Problems with applysing axial load at pipe after prescribed displacements

[CleteEng]

Hi.

I have tried several options for getting convergence for this analysis model in the last part where axial load is applied. I have applied 425 mm sideways displacement on a pipe OD 355 thickn = 19mm for the first 6 meters and wt = 38 mm on the last part.

After the displacement an axial load shall be applied with 300 kN. Is there a trick / method to improve the solving of this analysis?

For the time being the analysis stop when there is a rate change in prescribed displacements. Thanks in advance for any hints on solving this.

Kind regards Anders Austad

[S.LI]

Hi,

To apply a force following a PD, you can change the value of a load curve for the force.

Maybe you can share more information on your model, since I didn't follow your description very clearly.

Thanks.

[CleteEng]

Thank you.

I have attached a zip fil for the ach file at www.clete.no/temp

Then it is easier too see.

Most likely it is the contact that seem to be disturbing. 

[S.LI]

I looked through your model.

At 35%~40%, the maximum von Mises stress is 355 N/mm^2, which is almost the yield stress. After that, a 300KN force is applied. As you know, once a structure is yielded, its ability bearing loads will dramatically downgrade.

So in your problem, it's no way to find solutions for this 300KN on a yield structure.

This is why it never converges.

[CleteEng]

300 kN is axial load and the sideways displacement is generating a moment of approx 1100 kNm which is within elastic capacity of the pipe.

If we turn off contact and use bonded it converges.

I will solve this by applying loads instaed of prescribed displacements, as it seem to be something in the part where we change inbetween the 2 phaces that there is instability. 

[S.LI]

As you see, at time 70s, the root of part 3 is almost yield.

Contact and bonded could mean a lot of difference. For example, during a bending deformation, contact means only "compression" side bearing loads, but bonded means both "compression" and "tension" sides. So the maximum stress could be twice different.

[CleteEng]

Hi.

The prescribed displacements were introducing axial loads in addition to the 300 kN

Theese should not be there, but is still not critical.

If surface contact without friction is used the analysis model is solved OK.

Hence there are capacities in the section.

I guess that the problem occur due to "rapid" change in motion such that there are oscillation masses in the structure.

Theese has been tried to be avoided by several graphs for the load application, with no luck.

The analysis converge without friction and is stable.

I will try to investigate further how to make it solveable with friction. 

[S.LI · Accepted solution]

I didn't notice it's frictional contact.

In general, frictional contact is tougher than frictionless contact.