before and after impact condition of the CPVs. The Wasserstein function was used to calculate how much the after impact image has differed from the original one. In the end, a good correlation was obtained to the residual burst pressure.
The test data for static burst strength and load cycle fatigue strength of pressure vessels can often be well described by Gaussian normal or Weibull distribution functions. There are various approaches which can be used to determine the parameters of the Weibull distribution function; however, the performance of these methods is uncertain. In this study, six methods are evaluated by using the criterion of OSL (observed significance level) from Anderson-Darling (AD) goodness of Fit (GoF), These are: a) the norm-log based method, b) least squares regression, c) weighted least squares regression, d) a linear approach based on good linear unbiased estimators, e) maximum likelihood estimation and f) method of moments estimation. In addition, various approaches of ranking function are considered. The results show that there are no outperforming methods which can be identified clearly, primarily due to the limitation of the small sample size of the test data used for Weibull analysis. This randomness resulting from the sampling is further investigated by using Monte Carlo simulations, concluding that the sample size of the experimental data is more crucial than the exact method used to derive Weibull parameters. Finally, a recommendation is made to consider the uncertainties of the limitations due to the small size for pressure vessel testing and also for general material testing.
The presentation reflects the questions that have been raised by member states and NGOs.
As background information details on the question how to choose a suitable salvage pressure receptacle (SPR) are provided. This leads to the need for deleting the volume limitation for SPRs and to the discussions in 2014 and 2016 on this issue.
As an adequate substitution for deleting the volume limit for SPRs it is proposed to limit the pressure receptacles to be stored in by a maximum pressure-volume-value of 1.5 Mio bar-litres.
After some explanations to standardisation projects currently also tackling this pV-limit, some risk-based safety criteria for a limitation are mentioned with the focus of the consequence level and the general avoidance of catastrophic worst case scenarios.
First the acceptance rates resulting from the analysis of different minimum burst ratios and standards are displayed and explained. Especially the problems resulting from a further reduction if the minimum burst ratio are shown.
For a better understanding of these results the process of analysing the acceptance rate is explained in detail. This contains the alleged contradiction between safety of samples due to its scatter while the rejected sample provides a higher mean strength.
Then the importance of batch testing is explained and leads to the recommendation to have a strong interaction between provided mean strength and requested maximum of production scatter.
Based on the preliminary presentations in this task force it is repeated how a basic population scatters and its degradation influences properties statistically. There is a remainder on the issue of accepting borderline populations due to the minimum burst ratio requirement.
Then it is explained in detail how the analysis of minimum requirements by a Monte-Carlo-simulation works. The result with respect to the further reduced minimum burst ratio of 2.0 is summarised and means a two-step approach of two scatter levels requiring two different minimum burst ratios.
The methods for evaluating batch tests are analysed. A proposal for integration a statistical analysis is described in detail.
Finally, the positive effect of proof testing is shown and its dependence from the real ageing is shown.
This ends in the recommendation to have a stronger and quantitative look on real ageing and to follow the 2 proposed sets of criteria for a safe usage of hydrogen storage systems.
Faserverbundwerkstoffe (Komposite) zeigen außergewöhnliche Veränderungen ihrer Eigenschaften in Abhängigkeit von der Betriebsdauer. Daraus leitet sich auch die Notwendigkeit ab, eine Methode zu finden, die in der Lage ist, Festigkeiten und Restfestigkeiten von Komposit-Druckbehältern zu ermitteln. Dies kann mittels Lastwechselprüfung oder in anderen Fällen vorzugsweise mittels der langsamen Berstprüfung erfolgen. Für eine aussagekräftige Beschreibung der Festigkeit bedürfen die Ergebnisse beider Prüfverfahren einer statistischen Auswertung. Aber insbesondere die statistische Auswertung der berstbezogenen Festigkeitseigenschaften unterliegt neben der Frage der Stichprobengröße einer großen Unsicherheit mit enormen Folgen. Dies ist die Frage, welche Verteilungsfunktion der Streuung der Festigkeitseigenschaft konservativ zugrundegelegt werden darf. So ist zu untersuchen, ob die Gaußsche Normalverteilung (NV) die wahre Verteilung richtig beschreibt. Es könnten aber auch andere Funktionen, wie z. B. die Weibull-Verteilung notwendig sein, um eine Überschätzung der Überlebenswahrscheinlichkeit bzw. Zuverlässigkeit in dem interessanten Bereich von einem Ausfall pro 104 bis 108 Druckbehältern zu vermeiden.
Slow burst testing of samples as a method for quantification of composite cylinder degradation
(2014)
The current practise to focus periodic retesting of composite cylinders primarily on the hydraulic pressure test has to be evaluated as critical. The test itself always causes a certain amount of micro damage to the cylinders but does not necessarily deliver sufficient and evaluable information. Thus BAM Federal Institute of Materials Research and Testing (Germany) moves the focal point to a new approach for validation of composite cylinders, based on destructive sample tests parallel to operation. Statistical assessment of results of these destructive tests is employed for the estimation of remaining safe service life, based on reliability demands. The estimated service life is also used for the determination of re-test periods of the examined population of composite cylinders.
An essential aspect of this approach is the validation of current residual strength and its prediction at any point of service life. In cases of gas cylinders with very high cycle strength, residual strength cannot be quantified statistically by means of hydraulic load cycles. As a replacement, creep tests or burst tests may be employed. BAM suggests the 'slow burst test SBT' as a combination of these two test procedures. This is a compromise between the practicability of the (conventional) burst test and the practical relevance of sustained loads during service, to be tested in creep rupture tests. In this paper, a variety of 99 burst results of a cylinder design type used for breathing apparatus (CFRP with PE-liner) is evaluated. The influence of test procedure parameters and nature and intensity of artificial ageing on the test sample strength are analysed statistically. This leads to an evaluation of different procedures of artificial ageing and the recommendation to substitute conventional burst tests by slow burst tests for the assessment of composite pressure receptacles.
There is evidence that composite materials degrade under static loads. Consequently there is no way to design composite gas cylinders for unlimited service life. Yet this fact is not reflected by design standards for composite receptacles approved for the transport of dangerous goods or for stationary application. Unnecessary cost and weight result from this.
Due to the degradation aspects of composites it is necessary to provide appropriate retesting procedures, which allow rejection of used vessels as is common practice for metallic receptacles. The current procedure of hydraulic tests appears to be an expensive procedure with serious doubts regarding its usefulness for composite cylinders.
Consequently, an efficient method is needed, which allows limiting the life time to its safe maximum. It is thus recommended to combine retesting with the assessment of residual life time. Since relevant non-destructive methods are not yet sufficiently matured and standardized enough, a concept featuring destructive tests on composite cylinders parallel to operation should be considered. An alternative concept introduced here, proposes to substitute the hydraulic testing of 100% of a population by a probabilistic evaluation of destructive tests on samples parallel to operation. Intensified pre-fill inspections should be an additional component of such a concept.
The reduced volume method has found a particular zone to be used for evaluating the strength of unidirectional composite materials. To validate the method, modelling an experimental test where it has the same basic assumption of the model had been carried out. The result was favourable as the model can predict the average value from the experiment and also the time dependent effect.
This presentation explains how an PA-based approval concept could work and would substitute the most of deterministic requirements based on a risk based quantification of accepted failure rates.
The concept of a probabilistic approval approach is based on 5 separate types of requirements:
1. Testing and evaluation of properties at Beginning of Life (BoL)
2. Artificial aging with residual strength assessment
3. Testing against dedicated accidental loads
4. Surveillance of production quality
5. Surveillance of degradation/ check of estimated end of life (EoL)
Each of these aspects is visualised by photos and diagrams.
This presentation provides an impression of the boundary condition of the work on composite cylinders on a competent authority. It shows the various kinds of usage, the different tasks and duties around this kind of storage units and the wide spread of technical competence necessary for the substantial research when addressing the legislation. Based on this it shows how we react on changes in the regulations by research and how this influences the regulations, again. As an example, our probabilistic approach is mentioned and drafted for getting mentioned several times during the ongoing set of presentations.
Referring to the inter-departmental project group "Pressure Equipment" and the section "Pressure Equipment-pressure receptacles; Fuel Gas Storage Systems" developed from, it is shown here how research on safety aspects, technical regulations and the activities of the Federal Institute BAM in its role as competent authority interact to the advantage of safety of composite pressure vessels.
Conventional approval requirements exclusively ask for minimum strength values, which have to be met. The probabilistic approach estimates how likely none of the comparatively manufactured units fails during operation.
Both questions are juxtaposed and compared here with respect to the load cycle tests. The influence of the sample sizes is discussed additionally.