HELP!! How to connect Truss element to 3d body

[MarceloM]

Hello,

I'm trying to connect a heel to a truss element to simulate the plantar fascia, however I do not know how to perform this operation. What kind of constraint do i use? See the picture.

Thanks for the help!

 

[rb1957]

attach node to node, no?

another day in paradise, or is paradise one day closer ?

 

[MarceloM]

No. I have lines(Truss elements) and i am trying to connect them to the faces of the heel, I've already tried with tie but the lines do not move in conjunction with the heel.
My goal is to have this kind of stress (see picture).
Thank you!

 

[MarceloM]

This.

 

[goeasyon]

You can use the Constraint Equation to bond the node at the end of the line to the target surface element.

https://welsim.com

 

[MarceloM]

when i open the Constraint Equation i see:

| Coefficient | Set Name | DOF | CSYS ID |
What do i chose ?

 

[Erik Panos Kostson]

OK, what software are you using?

I guess it is ABAQUS, just use a coupling constraint. Pick the vertex on the truss and then one edge or a vertex on the bone.

As you see though in your image, there are very high stresses at these connections since you have a lot of force from the truss going into a small area (or zero if it goes into a single node). So stresses are not that reliable there. If you care about the local stresses there you need to model these areas in detail.

PS: If you are using ABAQUS there is a dedicated forum for that.

 

[MarceloM]

Yes i am using abaqus. Coupling constraint don't work i alredy try it and don't cause stress in the heel. The only constraint it works it's MPC Pin but I only have tension around the face and not in it (see pic).

 

[Erik Panos Kostson]

MPC could be OK too.

 

[rb1957]

what are the white lines representing ? Are they meant to guide the motion of the foot (without taking load).

why didn't Coupling "work" ? what behaviour are you expecting to see in the model ??

another day in paradise, or is paradise one day closer ?

 

[MarceloM]

The white lines represent plantar fascia. They are elastic and isotropic with YM of 350MPa and v = 0.3. Have as mesh elements T3D2. However they are not connected to the heel or fingers and I do not know how best to represent this design. The idea of the lines is to represent the tension in the fingers and heel when the foot is over the force of the body.

Thank you for the help.

 

[MarceloM]

When i use coupling i have this error ErrNodeMissingDofConstrDef(see pic).

 

[MarceloM]

I was able to do the Coupling with continuum distributing but I have a new problem "SOLVER PROBLEM. NUMERICAL SINGULARITY" in the nodes along the lines. See picture.

" ***WARNING: SOLVER PROBLEM. NUMERICAL SINGULARITY WHEN PROCESSING NODE
OKLINHAS-5-1.9 D.O.F. 2 RATIO = 49.8681E+12 .


***WARNING: SOLVER PROBLEM. NUMERICAL SINGULARITY WHEN PROCESSING NODE
OKLINHAS-5-1.9 D.O.F. 3 RATIO = 62.1061E+12 .


......"

 

[MarceloM]

Picture.

Thanks for any help!

 

[Erik Panos Kostson]

A chain link of many trusses can be singular meaning there is not enough stiffness. Try to use beams instead to see if it helps, and if it does convert them back to trusses, but do not subdivide them when meshing, that is have one single truss element only between the connection points on the bone.

 

[MarceloM]

I'm sorry but i did not understand what you mean by "but do not subdivide them when meshing, that is have one single truss element only between the connection points on the bone".
I have to do the simulation with truss elements(tension only).
I am conecting the nodes of the lines(truss elements) with the faces of the bones using Coupling (continuum distributing) and i'm still having the "SOLVER PROBLEM. NUMERICAL SINGULARITY".

Thank you for the help.

 

[MarceloM]

You can see the Coupling in the picture.

 

[Erik Panos Kostson]

As we said a series of connected truss elements is a chain link and can be singular meaning there is not enough stiffness and is thus instable (solver cannot solve this).

One needs to provide stiffness in the d.o.f indicated by your solver output.

One way that is easier as I said is to use one single truss element, so then it is not a chain link and not singular. One can have a no compression material property (there is a tick box for this in CAE, under properties, and mechanical -> elasticity -> elastic), so it only takes tension and no compression.

That is for me (last post on this from me ).

---------------------------
Are you new here:

https://www.eng-tips.com/market.cfm

Student, post here:

https://projectboard.engineering.co...=212727627.37.1537176080710&__hsfp=3148990692

 

[MarceloM]

I merge the trusses into a sigle part and i got the same warning.

"***WARNING: SOLVER PROBLEM. NUMERICAL SINGULARITY WHEN PROCESSING NODE
PART-1-1.51 D.O.F. 2 RATIO = 2.67279E+12.


***WARNING: SOLVER PROBLEM. NUMERICAL SINGULARITY WHEN PROCESSING NODE
PART-1-1.51 D.O.F. 3 RATIO = 2.61282E+12.


***WARNING: SOLVER PROBLEM. NUMERICAL SINGULARITY WHEN PROCESSING NODE
PART-1-1.37 D.O.F. 2 RATIO = 3.05130E+12.


***WARNING: SOLVER PROBLEM. NUMERICAL SINGULARITY WHEN PROCESSING NODE
PART-1-1.37 D.O.F. 3 RATIO = 1.85751E+12.


***WARNING: SOLVER PROBLEM. NUMERICAL SINGULARITY WHEN PROCESSING NODE
PART-1-1.36 D.O.F. 2 RATIO = 38.0830E+12 .


***WARNING: SOLVER PROBLEM. NUMERICAL SINGULARITY WHEN PROCESSING NODE
PART-1-1.36 D.O.F. 3 RATIO = 195.352E+09 .


***WARNING: SOLVER PROBLEM. NUMERICAL SINGULARITY WHEN PROCESSING NODE
PART-1-1.19 D.O.F. 2 RATIO = 15.0045E+12 .


***WARNING: SOLVER PROBLEM. NUMERICAL SINGULARITY WHEN PROCESSING NODE
PART-1-1.19 D.O.F. 3 RATIO = 2.17149E+12.


***WARNING: SOLVER PROBLEM. NUMERICAL SINGULARITY WHEN PROCESSING NODE
PART-1-1.11 D.O.F. 2 RATIO = 26.0087E+09 .


***WARNING: SOLVER PROBLEM. NUMERICAL SINGULARITY WHEN PROCESSING NODE
PART-1-1.11 D.O.F. 3 RATIO = 2.34719E+12.


***WARNING: SOLVER PROBLEM. NUMERICAL SINGULARITY WHEN PROCESSING NODE
PART-1-1.17 D.O.F. 2 RATIO = 26.0072E+09 .


***WARNING: SOLVER PROBLEM. NUMERICAL SINGULARITY WHEN PROCESSING NODE
PART-1-1.17 D.O.F. 3 RATIO = 2.34375E+12.

..."

Thank you for the help.

 

[Erik Panos Kostson]

Not sure why, difficult to say.
Can you send a picture like last time showing where these nodes are (they look like truss elements, because they indicate dof 2 and 3 which has the problem).