Surface-to-Surface Contact with friction

Genius
Genius

Hi

I have experimenting with contact and friction using the following model:

Problem.jpg

  • Two plates with size 100x100 mm, thickness 8mm
  • Upper plate is subject to horizontal load in one end and a spring reaction in the other end. At bottom the plate is subject to friction force against the bottom plate
  • Midsurfaces are made and meshed using element size 10x10mm
  • Coefficient of friction: 0.1

The external loads applied are shown in the table below, along with expected friction and spring reaction:

 

LC

Vertical load

Horizontal load

Friction capacity (f. coeff=0.1)

Horizontal force in friction

Horizontal force in spring

1 No slippage

10 kN

0.5kN

1kN

0.5kN

0

2 Slippage

10 kN

10kN

1kN

1.0 kN

9kN

 

Surface Contact - PGAP in FEM

The used files are attached. This approach approximates the reactions in the table above, and results seems OK. I was happy with this since I was experimenting with different use of the GAP element (CGAP/PGAP). For curiosity, I then tried to analyse the same using the contact definition in the SIM file:

 

Surface-to-Surface contact in SIM

See attached files.

 

Load case 1 – No slip

Horizontal contact friction force

Vertical contact force

 LC1 FX.jpg LC1 FZ.jpg

Note: Slip in outer part of contact area – very uneven distribution of shear

Nice and even distribution of contact pressure

 

Load case 2 –Slip

Horizontal contact friction force

Vertical contact force

 LC2 FX.jpg LC2 FZ.jpg

Nice and even distribution of shear

Nice and even distribution of contact pressure

 

The following nodal force report summarizes the total forces in the contact element between the two plates:

Node report contact.jpg

 

FEM contact values from the above nodal report

LC

Vertical load

Horizontal load

1 No slippage

10 kN

0.46kN

2 Slippage

10 kN

1kN

 

It is observed that total contact horizontal values is close to expected, a little bit too low for LC1 though.

The expected spring force based on this is then

LC

Expected FEM spring force

1 No slippage

0.04kN

(0.5-0.46)

2 Slippage

9kN

(=10-9)

 

 

 

Further investigation of the spring force - balancing the horizontal load gives however different results:

LC

Expected FEM spring force based on contact result

Actual FEM spring force

1 No slippage

0.04kN

0 kN

2 Slippage

9kN

2.61 kN

 

A plot of the spring force for load case 2 is shown below below:

LC2 spring force.jpg

 

Questions summarizing the above observations:

A

There is an imbalance between the applied horizontal load, contact friction force and horizontal spring. This is in particular noticeable for load case 2, giving a spring load of 2.6kN where 9kN is expected. What is my misunderstanding?

B

Why is the distribution of the friction force for load case 1 so uneven? I would expect this much more evenly distributed. This was also the result in the FEM analysis using FEM surface contact mesh

C

From Simcenter help: You should use Surface-to-Surface Contact in SIM to define contact between two surfaces. Unlike Surface-to-Surface Contact, Surface Contact Mesh generates contact (or gap) elements between the two surfaces”

In this example, maybe due to some misunderstanding by me, the GAP element approach seems to give more correct results. Why should I then not use the GAP element? Anyone with more information and opinions on this?

 

 

4 REPLIES 4

Re: Surface-to-Contact with friction

Experimenter
Experimenter

Hey, I was going through this experiment, can you suggest me about how can I be able to find out the pressure genrated between two parallel plates, If use Wireless pressure sensor. The I am never used about the terms which you have mentioned in your post, so any suggestion about how to place this sensor between 2 plates and how to get the required values If we use to test it through simulink. Any suggestion will be a great help.

Re: Surface-to-Contact with friction

Genius
Genius

Hi,

Not sure if I understand you, but the pressure between the plates is of course  the vertical load divided by the area:

p=10000N/(100*mm*100*mm)= 1N/mm^2

 

The nodal loading related to an internal element with size 10mm is then: f=p*10mm*10mm=100 N. This is observed in my plots.

 

I have no knowledge about Simulink.

 

BR

Geir 

Re: Surface-to-Contact with friction

Experimenter
Experimenter

Hey thanks for the reply, as I read the configuration parameters from end but I am receving the data of these sensors in bytes and I am trying to implement this formulae on sensor Application but there the values which I am getting is not according to what I expected till now.

Re: Surface-to-Contact with friction

Genius
Genius

I have checked with Siemens and the problem with my initial analysis is that there is sliding contact.  

The contact algorithm in SOL 101 does not takes this into account, being that SOL 401

(Multi-Step Non Linear) would be the proper way to perform this modelling.

 

It was relativelly easy to transfer my problem to SOL401 and I then got the answers I excpected. Problem solved!