Model non convergence after changing spring from linear to non linear

[NNBL]

Hello everyone

I have robot model  which is properly running for linear springs.  However when I run the model after changing the spring from linear to non-linear (i.e. uplift allowed in UZ+) it doesn't converge even for self weight. I tried applying rotational stiffness also, but that also didn't work. I am sharing robot model 

Regards

Nitin Bansal 

[Krzysztof_Wasik]

Hi @NNBL 

Corect parameters for load case 1

• Activate matrix update after each iteration
• Increase tolerance  (0.001 should work)

Refer to 

Non convergence while model calculation in Robot Structural Analysis

[NNBL]

Helllo Mr. Wasik, 

Many thanks for the solution.  It worked for the load case 1,  however it doesn't work for many other load cases. I am not sure how the tolerance is decided and what will its impact on the results? I have shared the robot file, can you please help in analyzing load combination 9100 & 9101?

Regards

Nitin Bansal 

[Krzysztof_Wasik]

Hi @NNBL 

Both cases converge with tolerance about 0.2 

I see that you have used imposed loads. Keep in mind that they are not working when applied in nonlinear supports (with uplift defined). Only rigid or elastisc(no uplift) supports can be used for nodes with imposed loads applied

[NNBL]

Hello @Krzysztof_Wasik 

I am not sure if it is correct to say that non-linear spring and imposed displacement doesn't work together.  I have analyzed multiple models in the past with such situations, they are running without any error. One of the models is attached for your reference. Please see and let me know if I am correctly understanding your message. 

Thanks & Regards

Nitin Bansal 

[Krzysztof_Wasik]

Hi @NNBL 

I would recommend following Robot Help limitations. Imposed displacement applied in the nodes  supported with supports with uplift can cause incorrect results.

Please have a look at the simple example below where imposed displacement defined for support with uplift, results in incorrect displacement

Example:

For two identical beams imposed displacement UZ=-10 cm was applied to nodes 1 and 3.

Nodes 2 and 4 are fixed.

Node 3 is supported with elastis support  in both UZ+and UZ-.

Node 1 is supported with elastic support in UZ- (UZ+ uplift considered)

For both beams displacement equal to imposed value (close to uz=10 cm  since support is elastic) is expected in nodes 1 and 3

After calculation expected results are obtained only for node 3 (supports with no uplift defined) while for  node 1 (uplift defined) incorrect displacement is presented. 

I would not recommend using imposed displacement, in nodes supported with supports, with uplift defined (following Robot Help description).

[NNBL]

Hi @Krzysztof_Wasik 

Thanks for your efforts!, 

I created the robot model similar to what you had created. Contrary to your observations, I am getting the same results for both the beams (see the image below). If possible can you please share your model, as I am bit concerned about many similar models on which my team is currently working. 

[Krzysztof_Wasik]

Hi @NNBL 

Please verify results in your mode after changing section to concrete beam 30x50.

[Simau]

Hi @Krzysztof_Wasik 

Please why i get same deformation in this example

[Krzysztof_Wasik]

Hi @Simau 

You have applide imposed displacement in nodes 1 and . We have discussed issue with imposed displacement application in nodes 2 and 4 (elastic supports with or without uplift defined)

Also bigger imposed displacement (UZ=-100 mm) was used.

[Simau]

@Krzysztof_Wasik 

Sorry,

my mistake!

[NNBL]

Hi @Krzysztof_Wasik 

Thanks for clarifying.

Can you please suggest any solution to this problem, as I to have change many models.

[Krzysztof_Wasik]

Hi @NNBL 

As described in Robot Help, imposed displacement works when applied in the node with rigid or elastic support defined. Please keep in mind that when imposed displacement is applied to node with elastic support defined, structure displacement might be different that imposed displacement value, especially when imposed displacement is not uniformly defined for whole model (node displacement equal to imposed displacement is obtained only for rigid support).

Example

When you would like to define imposed displacement for nonlinear elastic supports, I would suggest definition of separate load case and simulating expected initial structure displacement/deformation with forces  (not imposed displacements).