[arun7328]

Dear Sir,
I am trying to model the dynamic response of a flexible barge, in head waves, for a number of wavelengths. I have successfully predicted the heave and pitch motion of the barge modelled as rigid body. I used overset mesh for this purpose. All the results are compared with results published in various references which have validated the motion and responses of this barge. I even have the experimental results of this flexible barge.

I am now trying to set up co-simulation using Star and Abaqus and I am having a few problems.

1. During my first trials, I used symmetry condition, and modelled only one half of the barge. I have used morphing motion, defined by co-simulation, for the barge, and the overset mesh is defined as floating. The barge is modelled as a beam (half beam) in Abaqus, the particulars of which are similar to that used for the experiments. I have compared the beam properties in .dat file with the actual one and it matches. Before performing a co-simulation, I had evaluated the dry natural frequencies of the beam and compared the same with the results (both in paper and Euler beam) and they match very well for the first 7 modes. To represent the wetted surface, I have defined the half barge as a dummy surface and used coupling constraints to attach the nodes of the surface to the beam. I defined a boundary condition at the end nodes of the beam element (X,Y translation =0 , X,Z rotation=0. I even released Z rotation in another model). When I coupled the fluid and structural model using an Implicit scheme, the simulation failed to run giving an error that 'Negative cells found. Please correct your mesh'. I think the problem is because of the symmetry boundary condition. I have run the Abaqus model alone using pressure load and it works fine. Is it possible to use symmetry condition and run co-simulation? If so could you tell me what could be going wrong. I have attached my Abaqus model snap shot of half beam model. I have used 4 node surface element.

2. I have then modelled the full barge and replicated the same in Abaqus. I have attached the beam model as a picture. Since I could not define the dummy surface for the barge when it was closed at the top, I deleted the surface and assigned it as surface ( I am not able to define a closed body as a dummy surface. Is it possible?). I changed from implict to explicit. Now when I set up co-simulation It gives very weird, large deformation. In head waves I expect the barge to only have vertical rigid body modes and deformations. I get horizontal bending and after a few time steps it again gives the same error. I have assigned the same end boundary conditions as mentioned above . Is this because I have deleted the top surface of the barge in my Abaqus model? I am thinking the error is because there is no top surface in Abaqus model, the fluid model is not able to morph that surface. Is that possible? I am exporting Static pressure from star and importing displacements from Abaqus. I am using a time step of 0.003 sec for both Star and Abaus and coupling step of 0.003 sec. I even tried a coupling step of 0.012 sec but it just takes longer to give the same error.

Could you please kindly guide me on this problem? I have tried many things (time steps, coupling steps, half and full mdoels etc)

Thanks
Regards
Arun

[me3840]

I don't think a symmetry condition is a good idea for this. Doesn't that limit your twisting modes?

Where do the negative volume cells occur?

[arun7328]

Originally Posted by me3840

I don't think a symmetry condition is a good idea for this. Doesn't that limit your twisting modes?

Where do the negative volume cells occur?

Thank you for the reply. I also felt that symmetry is not good. May be it would be fine for vertical bending but not for twisting. I am now trying it on a full model. The negative cells occur in the overset mesh. After a few iterations the body twists in a weird way. Even when i run the simulation in calm water with the said BC's in structural model, I still get negative cells. I am having no clue as to why this is happening.

I have attached the picture of the iso surface. After a few time steps it distorts like in the pic and a few more it gives the error.

Regards
Arun