Hello,
I tried to obtain H-M interaction of clay surrounded by rock, however it continued to show an error message relating to convergence.
It worked fine using only Hydraulic modeling (checked only mass balance of water in equations solved), but adding stress equilibrium resulted in an error.
I checked all elastoplastic parameters and they seemed fine. I would like to have experts' help.
Could you please check my modeling and let me know what I missed?
Could you also explain time interval components (i.e., initial time step, partial time, partial time step) like how and when to use them?
Although the manual mentioned them briefly, it was little clear to me.
Thank you for your help in advance.
K
H-M Modeling
Moderator: CODE_BRIGHT Team
H-M Modeling
- Attachments
-
- Practice_1.gid.zip
- (63.79 KiB) Downloaded 310 times
Re: H-M Modeling
Dear CODE_BRIGHT Team,
I attach the latest file.
I changed mechanical boundary conditions and intervals in this model, but still got the same error message.
It continued to say that "time increment will be reduced due to problems with mechanical model. Convergence is not possible".
I checked mechanical properties and boundary conditions several times and they seemed fine.
As mentioned, using only the hydraulic option worked fine.
Could you please help me to find out reasons for the error message regarding convergence?
I really thank you for any prompt help in this matter.
Regards,
James K
I attach the latest file.
I changed mechanical boundary conditions and intervals in this model, but still got the same error message.
It continued to say that "time increment will be reduced due to problems with mechanical model. Convergence is not possible".
I checked mechanical properties and boundary conditions several times and they seemed fine.
As mentioned, using only the hydraulic option worked fine.
Could you please help me to find out reasons for the error message regarding convergence?
I really thank you for any prompt help in this matter.
Regards,
James K
- Attachments
-
- Practice_1.gid.zip
- (118.72 KiB) Downloaded 324 times
- code_bright
- Posts: 84
- Joined: Thu Dec 04, 2014 2:19 pm
- Location: Universitat Politècnica de Catalunya, Barcelona
- Contact:
Re: H-M Modeling
Hi!
You have several problems in your model. At least the following:
1) You cannot use gravity if you are using axisymmetry.
2) First interval finishes at time=10 and second interval starts at time=5.
Also, there are probably some problems in your boundary conditions.
Besides that, I recommend the following protocol in any model, specially if you do not have experience:
1) To start with an "equilibrium" interval in which the model boundaries are fixed and the are no fluxes. See if the model passes through this interval. If not, there is some problem in the definition of the model.
2) To make changes little by little, and see if each change works or if it gives any problem, so you will know then where the problem is.
You have several problems in your model. At least the following:
1) You cannot use gravity if you are using axisymmetry.
2) First interval finishes at time=10 and second interval starts at time=5.
Also, there are probably some problems in your boundary conditions.
Besides that, I recommend the following protocol in any model, specially if you do not have experience:
1) To start with an "equilibrium" interval in which the model boundaries are fixed and the are no fluxes. See if the model passes through this interval. If not, there is some problem in the definition of the model.
2) To make changes little by little, and see if each change works or if it gives any problem, so you will know then where the problem is.
- code_bright
- Posts: 84
- Joined: Thu Dec 04, 2014 2:19 pm
- Location: Universitat Politècnica de Catalunya, Barcelona
- Contact:
Re: H-M Modeling
About the variables in the interval data window, the interval starts at a given "initial time" and ends at a given "final time". The first time step would be given by "Initial time step" and there could be some convergence problems if this time step is not small enough, although if it is much smaller than neccesary, the calculation may take more time than it should.
The "partial time" is a time between the initial and final times, after which the time step will be lower than any given value in "partial time step". This is useful sometimes to avoid convergence issues and other times to have more data in the results, because if the time step is not limited, it will get bigger and bigger as long as the convergence is achieved easily, making that way the calculation faster, but giving, as I said before, less density of results, specially at the end of the time interval.
I hope you have now a more clear picture about the time interval data window.
Regards,
Alfonso
The "partial time" is a time between the initial and final times, after which the time step will be lower than any given value in "partial time step". This is useful sometimes to avoid convergence issues and other times to have more data in the results, because if the time step is not limited, it will get bigger and bigger as long as the convergence is achieved easily, making that way the calculation faster, but giving, as I said before, less density of results, specially at the end of the time interval.
I hope you have now a more clear picture about the time interval data window.
Regards,
Alfonso