Hello all,
I am modelling a deep excavation using PLAXIS. At some particular excavation level, it has been that the analysis is taking quite a long time even if the global error is less than the tolerance specified in the numerical parameters. Snap shot of the running analysis is attached below. Is there anything else that influences the analysis? Suggestions would be really helpful.
Regards
Prasanthi
Reasons for possible delay
-Multi-layered soils
-using soil models such as the Hardening model and HS small models are very iterative and will take much more time than Mohr-Coulomb
-The load might be very high, try a smaller load just to check
-The coarseness/Finnes of the mesh
this might happen through following analyses:
coupled time-dependent analyses
very fine mesh generation used
consolidation analysis with high plastic deformation
Thank you all,
In this model, multi-layered soil is present and soil model used is Mohr-Coulomb and the analysis is "plastic". I tried with different coarseness factors. If the graph between Mstage and Disp. in the above image is closely observed, it is clear that the graph is horizontal. As fas as I know, constant value of Mstage with increasing displacements is an indication of Non-convergence of Model. My doubt is, global error is less than the tolerance but still the program is not converging. What might be the reason for this (or) is anything wrong in my understanding. Suggestions would be helpful.
Regards
Prasanthi.
Hello Prashanthi,
I'm having the same problem, have you managed to find a solution?
Your help would be much appreciated. Thank you,
Dina
I think one way to know if the failure mechanism develops is through phi-c analysis
https://www.plaxis.com/content/uploads/import/kb/kb-downloads/PHI-C%20REDUCTION%20AND%20COMPARISON%20WITH%20BISHOP%27S%20METHOD.pdf
or the last step in this video
https://www.youtube.com/watch?v=2eIu802ghq8
Dear Prashanthi,
The following points may be useful to you,
1. Have you started your analysis with correct mesh size and correct mesh density..this can be achieved by trial analysis ..
2. Soil model..Mohr-Coulomb model is good upto a shallow depth of excavation..you think for other kind of model i.e. hardening model is best for deep excavation..
3. Computational time..it is purely depends on type of soils in a multi-layered soil medium, mesh size and density (This can be corrected by doing trail analysis - find out correct size and density for your analysis), consolidation analysis - coupled analysis (soil-water interaction analysis). You know very well that load increases then displacement increase and may reach to plastic flow condition ( it is quite possible in higher excavation load level)..
4. You have mentioned that the time taken for completing the analysis is more even global error is less than tolerance error...yes, it is possible in excavation kind of problem mainly at higher excavation load level..Since analysis performed in incremental form in each incremental steps, if the outer balancing forces or displacement is higher than the acceptable limits then this force will be carried forwarded to next iterations..so the iteration process will be continued up to the stage at which the balancing forces is in acceptable limit (this stage is called your solution is converged)..Some time, you may force to converge the analysis by defining convergence criteria..but it is not acceptable for the forced convergence mainly on excavation kind of problem..you know the reason, why we should not force the analysis to converge..
5. I feel you should not worry about the time taken to complete the analysis..if you understand the above mentioned points then you may not worry about the time but results obtained from the analysis needs to be exact solution..hope you understand all these points as mentioned above.. all the best in your research..
The problem is more simple than you think. Just look at the number of inacurate plastic points that you have and the most important CSP is close to zero. Don't worry the output is "the key" of your answer here.
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Publication