Dynamic/Implicit. How to define a rolling condition with an initial speed and good contact? Please!

[omar.el92]

Hi everybody,

I really need your help, i really have been struggling with this issue for weeks now.

I want to define a rolling condition of a wheelset travelling at 70Km/h on a rail track.

I have defined three steps:

1)Static step: contact initiation with an imposed vertical displacement of the wheelset;

2)Static forces initiation: removal of the displacement and contemporary application of the static loads at the bearings.

3)Dynamic/implicit step: 

  T=1s;

  Increment size:0.0005s

  NLGEOM ON

  Velocity BC (rotation and longitudinal speed) on a RP coupled (distributed      coupling) with the wheelset axis. Instantaneous load.

I have tried tons of strategies, but i never got a realistic rolling with good contact: either the axis deforms, the velocity field oscillates, the contact saturates completely.

I have attached some pictures of the model. The V field is after 0.002s of integration, and as can be seen it's not following the BCs. 

I always have CSHEARF=(frictioncoefficient)*CNORMF.

You can also find attached the .inp file of the model.

Please i'm really looking forward for your help!

Regards,

Omar

 

[Mustaine3]

Have you tried to do an Import-analysis with the model from 2) to 3) and apply initial conditions for the explicit job?

 

[omar.el92]

Thank you for the answer Mustaine3!

It's actually a Dynamic/Implicit step.
You mean from the static part and the dynamic one? What would be the benefit of making an import of that part? The BC are already applied for the implicit job. Thank you for the cue.
Omar

 

[Mustaine3]

When you do it all in one analysis, but then you have to increase the velocity slowly (=not instantaneous), otherwise the inertia will create resistance and the elements will collapse.


When you do Std-to-Std import, you can try predefined field to start the analysis with the desired velocity. But I don't know, if the initial velocity can be applied to imported structures. Try it.