TY - JOUR

T1 - A study on the energy transfer of a square prism under fluid-elastic galloping

AU - Jayatunga, Hewawasam Gamage Kasun Gayantha

AU - Tan, Boon Thong

AU - Leontini, Justin Scott

PY - 2015

Y1 - 2015

N2 - In this paper, power transfer of an elastically mounted body under the influence of fluid-elastic galloping is analysed.The quasi-steady state model equations are first analysed to find suitable governing parameters. It is shown that, as well as Re, the system is a function of three dimensionless groups: a combined mass-stiffness parameter, ?1; a combined mass-damping parameter, ?2; and mass ratio, m*.Data obtained by numerically integrating the quasi-steady state equations show that for high values of ?1, the power extracted from the flow is a function of ?2 only. For low values of ?1, the power extracted is still a strong function of ?2, but is also a weak function of ?1. For all the cases tested, the power extracted was independent of the value of m*.These results are then compared to results of direct numerical simulations. It is found that ?1 has a much stronger impact on the power extracted than predicted by the quasi-steady state model. The error is shown to be an inverse function of ?1. The failure of the quasi-steady state model at low ?1 is hypothesised to be due to the stronger influence of vortex shedding, which is not accounted for in the quasi-steady model. Spectral analysis of the DNS cases at low ?1 shows a significant response at the vortex shedding frequency. The strength of the vortex shedding response is also shown to be an inverse function of ?1.Even though the quasi-steady state model does not accurately predict the power extracted, it does predict the parameter values at which maximum power transfer occurs reasonably well, and both the quasi-steady model and the direct numerical simulations show that this value is basically independent of ?1.

AB - In this paper, power transfer of an elastically mounted body under the influence of fluid-elastic galloping is analysed.The quasi-steady state model equations are first analysed to find suitable governing parameters. It is shown that, as well as Re, the system is a function of three dimensionless groups: a combined mass-stiffness parameter, ?1; a combined mass-damping parameter, ?2; and mass ratio, m*.Data obtained by numerically integrating the quasi-steady state equations show that for high values of ?1, the power extracted from the flow is a function of ?2 only. For low values of ?1, the power extracted is still a strong function of ?2, but is also a weak function of ?1. For all the cases tested, the power extracted was independent of the value of m*.These results are then compared to results of direct numerical simulations. It is found that ?1 has a much stronger impact on the power extracted than predicted by the quasi-steady state model. The error is shown to be an inverse function of ?1. The failure of the quasi-steady state model at low ?1 is hypothesised to be due to the stronger influence of vortex shedding, which is not accounted for in the quasi-steady model. Spectral analysis of the DNS cases at low ?1 shows a significant response at the vortex shedding frequency. The strength of the vortex shedding response is also shown to be an inverse function of ?1.Even though the quasi-steady state model does not accurately predict the power extracted, it does predict the parameter values at which maximum power transfer occurs reasonably well, and both the quasi-steady model and the direct numerical simulations show that this value is basically independent of ?1.

U2 - 10.1016/j.jfluidstructs.2015.03.012

DO - 10.1016/j.jfluidstructs.2015.03.012

M3 - Article

VL - 55

SP - 384

EP - 397

JO - Journal of Fluids and Structures

JF - Journal of Fluids and Structures

SN - 0889-9746

IS - May 2015

ER -