I modelling design of umbrella arch and pipe arch in Plaxis 3D. But I could not excavate while actived embedded beams. We encounter failure slow convergence (204). I read reference manual page 486. But I could not solutions this model. I will appriciate if you will answer my question or send an example for staged excation in embedded beams aplication.
Thanks for your support.
I think there is a singularity will be happened during active embedded beams. you need to check the geometry of the structure (tunnel) and I think you find a very small part of the beams in the tunnel and this causes the singularity. You need to plot the structure in different way.
Hi Mr. Bakr,
I think a problem on version 2018.01 because I checked always geometry, lines and points but I encounter always this problem. I cannot calculate this version with embedded beams or without embedded beams. If you want I could share to you this sample.
Thanks for interest
I had the same error with similar model (tunnel with embedded beams). My solution: I changed to Pardiso solver type instead of Picos solver. What was interesting: This error came up on the system with new AMD 12-core CPU and WIN 10 while on older system with i7 and WIN7 Picos worked fine. In both cases 2018.01 version was used.
most likely , there is a problem with applying the boundary conditions
Mr. Aras,
I know it may be late, my opinion is that in this case is not mesh related, but rigidity related, something very small but very rigid (embedded beams) next to something large and less rigid (soil).
I see a workaround using volumes (square shaped that have the same perimeter as the circle).
It may seem harder, but volumes work best with volumes, whereas with plates or beams may not always work. For axial forces in the volumes use Plaxis "free cross section" feature to create perpendicular cross section of a particular former beam, now turned volume, and Plaxis will provide you with an effective normal stress (i.e. the axial force).
I am curious how you worked it out.
Kind regards
The same problem with model of pile foundation in step of pile construction, I hope to fine solution. Is the reason with pile interface of what?
Mr. Aziz,
Try changing the Solver Type from Picos to Classic. It works for my in some models. In those cases I think it's interface related, when Rinter small, for example 0.01, because it modifies the friction angle and soil cohesion.
If you choose to change from Picos to Classic, you should also change the Solver Type of all subsequent Phases to Classic.
Hope it does the trick for you, as well.
Kind regards
Hi Mr. Aziz,
I recommend classic or pardiso solver for your model. But classic slower than pardiso. Firstly you choose solver type and change max steps, tolerated error, max unloading steps, max loading friction per steps and activated use line search and use gradual error reduction on numerical control parameters. If you change parameters (values) I think you can calculate model.
Thanks
Moreover, the same error is possible due to the overlapping of elements like surfaces. You can check the positions of surfaces.
Thanks.
- Badges
- Science topic
- Similar topics
- Biological Science
- Physiology
- Physiological Processes
- Stress