RE: questions and application of ZeroLengthContactASDimplex

[ryancwl95]

Hi

I am modelling a tubular pile embedded in soil, and I am keen on using the ZeroLengthContactASDimplex elements to model pile-soil interaction. I have some questions about the inputs for this element. I watched the webinar on the use of these contact elements, but I still have some uncertainties that I hope you can help me with.

ZeroLengthContactASDimplex TCL code:
element ZeroLengthContactASDimplex $eleTag $n1 $n2 $Kn $Kt $mu <-orient $nx $ny $nz> <-intType $type>

1.) What should be the range of $Kn & $Kt values? In some resources, the values for $Kn and $Kt are reported to be the same, ranging from 1e12 to 1e16. I would like to know if this is the correct range of values to use for these variables. In one of the webinar sessions, different values for $Kn and $Kt were discussed. Could you explain why these values can differ in certain cases and when they should be the same? Also, is there a sure way to calculate these values ($Kn & $Kt), it is not explicitly shown or explained in the manual.

2.)The image shown below is a representation of my model. I want to simulate the contract interaction using ZeroLengthContactASDimplex between the surface of the pile (blue) and the surrounding soil (red). Since this is a curved surface, can I confirm that I should be using "from element" for the orientation type rather than " from local axes"?

Screenshot 2025-07-10 142434.png (194.08 KiB) Viewed 3085 times

3.) What are the orientation values $nx$, $ny$, and $nz$ in the input TCL code to model the "from element" orientation type?

I look forward to your reply. Thank you in advance for your help.

Sincerely,
Ryan

[STKO Team]

1.) What should be the range of $Kn & $Kt values? In some resources, the values for $Kn and $Kt are reported to be the same, ranging from 1e12 to 1e16. I would like to know if this is the correct range of values to use for these variables. In one of the webinar sessions, different values for $Kn and $Kt were discussed. Could you explain why these values can differ in certain cases and when they should be the same? Also, is there a sure way to calculate these values ($Kn & $Kt), it is not explicitly shown or explained in the manual.

They are penalty stiffnesses: In theory they should be INF, but they can't, so they should be high enough wrt the typical stiffness of your model, but not too high to avoid numerical instabilities.
Make it 2 or 3 orders of magnitude higher that then typical stiffness of your model (for example take the typical stiffness as E*L, where L is the size of your mesh)

2.)The image shown below is a representation of my model. I want to simulate the contract interaction using ZeroLengthContactASDimplex between the surface of the pile (blue) and the surrounding soil (red). Since this is a curved surface, can I confirm that I should be using "from element" for the orientation type rather than " from local axes"?

That's only relevant if you model it from STKO. If you model it in pure Tcl, just put the components of the contact vector, i.e. the unit-vector that goes from the master node to the slave node

[ryancwl95]

[/quote]
That's only relevant if you model it from STKO. If you model it in pure Tcl, just put the components of the contact vector, i.e. the unit-vector that goes from the master node to the slave node
[/quote]

Hi, thank you for your response. Could you please explain what you mean by "contact vector values" and how it should be assigned for a curved surface, like in the example I provided? Additionally, what should the orientation values $nx, $ny and $nz be in input TCL code? Since it is a curved surface and the contact can potentially move in three directions, should the orientation be set to -orient 1 1 1?

Thank you for your help thus far, I look forward to your reply.

Regards,
Ryan

[STKO Team]

Hi, thank you for your response. Could you please explain what you mean by "contact vector values" and how it should be assigned for a curved surface, like in the example I provided? Additionally, what should the orientation values $nx, $ny and $nz be in input TCL code? Since it is a curved surface and the contact can potentially move in three directions, should the orientation be set to -orient 1 1 1?

N = [nx, ny, nz]
N is outward unit normal vector of the contact master surface

[ryancwl95]

N = [nx, ny, nz]
N is outward unit normal vector of the contact master surface

And if the interface elements are applied to a curve surface of a pile, should the values of orientation be -orient 1 1 0 ?

[STKO Team]

And if the interface elements are applied to a curve surface of a pile, should the values of orientation be -orient 1 1 0 ?

No, those three values are the x,y and z component of the outward normal vector at a given point, thus cannot be constant on a curved surface.

Hint: imagine your curved surface is a cylinder whose central axes is allined with the global Z axis, the -orient components should be: cos(alpha), sin(alpha), 0.