Why is my part not rotating unless I make it a rigid body.

matthijs

I ran into an issue with a simulation where a bottle was supposed to fall over. I put the bottle on an edge with its COG not above the edge. I applied gravity to the bottle, so I was expecting gravity to cause the bottle to tip over and fall of the edge. Instead, the bottle remained perfectly upright and did not move. 

Some additional information about the simulation:

• I used aluminium for the bottle and the edge
• The edge surface is completely fixed 
• I applied a sliding contact between the bottles and the edge (TYPE7)
• Both the bottle and the edge are meshed using a shell mesh

Eventually I tried using a RBODY. The results of this test can be seen below. All nodes in the blue bottle are connected by a RBODY. This did have the desired effect. However, the blue bottle can now no longer deform. For the purpose of this simulation this is not an issue, but I was wondering if there is a way to have the bottle fall over without using a RBODY. Also, I would be interested in an explanation as to why the bottle without the RBODY refuses to rotate.

Figure - bottles at t = 0.0 s (LEFT), bottles at t = 0.2 s (RIGHT)

Polyvios Romanidis

Matthijs Quist said:

Hi Polyvios,

The _0000.out file does not show any warnings or errors, so I don't think there is something wrong in terms of penetrations. Perhaps it has something to do with the contact or gravity as you mentioned. 

I have attached my model file. Hopefully you can figure out my mistake.

Thank you in advance,

Matthijs

Hi Matthijs,

Thanks for sharing the model.  What you describe is indeed strange and normally both items (with and without /RBODY defined) should fall as expected.

We are investigating now what is the exact cause of this error, but normally everything should work.

For now I can propose you a way to overpass this problem.  You can change your material for /MAT/LAW1 (ELASTIC) to /MAT/LAW2 (ELASTO-PLASTIC) and set a yield point equal to 1e+30.  In this way your material will have an elastic behavior as now and the bottle will fall normally as expected.  It is also worth to use more than 1 integration points in /PROP/SHELL, so I will recommend N = 5.

Polyvios

Polyvios Romanidis

Hi Matthijs,

Both models should work normally and /RBODY formulations should not bring any effect like the one you have described.

Probably, an error occur in the set up of the model maybe with initial penetrations or the application of gravity load.  From the data you provided is not clear what is the fault.

If you could share the model, I could find out what is wrong and give you some feedback!

Thanks,

Polyvios

matthijs

Polyvios Romanidis said:

Hi Matthijs,

Both models should work normally and /RBODY formulations should not bring any effect like the one you have described.

Probably, an error occur in the set up of the model maybe with initial penetrations or the application of gravity load.  From the data you provided is not clear what is the fault.

If you could share the model, I could find out what is wrong and give you some feedback!

Thanks,

Polyvios

Hi Polyvios,

The _0000.out file does not show any warnings or errors, so I don't think there is something wrong in terms of penetrations. Perhaps it has something to do with the contact or gravity as you mentioned. 

I have attached my model file. Hopefully you can figure out my mistake.

Thank you in advance,

Matthijs

Polyvios Romanidis

Matthijs Quist said:

Hi Polyvios,

The _0000.out file does not show any warnings or errors, so I don't think there is something wrong in terms of penetrations. Perhaps it has something to do with the contact or gravity as you mentioned. 

I have attached my model file. Hopefully you can figure out my mistake.

Thank you in advance,

Matthijs

Hi Matthijs,

Thanks for sharing the model.  What you describe is indeed strange and normally both items (with and without /RBODY defined) should fall as expected.

We are investigating now what is the exact cause of this error, but normally everything should work.

For now I can propose you a way to overpass this problem.  You can change your material for /MAT/LAW1 (ELASTIC) to /MAT/LAW2 (ELASTO-PLASTIC) and set a yield point equal to 1e+30.  In this way your material will have an elastic behavior as now and the bottle will fall normally as expected.  It is also worth to use more than 1 integration points in /PROP/SHELL, so I will recommend N = 5.

Polyvios

matthijs

Polyvios Romanidis said:

Hi Matthijs,

Thanks for sharing the model.  What you describe is indeed strange and normally both items (with and without /RBODY defined) should fall as expected.

We are investigating now what is the exact cause of this error, but normally everything should work.

For now I can propose you a way to overpass this problem.  You can change your material for /MAT/LAW1 (ELASTIC) to /MAT/LAW2 (ELASTO-PLASTIC) and set a yield point equal to 1e+30.  In this way your material will have an elastic behavior as now and the bottle will fall normally as expected.  It is also worth to use more than 1 integration points in /PROP/SHELL, so I will recommend N = 5.

Polyvios

Hi Polyvios,

Thank you for the suggestion. Changing the material and the number of integration points does resolve my issue. Apparently the number of integration points was the issue. An elastic material with global integration, (N = 0), also worked.

Using membrane elements (N = 1) for this application was wrong in the first place so it does not really matter for me that they do not rotate.

Anyway, thanks for your help,

Matthijs