The first reported observations of a working ART (Anti-Roll Tank) were made by Watts in 1883 with a free surface tank placed on board the HMS Inflexible during a series of voyages (Watts, 1883). The tank was not designed for this ship, it was just fitted in the available space, and the working principles of the device were not yet understood. In 1911, Frahm designed a new type of tank, the U-type, based on a series of scientific assumptions, among which that the tank should be in resonance with the natural roll period of the ship (Frahm, 1911). Though this was based on the assumption that the response of the tank would counteract the wave induced roll excitation, rather than acting as a damper, it was the start of a better understanding of the ART design.The response of ART and tank sloshing was thoroughly investigated in the fifties and sixties, both for marine applications and for rocket fuel tanks (Stigter, 1966; van den Bosch, Vugts, 1966; Honkanen, 1971; Abramson, Silverman, 1966). It became evident that the response of such a tank is strongly non-linear, both as function of the excitation frequency and as a function of the amplitude. Linear methods were developed at the time, such as equivalent pendulums, but those showed unacceptable simplifications (Chu et al., 1968).