Three point bending test

[Yaakov]

Dear FEAP Team & Users,

I want to make a three point  bending  test in FEAP .  I used the command Rigid,,2 and Rigid,3 in input file.
I could not get the results.
I have attached the input file.
can you please let me know error in the input file.
thanks very much for your suggestion!

Kind regards,
Yaakov

[Prof. R.L. Taylor]

What do you want to be rigid?  The circular object?

[Yaakov]

Dear Prof. Taylor,

Thanks very much for your answer!

 Only the circular object is to be modelled as rigid body.

[Prof. R.L. Taylor]

The setting of element groups to rigid requires the list of ELEMents to be split into separate groups: one for each rigid body and the rest for the flexible body.  Thus, all the elements for the circular disk are put into a group as follows:

RIGID,1
ELEMENTs
  list of connections for all elements to be grouped as a single rigid body
!blank record

FLEXRible
ELEMENTs
list of flexible elements.

Note you can also declare a surface to be RIGId in the contact and not have to model the disk.

[Yaakov]

Dear Prof. Taylor,

Thanks very much for your support!
I have an additional question:

can I define the rigid body in this way:

MATErial 2
SOLId
ELAStic ISOTropic 1+e100000 0

I have checked and there is almost no strain & stress values in Part 2 (Rigid Part).

Wish you a happy weekend ahead.

Kind Regards,
Yaakov

[Prof. R.L. Taylor]

There are two reasons this is not an ideal solution.

1.) It is not rigid and you may have ill-conditioning between parts of the analysis that have much smaller properties.

2.) All the nodes in the "stiff" body have unknowns and you solve for a lot of equations unnecessarily; Rigid bodys have 3 equations in 2-d and 6 in 3-d so are more efficient.  However, they are harder to set up for the analysis since they have separate data definitions for loads and restraints -- they are not on the nodes.

[Yaakov]

Dear Prof. Taylor,

Thank you very much!

Kind  regards,
Yaakov

[Yaakov]

Dear Prof. Taylor,

I have used new commands (flex + rigi) in input file, it shows me wrong results (rigid body has stress&strain ...)
If I use RIGI,3 (end mesh), It shows more bugs.
 
I have attached the new input file IEX8.

thanks very much for your suggestion!

Kind regards,
Yaakov

[Yaakov]

Dear All,

There is still no right solution for the typical 3 point bending test in FEAP.

we checked that the rigid body definition should be correct (feap manual84).


Warm regards,
Yaakov
 

[Prof. R.L. Taylor]

Your mesh IEX8 has several errors.

First half of the elements on the ball are rigid and half flexible.  This probably will not work very well, but you can try later.

Secondly, the problem needs to be finite deformation for the rigid body to work as you describe (the solution does not have to produce finite deformations -- but contact is really described for this as is rigid body behavior).

There is a command to "fix" a rigid body, you do not do it by fixing the nodes with BOUN,  use RBOU (see appendix in manual), currently the application of a displacement to a rigid body is not coded, thus it is necessary to move the flexible body (maybe you can figure out how to implement applying RDIS (the command is there, but not the code to do it).

With some changes to the input file it seems to run now for me.

[Yaakov]

Dear Prof. Taylor,

thanks very much for your answer!

The new input file works very well now. I would ask, I can get the values of the strain/stress in rigid body, are the solutions logical/correct?

Thanks very much! & Warm Regards,
Yaakov
 

[Prof. R.L. Taylor]

Except for the computation of geometric and inertia values for a rigid body we do not process elements to output or compute local properties, thus it is not possible to print the "zero" strain/stress values in a rigid body.

You can print the displacements for a rigid body  and if no rotations have occurred then the values should be constant.  If the body rotates, then the displacements at nodes will reflect this.

[Yaakov]

Dear Prof. Taylor,

Thanks very much for your answer!

I would like to define the displacement (RBOU) on the rigid body as I need the displacement-force curve for ASTM numerical experiment.

Kind Regards,
Yaakov

[Prof. R.L. Taylor]

The boundary displacements of the beam are the negative of the rigid body displacement, so you have what you need with the analysis the way it is.

[Yaakov]

Dear Prof. Taylor,

Thanks very much, I will do!

Warm Regards, Yaakov