contact analysis 1

[cameprak]

Hello to all,

I have to analyze Nylon plastic part closely wrapped (all 4 sides) with Expanded Polypropylene material(used for energy absorption).

static load will be applied directly on EPP part and then transferred to plastic part and again EPP. Here, should i do contact analysis. How important is this contact with or without friction. How to approach this problem to solve by FEA.

Thanks in advance for help.

cameprak

 

[xerf]

If the surfaces are in contact all the time (without separation and significant penetration) it might not be necessary to model contact, unless you need to take into account friction.

From my personal experience, including frictional contact for surfaces with intermitent contact may lead to a very non-linear response which in turn requires many small time increments in order to attain solution convergence.

If you're using CAE it is not a big deal to try different modeling approaches of the interfaces between the two materials, since there's no change in model geometry.

The modeling approach should be chosen based on if it is relevant to the overall physics you want to capture in your model.

In general, including contact modeling leads to responses with significant non-linearities which require a longer computational time.

 

[xerf]

If are not using ABAQUS you can ignore the sentence about using CAE. I thought I was replying on ABAQUS forum. My bad.

Best.

 

[cameprak]

Hi Xerf,

thanks for your immediate reply. You are right. We are planning to do using Abaqus.

1. I m not much interested in friction. Interest would be failure of Nylon part and EPP interms of stress and deformation and amount of energy taken by EPP. If we dont model contact, how load transfer happens from EPP to PA part. Avoiding friction has what kind of effect.

2. You explained about different modelling approaches. Could u explain littlebit further.

3. plastic part is with even thickness. Epp has bulk material(40 mm) in most of the places. Should i go for 3d tetra mesh (2nd order) or shell is possible.

4. Any idea abt 3d tetra meshing time with proper catia model and quality of results compared to Hexamesh (in Abaqus).

I have limited knowledge abt abaqus and i dont do this analysis in house. Design is done by me and partly responsible for development.

Thanks in advance.Any more info needed pls feel free.

cameprak

 

[xerf]

If you are using ABAQUS, then you might be interested on the ABAQUS forum i.e.: forum799

I think your questions span at least one textbook.

Contact in FE models is implemented using different formulations and may vary from FE software to FE software. ABAQUS contains different contact formulations that can be used.

1.For details see:
ABAQUS Theory Manual -> 5.1 Contact modeling
and see 2.

2.ABAQUS Analysis User's Manual
->Part VIII: Constraints
->Part IX: Interactions

To make a long story short, if you have a system made of 2 distinct materials you can:
- assign different material reponses to different elements (regions) which share common nodes.
- assign different material reponses to different elements (regions) with overlapping distinct nodes and then constrain the overlapping nodes to have the same displacements (as a generalization you can constrain only some degree of freedom, e.g. only the displacement in X-direction). Also, in general, constraints may be introduced between non-overlapping nodes.
- assign different material reponses to different elements (regions) and then use a contact formulation for the boundaries of dinstinct regions. There'are various contact formulations (as I said you can find separate texbooks on computational contact mechanics), some of them are based on contact elements, some of them are based on tracking the relative positions between pairs of user defined surfaces (i.e. FE surface=sets of nodes) and introducing/removing constraints between the nodes when penetration/separation is detected.

3. I cannot tell. There are many aspects to be considered. Usually, using structural elements such as shells is based on some assumptions regarding the stress/strain variation in the directions that are lumped (e.g., thickness direction for shells).

4. I do not have any idea. Posting this question on ABAQUS and CATIA forums may be useful for getting an answer.

Best.

 

[cameprak]

Hi,

xerf, thanks for your input.

cameprak