Rubber Marine Fender

[jhardy1]

I am trying to model a hollow cylindrical marine berthing fender, similar to an over-sized cotton reel. The fenders are loaded in axial compression, and can accommodate 50% axial shortening and more, by bulging out at the mid-point. The fenders absorb very large amounts of ship berthing energy by large axial compression with minimal axial load increase. This is achieved by the fender wall “buckling”, which results in the axial load remaining fairly constant from 25% axial compression up to 50% axial compression and beyond.

(The fenders are similar to the Bridgestone Super Cell, details of which can be found here:

http://www.bridgestone.co.jp/english/diversified/marinefender/t_s_cell.html

Please note: The actual fenders are NOT Bridgestone – I don’t want to be accused of slandering Bridgestone’s good name on this forum!)

Due to various issues of proprietary information, contractual issues, etc, we have not been able to get the manufacturer’s data on actual rubber properties. I have been trying to undertake a preliminary investigation of the fender using “generic” rubber properties, but so far, I can’t get my preliminary models to run past about 25% to 30% axial compression before I get a convergence problem.

I can think of 2 possibilities, but playing with these hasn’t helped me yet:

a) Maybe my “generic” rubber properties are totally unsuitable for a rubber fender of this type?
b) I suspect there could be some “volumetric locking” happening in parts of the model subject to very high strains. I have tried refining the mesh in these critical areas, but so far, there is no indication that such refinement is improving the convergence problems. The deflected shape seems to form a local crease in the fender, where it folds back on itself. I suspect that maybe the local strains around this crease become a real singularity in my model, in which case, perhaps no amount of mesh refinement will lead to a converged solution? Or should I persist with more and more local mesh refinement until a converged solution is achieved?

Can anyone offer any particular insights; specifically:

a) Do you know of any web sites where I can find real engineering properties for engineering rubbers, suitable for use in FEA? I am looking for Mooney-Rivlin parameters or similar, Bulk Modulus, etc, for rubber with a Shore A Hardness of 77; Elongation of 350%, and a Tensile Strength of 160 kg/cm2 (16 MPa).
b) Does anyone know of a modelling strategy which can address local singularities which arise where the rubber deforms so much that it creases sharply?

Any help greatly appreciated.

 

[harry123456]

Hi
which software are you using? ABAQUS example problems manual has a lot of generic rubber properties. They have most with Mooney Rivlin and one or two with stress-strain data.
harry

 

[jhardy1]

Harry,

Thanks for the tip. I have access to the ABAQUS documentation. I will go through it and see what I can find for specific rubber properties.

I have been trying a range of what seem to me to be plausible proprieties, so far without success. I was wondering if part of my problem is that fenders of this type will only work within a limited range of properties?

 

[harry123456]

Hi,
I think that a general rubber should be good enough to make a trial run. I am also using a hyperelastic material and since I do not have material properties I am using a "generic" rubber mooney -rivlin model. As far as I am guessing the rubber material model should not make a big difference. Maybe you can look at this example in the manual

Self-Contact in automotive components: rubber gasket


Also, what are you using Abaqsus Standard or Explicit? That would also make a difference. I tried using Standard but it wouldnt work since my rubber component was compressed heavily and underwent lot of distortion. So I had to use Explicit with distortion control to make it work.
harry