[arun7328]

I am having a big problem with the co-simulation tool. I am trying to predict motion and response of a flexible barge and S-175 containership using the coupling. I am comparing with experiments and 2-D hydroelastic predictions. The motions compare well when compared to experiments and 2-D predictions. But when I compare the bending moments they are hugely under-predicted using the coupling. There is about 50% under-prediction in the bending moment which is not at all acceptable. I am only working on head seas so interest is symmetric responses alone.
I have tried various things and also checks to see where the problem is arising from. I have varied various co-simulation parameters( exchanges, time-step, inner iterations) but the results do not change. I am pretty confident with my CFD model. The refinement on free surface, around the body etc. The waves are really stable for more than 10 wave periods and minimal decrease in wave amplitude is seen. In my FE model, I have used beam elements to represent the flexible backbone used in experiments. The beam elements are linked to dummy surface elements which represent the actual wetted surface of the ship. The beam nodes are free to translate in X and Z direction and rotate in Y (longitudinal axis is X, heave is Z and athwart ships is Y). My dry hull natural frequencies match well with experiments and 2-D predictions. I export pressure from Star and import the nodal displacements from Abaqus. It looks like my coupling / time-step is converging well. I then calculate the Bending moments using values from about 8-10 wave periods.

I would be grateful if you could share your experience in using CFD/FEA coupling. Could you please guide me on what could be going wrong in my case? Have you encountered a similar problem when carrying out analysis on ships?

I have been working on the 2 cases for over 3 months now and did not get anywhere. Your expert comments would be immensely helpful.

Thank You again

Regards
Arun