Debugging constraints on rigid bodies

[wwang123]

Hello!

I've run into a problem in the loads/solutions of my model. I want to model a multilayered composite out of shell281, of which one layer is non-uniform (i.e. it has a non-zero thickness in some areas and a zero thickness in others). To accomodate this, I've created separate areas associated with different multilayered sections that has the last layer and does not have the last layer.


I've been able to run the same loads/solution/postprocessing code for a similar file. However, unlike that aforementioned file, the similar file is just composed of one area associated with one multilayered section as all thicknesses are uniform.

The rest of the system consists of a target surface (targe170) and contact surface (conta174) in the set up of a 4-point bending test.

I'm getting the following error:

"The value of UZ at node 1824 is 3387349.3. It is greater than the current of 1000000. This generally indicates rigid body motio nis a result of an unconstrained model. Verify that your model is properly."

Suggestion/hints as to what this entails/means (e.g. commands I need to change or check this)?

I've tried to find this node in the GUI, but it doesn't seem to exist when activating plotctrls.
If more information is needed, please post!



I've been able to run a similar file with the same code for loads, solutions, and postprocessing

 

[lopessilva]

Hello,

I would clear the mesh and perform the "glue" boolean operation on the areas. If glue is not possible, try "overlap". Then do the rest of the procedure again and solve the model. It may happen that ANSYS does not recognize the connection between the several layers or areas you have. If it does not work, you may have
to further constraint the model, either by contraining more DOF, if it is still adequate to the physical behaviour of the phenomenon you want to to study, but instead of messing around with the DOF thing, I would try to constraint the model better (you can do it on solution->loads->displacement). Apply it on keypoints (simply supported). If it still can´t solve the model, apply it on the entire area (cantilevered).

regards,
Hugo