Computational modelling and experimental tank testing of the multi float WaveSub under regular wave forcing

Emilio Faraggiana, Craig Whitlam, John Chapman, Andrew Hillis, Jens Roesner, Martin Hann, Deborah Greaves, Y-H Yu, Kelly Ruehl, Ian Masters, Graham Foster, Gareth Stockman

Research output: Contribution to journalArticlepeer-review

21 Citations (SciVal)
47 Downloads (Pure)


A submerged wave device generates energy from the relative motion of floating bodies. In WaveSub, three floats are joined to a reactor; each connected to a spring and generator. Electricity generated damps the orbital movements of the floats. The forces are non-linear and each float interacts with the others. Tuning to the wave climate is achieved by changing the line lengths, so there is a need to understand the performance trade-offs for a large number of configurations. This requires an efficient, large displacement, multidirectional, multi-body numerical scheme. Results from a 1/25 scale wave basin experiment are described. Here, we show that a time domain linear potential flow formulation (Nemoh, WEC-Sim) can match the tank testing provided that suitably tuned drag coefficients are employed. Inviscid linear potential models can match some wave device experiments; however, additional viscous terms generally provide better accuracy. Scale experiments are also prone to mechanical friction, and we estimate friction terms to improve the correlation further. The resulting error in mean power between numerical and physical models is approximately 10%. Predicted device movement shows a good match. Overall, drag terms in time domain wave energy modelling will improve simulation accuracy in wave renewable energy device design.
Original languageEnglish
Pages (from-to)892-909
Number of pages28
JournalRenewable Energy
Early online date2 Jan 2020
Publication statusPublished - 30 Jun 2020


Dive into the research topics of 'Computational modelling and experimental tank testing of the multi float WaveSub under regular wave forcing'. Together they form a unique fingerprint.

Cite this