Problem in interaction between roller and laminate.

[fifiariff]

Hii all,

I am trying to model four point bending of composite laminate with a hashin damage model using Abaqus 2017. The configuration of the CFRP laminate is [0/90]24. After the job completed, the result shows the rollers are not moving down as I expected (as shown in the figure)
Can someone help me to solve this problem?

Thank you.

 

[FEA way]

There is a deflection so it must be working at least to some extent. But you are showing the results at step time 0.1256. Is that the whole analysis ?

 

[fifiariff]

No, the step time is 1. But until the end of the time, the roller still not moving down.

 

[FEA way]

But the beam keeps bending ? What is the deformation scale factor - 1 ? How is the roller movement defined ?

 

[SWComposites]

pls show the applied loads and boundary conditions.
is this a static or dynamic analysis?
why are you modelling the rollers and not just applying loads?

 

[fifiariff]

Yes, the beam keeps bending.The roller movement defined by the roller keeps attached the beam until end process.
This is static analysis. I defined the rollers as a rigid body. Then I model the frictionless interaction between the roller and beam.
So,I model the roller to check the effect of the roller to the beam.
The applied loads and boundary condition are shown in the attached figure.

Thank you.

 

[FEA way]

So there’s a rigid body reference point (with prescribed displacement) attached to the rigid roller. Then there are rigid bottom supports with all DOFs fixed. But what is that additional rigid body marked by you (in the middle of the beam, with U1=UR2=UR3=0) ? Is that a rigid surface ? What’s the purpose ?

 

[SWComposites]

I model the roller to check the effect of the roller to the beam. complete waste of time. real rollers are not rigid. specimens are actually tested with pads/blocks between the rollers and the specimen. See ASTM D7249 and

https://www.compositesworld.com/articles/flexure-testing-of-sandwich-composites

All the rigid elements just complicate the model and apply unrealistic boundary conditions and can lead to numerical problems.