Transient Fluid-structure Interaction Research Articles

Transient fluid-structure interaction in naval applications using the retarded potential method

Submerged structures subjected to underwater shock remain a difficult and costly problem to analyze. In this work, the retarded potential (RP) method, which is a boundary element technique, is employed to discretize the fluid domain. The retarded potential program RPF is coupled to the ABAQUS non-linear finite element program to form ABAQUS/RPF. The Figueiredo method is used to convert the standard RP differential-delay equations for the fluid to simply delay equations, which are inherently more stable. ABAQUS/RPF is applied to both a rigid and elastic sphere. The Figueiredo approach is seen to extend significantly the stability range of the standard RP by a factor of approximately 10–20. Integration of the constant-pressure RP fluid element is shown to be very important to achieve this level of accuracy. Efforts to improve the Figueiredo approach using higher-order finite difference operators appear to have no effect on accuracy or stability. Stability of the coupled system is achieved by simply not extrapolating the structural accelerations.

Stabilizing the retarded potential method for transient fluid–structure interaction problems

Stabilizing the retarded potential method for transient fluid–structure interaction problems

Stabilizing the retarded potential method for transient fluid-structure interaction problems

The Retarded Potential (RP) method, which is a boundary element technique and non-local in both space and time, is employed to discretize the fluid domain for the analysis of transient fluid–structure interaction problems. The retarded potential analysis program RPFS is coupled to the ABAQUS non-linear finite element code to form ABAQUS/RPFS. The standard RP is inherently unstable for time steps below a critical time step that is equal to the maximum distance in the fluid divided by the wave speed. A technique referred to as the Figueiredo method is used to convert the standard RP differential-delay equations for the fluid to simply delay equations, which are more stable. The Figueiredo approach extends the stability range of the standard RP by a factor of approximately 10–20, but this time step is still not small enough to be useful for analysis. Digital signal processing methods are used to further stablize the response of the fluid by removing the oscillating high-frequency noise in the time histories of the solution without introducing phase shifting or any significant damping. Stability of the coupled system is achieved by not extrapolating the structural accelerations. ABAQUS/RPFS is applied to both a rigid and elastic sphere subjected to a plane wave, and the results using the full time histories required are completely stable and quite accurate. With this procedure, the retarded potential method may yet prove to be a valuable analysis tool for transient fluid–structure interaction problems. © 1997 John Wiley & Sons, Ltd.

Comparison of ICECO code predictions with flexible vessel experiments

Some comparisons of ICECO code predictions with experimental data concerning transient fluid-structure interaction are given. The test results are taken from flexible vessel experiments conducted by Stanford Research Institute under the direction of Argonne National Laboratory. Two different experiments are considered: one with a rigid core barrel, and one with a flexible core barrel. Both experiments are performed in simple reactor vessels with a well-defined energy source and simple boundary conditions. Correlations of pressures and impulses are made at all available gauge stations. The permanent deformations of the core barrel and the cylindrical vessels are compared with ICECO predictions. The effects of core barrel flexibility on the wave propagation and vessel deformation are also investigated. The agreement between the analysis and experiments is found to be quite good.