## Fluids

#### Ansys aqwa. Joints close loop in time response. Can we solve this in aqwa?

Tagged:

• Mohamed Hassan
Subscriber

Dear All, I am trying to model a motion compensated gripper frame in anays aqwa. The motion compensated frame objective is to compnestae relative motion between floating installation vessel and floating spar in order to install wind turbine above the flaoting spar.

For now i am trying to use a mchnical connection that enable me to have both the vessel and the spar rigidly connected in X and Y directions and allowing the other movements. in other words, both the vessel and the spar are free to move in Z, rotate (X,Y,Z) indpendnent of each other.

I have tried several connections, the closest connection, that could suits my requst is ball and socket, however, it rigildy connect the two structure at Z direction.

How can this be solved in Aqwa? is there any other way of model like two spring damper systems in X and Y directions instead? or any other advise?

• Mike Pettit
Ansys Employee

Hello,

For this system you could use a 'dummy' structure between the vessel and the spar. Connect the spar to the dummy structure using a ball and socket joint, so that the spar can rotate freely in all directions. Then connect the dummy structure to the vessel using a hinge joint, where the hinge axis is parallel to the XY plane and is positioned a long way away (e.g. 1000 m) from the vessel/spar. In this way the motion of the dummy structure relative to the vessel is almost linear in the Z direction as it travels along a small arc of a large circle around the hinge axis (assuming that there is not a very large relative Z motion between the vessel and spar).

To create the dummy structure you can add an extra part in the geometry editor (it doesn't matter what the part looks like), import the updated geometry into Aqwa Workbench, and then suppress the surfaces of the new part. You will need to include a mass definition for it, which should be subtracted from the spar mass. The mass of the dummy structure should not be very small, as this may cause numerical stability problems in the time domain calculations.

I hope this helps!

Mike