Piecewise system with discrete elements

ZacLee

Hello Code_Aster users,

I am trying to build a discrete model with piecewise nonlinearities (reference in the image bellow). The idea is to model a mass that is free to move a delta distance before coming into contact with a linear spring. I’m encountering some issues with my current approach using DIS_CHOC elements. Does anyone have any suggestions that could help me model this?

Best regards,

![
](https://)

jeanpierreaubry

ZacLee

i cannot see why this would not work
i have used this quite a few times with CABLE and MEMBRANE structures without any trouble
sharing you files would help

@"miib"
DIS_CHOC has to appear in

STAT_NON_LINE(
	COMPORTEMENT=(
		_F(RELATION='DIS_CHOC', DEFORMATION='PETIT_REAC', GROUP_MA=('c3',),),
	),

preceeded by a material definition like this

atoil=DEFI_MATERIAU(
	ELAS=_F(E=100.0, NU=0.3,), #not too still
	DIS_CONTACT= _F(
		RIGI_NOR=1e3,
		DIST_1=2,DIST_2=2,
	),
);

ZacLee

jeanpierreaubry

runcase-1-stage-1.txt

5kB

Thank you for your response. What I really want to simulate is a piecewise absorber—a mass free to move between two springs. I’ve attached my file for reference. Should I also apply a DIS_CONT material to the mass? I’m also struggling with determining the force profile. It should be a piecewise force, but when I output the reaction at a node, I’m getting a linear function. Thank you for your help.

P.S. I’ve attached a picture of the model I created in Code_Aster for this simulation.

jeanpierreaubry

ZacLee
thanks
but without the mesh impossible to give any advice

ZacLee

Dear jeanpierreaubry, I appologize for my mistake. Please, find attached my Mesh in a zip file

ZacLee

jeanpierreaubry The file is not supported in the chat conversation

ZacLee

jeanpierreaubry

global-mesh.txt

143B

Please, find the mesh file

jeanpierreaubry

ZacLee
what format is this mesh?
we do not know where the groups are allocated
this is not very useful, not to say useless
please be serious!

ZacLee

Dear jeanpierreaubry I've exported in a .DAT format because of the size of the file. I'm having some strugglings to upload a .MED file due to his size. Apologize again for my ignorance. I'm just learning to use the software, That's why I'm doing many mistakes.
My zip files with .MED file are being shaddowed on the chat
Cordially

jeanpierreaubry

ZacLee
use this, it will print the mesh in the mess file and share it here

mesh=LIRE_MAILLAGE( 
    INFO=1,
    INFO_MED=2,
    UNITE=20,FORMAT='MED',
);

otherwise i notice that the mesh seems to extend on 9 meter with a mass of 136.9 g, that seems very curious to me
and it seems to me that
there should be an ELAS value in DEFI_MATERIAU
in COMPORTEMENT
RELATION should be DIS_CHOC and not DIS_CONTACT
and there should be DEFORMATION='PETIT' specified
did you read the .mess file there probably are some warnings in it

ZacLee

Dear, jeanpierreaubry

Please find attached my message file with the mesh printed following your recommendation. I'm probably getting the wrong physics because I am measuring the wrong physical quantities. Normally, the reaction at the node should follow a piecewise function, as it only starts moving after contact. If you can help me solve this or suggest another way of modeling it, I would be grateful.

message.txt

257kB

Best regards

jeanpierreaubry

ZacLee
i will hve a look
what the units and dimensions?

ZacLee

Dear,jeanpierreaubry

All units are in SI: MKS. The two springs have 4m lenghts and the the damper on wich the mass is attached is 4.5m lenght. Please, npte also that the gap betwen the two springs is 1m.

Thank you.

Best regards

jeanpierreaubry

ZacLee
then the values of stiffness and mass are way out of the physics in my opinion
neverheless
here is how to modify the mesh

i have added 2 line elements named 'choc' between MASS and the ends of the springs
the attached file is the gmsh .geo file
but it is easy to find

ex1.txt

585B

out what is what

jeanpierreaubry

jeanpierreaubry
and here is the command file that goes with it
i think i have results that look realistic with the strange stiffness and mass
i modified some of them

ex1-comm.txt

4kB

ZacLee

Dear jeanpierreaubry,

Thank you verymuch for your help. The system is a piecewise Non linear energy Sink (NES). Normally, the stifness is low compared to a normal absorver as we want to transfer the energy from a primary system to the NES. I'll try to execute the files you sent to me.

Best regards

ZacLee

Dear, @jeanpierreaubry
I apologize for bothering you again. After recreating the model as you suggested (I followed your instructions exactly), I received an error message regarding a node that I did not create. The error indicates that there is a displacement in the DX direction for a node that does not exist in the node set.

Could you help clarify this issue for me?

Best regards,

message.txt

48kB

jeanpierreaubry

ZacLee
look at the message file under LIRE_MAILLAGE

the mesh has been badly build
you have many nodes that should not exist
there are intermediate nodes between the ends of the springs, for example N9 to N22
something is wrong within the hypothesis at mesh time
as far as i am concerned i never use salome so i cannot help
but your previous mesh did not had that fault

in France where i live it is now bed time

ZacLee

Dear, jeanpierreaubry
I've got the model to work. i'm having this results

The mass displacement is purely linear as if we hadn’t the gap before the contact. This is the problem I was having in the beginning. If you still have the results you’ve got from this model, I would appreciate if you can share.
Maybe it’s just a lack of understanding of the physical behavior from my side. Nevertheless, thank you for your help. Your advices were very helpful.
Best regards

jeanpierreaubry

with this in BC do not you have too many nodes fixed?

_F(DZ=0.0,GROUP_NO=('FIXE_N', 'MASS', 'U_SPRING_CN', 'U_SPRING_FN', 'B_SPRING_CN', 'B_SPRING_FN'))),

ZacLee

Dear,jeanpierreaubry

If the contact nodes CN are free to move move in Z direction, I’m getting a rigid body mode. So the calculations is being ended with an error message. That’s why I modelled only the mass as a free to move in Z direction.