Layer | Fill | Outline |
---|
Map layers
Theme | Visible | Selectable | Appearance | Zoom Range (now: 0) |
---|
Fill | Stroke |
---|---|
Collaborating Authors
Results
Influence of Sampling Rates On Sloshing Pressures Prediction
Fillon, B. (Research Department - Marine Division - Bureau Veritas) | Henry, J. (Research Department - Marine Division - Bureau Veritas) | Baudin, E. (Research Department - Marine Division - Bureau Veritas) | Diebold, L. (Research Department - Marine Division - Bureau Veritas) | Parmentier, G. (Research Department - Marine Division - Bureau Veritas) | Derbanne, Q. (Research Department - Marine Division - Bureau Veritas)
ABSTRACT: The aim of this study is to evaluate the influence of different sampling rates on sloshing peak pressures prediction. This study is performed for a long-duration sloshing test: 480 hours at full scale, generated by 96 five hours individual tests, at a low partial filling for one critical sea state regarding sloshing. Moreover, this raw signal at a 20 kHz sampling rate is degraded at two other sampling rates of 10 kHz and 5 kHz. Then, the influence of these 3 sampling rates (5, 10 and 20 kHz) on sloshing peak pressure prediction is investigated through comparison with the reference sloshing pressure peak prediction coming from the peak memory cards which is equivalent to 100 kHz. This comparison relies on statistical tools such as different fitting distributions, goodness of fittings, estimated pressures and confidence intervals. INTRODUCTION Sloshing model tests, submitted by the designer, are standard part of Bureau Veritas comprehensive sloshing assessment. These sloshing model tests determine the sloshing loads, here pressures to be applied on the Cargo Containment System (CCS) (Bureau Veritas, NI564, 2011). Both observations of the few sloshing events which occurred at sea and sloshing model tests clearly indicate variability of sloshing pressures (Gervaise, De Seze & Maillard, 2009). This stochastic behavior of sloshing pressures result in a flat tail exceeding probability curve. As a consequence, a small change in the probability level can have strong influence on the expected pressure. During last ISOPE Conference (Fillon & al., 2011), it was shown, for a long duration sloshing test of 480 hours at full scale, at low partial filling which is the most critical in terms of sloshing loads, that the Generalized Pareto (Gpa) distribution is more suitable than the 3- parameter Weibull and the Generalized Extreme Value (Gev) distributions for fittings of pressure peaks.
- Energy > Oil & Gas (1.00)
- Transportation > Freight & Logistics Services > Shipping (0.34)
ABSTRACT: This paper aims to show an alternate solution focused on the impact maximum record and opens to a way of carrying on comparative studies of sloshing model tests through impact pressure statistical assessment. After describing the measurement system, aiming to reach high sampling rate performances through dedicated crest memory circuit, some case studies results will be gathered to enable a pragmatic assessment of such system. INTRODUCTION From the recent years, the problematic of methane transportation has raised numerous studies towards the knowledge and assessment of sloshing phenomenon. The complexity of this field is demanding high level engineering in several domains such as mechanics, hydrodynamics and fluid-structure interactions. Methodologies have been issued to define the framework and the different steps. One of the key points in these studies is the significant role of the sloshing model tests. Although recent works have focused on the global forces (Moirod et al., 2010), up to now the associated measurements are, in most of the cases impact pressure measurements, relying on a statistical approach. Indeed as regards to the numerous configurations of sea states and filling levels of the tank that cause sloshing, the statistical approach has always been used, since the early studies in the 80's. On a structural point of view, the energy transmitted to the structure basically involving pressure versus time on a defined area, is a key parameter. The extraction of pressures associated to return periods is the main issue of a model test post-processing. It leads to qualify the severity of a test and consequently the couple sea-state/ship configuration. This statistical process is first relying on the maximum pressure extracted from a given impact of the recorded signal on a given area and time window. Consequently, the accuracy of this value might be thought of major importance.