Abaqus - contact between two parts doesn't working. Any advice?

[P.M.L]

I am a new user with the Abaqus software and I am trying to perform a 2D simulation of a person sitting on a cushion, the cushion being resting on a rigid surface. The simulation is defined as Abaqus / explicit dynamics, and I'm defining the surfaces of both the cushion and person using the contact penalty method. However, when viewing the results, the "body" passes through the cushion without contact. Any advice that I should analyze better?

 

[Erik Panos Kostson]

It looks like the contact is not activating. It is hard to say what is wrong. If you have the input or cae file perhaps someone could find out why.

 

[P.M.L]

Hi Erik Panos,
First, thanks for the reply! I'm attaching the file as you suggested!

 

[Erik Panos Kostson]

You are welcome. Unfortunately I do not have abaqus (hopefully someone else can do that).

What I would recommend (since there is an "issue" with the contact) is to use perhaps a different constraint option (e.g., kinematic instead of penalty).

Also perhaps try out a General Contact (explicit) with all surfaces active (default option I believe).

 

[P.M.L]

Hi Erik Panos,

I made the changes you suggested and unfortunately there were no positive effects. The problem may not be the contact method. Anyway, thanks for the help!

 

[Dave442]

are you sure contact isn't established - you are generating stresses in the cushion. try to plot contact pressure etc.

how have you defined the step time and increment for the explicit step? are you trying to run a quasi-static analysis?

 

[P.M.L]

Thank you for all the help.

 

[Erik Panos Kostson]

Interesting paper. Many different units, tricky. If you are using consistent units, Ton,N,mm, K (and not SI, see Link), then density of steel is 7.8E-9 ton/mm3, 15 kpa is 15E-3 Mpa (gravity is 9.8E3). Try that and see how it goes, Otherwise I will ask a friend to have a look.

 

[Erik Panos Kostson]

With the following changes it works:
- change to consistent N, mm, Tons, MPa or SI units (as described in my previous post).
- plane strain condition to be added to the section definition (edit sections and tick plane strain box).
- change to plane strain elements (CPE4R).
- remove/delete point load on the muscle and add perhaps a pressure load instead (point load on a 2D part is never realistic, since it acts on a node that does not have any extension/dimensions) that is activated in 2nd step (after the first gravity step when the part has rested on the cushion, this pressure load can be ramped up). If you need to input a harmonic load just add an amplitude table describing e.g. a sinusoidal function instead of ramping it up). You can add also the constant pressure load on the muscle (ramp it up), and then use an acceleration/gravity load (3rd step) to excite the whole structure using a harmonic acceleration excitation via an amplitude table (e.g., 1 Hz excitation, like one of the load cases in the paper).

 

[P.L.M]

Hi Erik Panos Kostson,

Here is P.M.L, I'm having trouble accessing my account, I believe it's because I posted the article link.
However, I would like to thank you for the help, set the units and it worked.
My last question! in that link you sent me, in case I configure the parameters of the Hyperelastic material in Pa, which unit am I working on? would it be in the SI? kg, m, s...

And again, thanks for the help!

 

[Dave442]

this is useful:

https://www.dynasupport.com/howtos/general/consistent-units

 

[Erik Panos Kostson]

Hi

Good to hear that you are OK.

If you set the units to Pa, then you could work in SI units, which is good if one uses a software that does not let one choose units (in Strand7 and other FEA software, one can choose the units), and where the standard is free (some areas of engineering, one might need to use kPa, instead of Pa and so on).

In this case (if you choose SI units) though bare in mind that you need to change/model the part in meters (I assume the parts are in mm), density to kg/m3, force to N, elastic coef. to Pa, gravity to 9.8 m/s2 (instead of 9800 mm/s2) and so on.