The intricate physical mechanisms involved during sloshing impacts in LNG tanks lead to biases in sloshing model tests when the impact loads are predicted. In order to increase the understanding of these biases, a new state-of-the-art facility dubbed the Multiphase Wave Lab (MWL) has been established. In the MWL, impact tests are performed within an autoclave (15 m long x 2.5 m diameter), whose purpose is to provide an accurately controlled environment in which the pressure, temperature and gas composition can be controlled and monitored. Wave impact tests are performed by generating waves in a flume which is located inside the autoclave. In this paper, we present the capabilities of the MWL to control the temperature, ullage pressure and gas composition in the autoclave. We study also the quality of the global flow repeatability by means of a breaking wave which is created with a wave-focusing technique. We quantify the repeatability of the waves with a Sobolevnorm-like criterion on the frequency domain and evaluate the repeatability for different ullage pressures. Preliminary experiments show a good degree of repeatability, in accordance with high-speed recordings of the impacting waves.
Ships that transport liquefied natural gas (LNG), floating units that produce or re-gasify LNG and ships that use LNG as a fuel need to be equipped with dedicated tanks to hold the LNG and minimize the heat transfer between the LNG and the environment (LNG is stored at -162°C at atmospheric pressure). In terms of containment, membrane LNG tanks such as the NO-line and the Mark-line tanks (GTT, France) are widely used because they utilize the hull space efficiently. However, these structures are more sensitive to sloshing impacts inside the tanks. Thus, it becomes of great importance to quantitatively assess the magnitude of these impacting forces in the design stage. Currently, this is done by performing sloshing model tests at reduced scale (Gervaise et al., 2009). These tests, however, are inherently biased. Empirical scaling factors based on feedback at sea are then used to account for these biases. Dedicated measurements on-board LNG carriers have shown that the long-term statistics of the impact loads as defined by the prevailing sloshing assessment methodology are conservative (Lund-Johansen et al., 2011; Pasquier & Berthon, 2010). However, the short-term statistics derived from the tests do not consistently represent reality (Karimi et al., 2014). Resolving this inconsistency between the scaled and the full-scale measurements would lead to optimized tank designs in terms of strength and thermal capacity, as well as to increasing the operating envelope of the vessels (e.g. operating with partially-filled tanks).