[CIG-SHORT] Fault direction in 3D fault
Jiangzhi Chen
jzchenjzarthur at gmail.com
Sun May 4 19:21:33 PDT 2014
Thank you very much. I will run it and see if it solves my problem. I am not sure if I can beat down the condition number, but I will try smoothing with some other schemes.
Jiangzhi
On May 4, 2014, at 6:58 PM, Charles Williams <willic3 at gmail.com> wrote:
> It appears that the problem is that your fault at the bottom of the slab is completely horizontal over much of its extent. This causes problems for PyLith when the default up_dir [0,0,1] is used. The way to diagnose this problem is to have PyLith output the orientation information for the faults, as Brad suggested previously. When you plot the strike_dir as vectors, you will see it varies quite a bit for the horizontal portion of the fault, because the direction is undefined. The solution is to choose a vector that does not align with the fault normal vector. I chose a vector that points slightly to the west. When this is done, the various fault slip directions are well-defined.
>
> I’m attaching a slightly revised interseismic.cfg file that has a few changes. Try this out and see if the results are reasonable. Also, I would spend some time trying to improve the quality of your mesh.
>
> Cheers,
> Charles
>
>
> <interseismic.cfg>
>
> On 5/05/2014, at 12:52 pm, Jiangzhi Chen <jzchenjzarthur at gmail.com> wrote:
>
>> After I generated the mesh, I smoothed it using
>>
>> cleanup volume all
>> volume all smooth scheme condition number beta 2.0 cpu 4
>> smooth volume all
>>
>> Maybe I should give it a smaller beta number.
>>
>> On May 4, 2014, at 5:30 PM, Charles Williams <willic3 at gmail.com> wrote:
>>
>>> I’m trying to run your problem right now, but it’s extremely slow. After looking at your mesh, I realized that the quality is extremely poor, especially near the fault. This will probably affect the convergence speed as well as the accuracy of your solution. Did you smooth the mesh at all?
>>>
>>> Cheers,
>>> Charles
>>>
>>>
>>> On 5/05/2014, at 10:29 am, Jiangzhi Chen <jzchenjzarthur at gmail.com> wrote:
>>>
>>>> I used field split in calculation, and used a specific solver. The linear solver usually converges after 3 or 4 iterations.
>>>>
>>>> [pylithapp.timedependent.formulation]
>>>> split_fields = True
>>>> matrix_type = aij
>>>> use_custom_constraint_pc = True
>>>> [pylithapp.petsc]
>>>> fs_pc_type = fieldsplit
>>>> fs_pc_fieldsplit_type = schur
>>>> fs_pc_fieldsplit_schur_factorization_type = full
>>>> fs_pc_fieldsplit_real_diagonal = true
>>>> fs_fieldsplit_0_pc_type = gamg
>>>> fs_fieldsplit_0_ksp_type = preonly
>>>> fs_fieldsplit_0_ksp_rtol = 1.0e-10
>>>> fs_fieldsplit_1_pc_type = none
>>>> fs_fieldsplit_1_ksp_type = minres
>>>> fs_fieldsplit_1_ksp_rtol = 1.0e-10
>>>>
>>>> # Convergence parameters.
>>>> ksp_rtol = 1.0e-8
>>>> ksp_atol = 1.0e-12
>>>> ksp_max_it = 100
>>>> ksp_gmres_restart = 50
>>>>
>>>> # Linear solver monitoring options.
>>>> ksp_monitor = true
>>>> ksp_view = true
>>>> ksp_converged_reason = true
>>>>
>>>> # Nonlinear solver monitoring options.
>>>> snes_rtol = 1.0e-8
>>>> snes_atol = 1.0e-12
>>>> snes_max_it = 100
>>>> snes_monitor = true
>>>> snes_view = true
>>>> snes_converged_reason = true
>>>>
>>>> I attached the mesh and pylith configurations.
>>>>
>>>> <test.zip>
>>>> On May 4, 2014, at 2:56 PM, Charles Williams <willic3 at gmail.com> wrote:
>>>>
>>>>> Hi Jiangzhi,
>>>>>
>>>>> The up_dir is just used to help PyLith figure out how your coordinate system is set up. Not specifying it for your slabbot fault may be causing some problems, but I also suspect there may be some inconsistencies between your boundary conditions and your applied fault slip. Also, when you run the problem are you getting convergence in the linear solution?
>>>>>
>>>>> Cheers,
>>>>> Charles
>>>>>
>>>>>
>>>>> On 5/05/2014, at 9:38 am, Jiangzhi Chen <jzchenjzarthur at gmail.com> wrote:
>>>>>
>>>>>> Thanks for your advice, I am still trying to get the slab move correctly. The part I am confused is that the slab just sinks down, as shown in the gif figure in my previous email. Is this up_dir related to the fault type, i.e., reverse fault or normal fault? In my case, the slab top is a reverse fault, and slabbot is a normal fault. Maybe the reason is that I did not specify up_dir for the slabbot fault.
>>>>>>
>>>>>> Jiangzhi
>>>>>>
>>>>>> On May 4, 2014, at 2:20 PM, Charles Williams <willic3 at gmail.com> wrote:
>>>>>>
>>>>>>> Hi Jiangzhi,
>>>>>>>
>>>>>>> Did you make any more progress on this? Were the strike_dir, dip_dir, etc., consistent with what you thought? These are just direction cosines, so you can plot them as vectors to see what they look like.
>>>>>>>
>>>>>>> Also, I noticed that you only specified up_dir for one of your faults. I believe that the default is [0,0,1], but it’s probably a good idea to specify this as well for consistency.
>>>>>>>
>>>>>>> Cheers,
>>>>>>> Charles
>>>>>>>
>>>>>>>
>>>>>>> On 3/05/2014, at 8:03 am, Jiangzhi Chen <jzchenjzarthur at gmail.com> wrote:
>>>>>>>
>>>>>>>> Hi Brad and Charles,
>>>>>>>>
>>>>>>>> Thanks for your advice. First I tried Brad's suggestion, but the slab still went to wrong direction (screen1 and screen2). To be clear, the model simulates the Cascadia subduction zone, X is easting, Y is northing, and Z is up. Currently I have slabtop and slabbot both creeping at 8 cm/yr, and the slab is supposed to go down with slight rotation because of the geometry of the subduction boundary.
>>>>>>>>
>>>>>>>> Then I tried the suggestion from Charles, changed the bc_dof on the side to 1, and bc_dof on the bottom to 2, with up_dir [0,0,1]. This time, the slab did go down, but it just sank all way down (slab.gif).
>>>>>>>>
>>>>>>>> Jiangzhi
>>>>>>>>
>>>>>>>> On 05/01/2014 04:53 PM, Brad Aagaard wrote:
>>>>>>>>> Jiangzhi,
>>>>>>>>>
>>>>>>>>> If this is the same model you used a while ago, then I believe you use [0,1,0] as the up_dir for the domain. You need to set the up_dir to [0,1,0] for all faults and boundaries.
>>>>>>>>>
>>>>>>>>> You should request that the fault orientation be information be written so that you can see the strike_dir, dip_dir, and fault normal directions for your geometry. This will help you debug the fault slip directions. See examples/3d/hex8/step06.cfg for how to request the fault directions be output.
>>>>>>>>>
>>>>>>>>> Brad
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> On 5/1/14, 4:25 PM, Charles Williams wrote:
>>>>>>>>>> Hi Jiangzhi,
>>>>>>>>>>
>>>>>>>>>> One problem that I see is that I think you’ve mixed up the bc_dof for your bottom boundary (should be 2) compared to your front and back boundaries (should be 1). Let me know if this helps.
>>>>>>>>>>
>>>>>>>>>> Cheers,
>>>>>>>>>> Charles
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> On 2/05/2014, at 10:26 am, Jiangzhi Chen <jzchenjzarthur at gmail.com> wrote:
>>>>>>>>>>
>>>>>>>>>>> Hi everyone,
>>>>>>>>>>>
>>>>>>>>>>> I have a 3D subduction model simulating the interseismic deformation, where I have both fault_slabtop and fault_slabbot creeping. I have set fault_slabtop with reverse_slip of 8 cm/yr and fault_slabbot with reverse_slip of -8 cm/yr, and used the up_dir parameter [0,0,1], but still the subduction is in wrong direction. I am not sure what is wrong. The configuration files are attached. Thanks.
>>>>>>>>>>>
>>>>>>>>>>> Jiangzhi
>>>>>>>>>>> <screen.png><pylithapp.cfg><interseismic.cfg><mat_conmantle.spatialdb><mat_oceanlitho.spatialdb><mat_oceanmantle.spatialdb><mat_conlitho.spatialdb>_______________________________________________
>>>>>>>>>>> CIG-SHORT mailing list
>>>>>>>>>>> CIG-SHORT at geodynamics.org
>>>>>>>>>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>>>>>>>>
>>>>>>>>>> Charles A. Williams
>>>>>>>>>> Scientist
>>>>>>>>>> GNS Science
>>>>>>>>>> 1 Fairway Drive, Avalon
>>>>>>>>>> PO Box 30368
>>>>>>>>>> Lower Hutt 5040
>>>>>>>>>> New Zealand
>>>>>>>>>> ph (office): 0064-4570-4566
>>>>>>>>>> fax (office): 0064-4570-4600
>>>>>>>>>> C.Williams at gns.cri.nz
>>>>>>>>>>
>>>>>>>>>> _______________________________________________
>>>>>>>>>> CIG-SHORT mailing list
>>>>>>>>>> CIG-SHORT at geodynamics.org
>>>>>>>>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> _______________________________________________
>>>>>>>>> CIG-SHORT mailing list
>>>>>>>>> CIG-SHORT at geodynamics.org
>>>>>>>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>>>>>>
>>>>>>>> <screen1.png><screen2.png><slab.gif>_______________________________________________
>>>>>>>> CIG-SHORT mailing list
>>>>>>>> CIG-SHORT at geodynamics.org
>>>>>>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>>>>>
>>>>>>> Charles A. Williams
>>>>>>> Scientist
>>>>>>> GNS Science
>>>>>>> 1 Fairway Drive, Avalon
>>>>>>> PO Box 30368
>>>>>>> Lower Hutt 5040
>>>>>>> New Zealand
>>>>>>> ph (office): 0064-4570-4566
>>>>>>> fax (office): 0064-4570-4600
>>>>>>> C.Williams at gns.cri.nz
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> CIG-SHORT mailing list
>>>>>>> CIG-SHORT at geodynamics.org
>>>>>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>>>>
>>>>>> _______________________________________________
>>>>>> CIG-SHORT mailing list
>>>>>> CIG-SHORT at geodynamics.org
>>>>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>>>
>>>>> Charles A. Williams
>>>>> Scientist
>>>>> GNS Science
>>>>> 1 Fairway Drive, Avalon
>>>>> PO Box 30368
>>>>> Lower Hutt 5040
>>>>> New Zealand
>>>>> ph (office): 0064-4570-4566
>>>>> fax (office): 0064-4570-4600
>>>>> C.Williams at gns.cri.nz
>>>>>
>>>>> _______________________________________________
>>>>> CIG-SHORT mailing list
>>>>> CIG-SHORT at geodynamics.org
>>>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>>
>>>> _______________________________________________
>>>> CIG-SHORT mailing list
>>>> CIG-SHORT at geodynamics.org
>>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>>
>>> Charles A. Williams
>>> Scientist
>>> GNS Science
>>> 1 Fairway Drive, Avalon
>>> PO Box 30368
>>> Lower Hutt 5040
>>> New Zealand
>>> ph (office): 0064-4570-4566
>>> fax (office): 0064-4570-4600
>>> C.Williams at gns.cri.nz
>>>
>>> _______________________________________________
>>> CIG-SHORT mailing list
>>> CIG-SHORT at geodynamics.org
>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>
>> _______________________________________________
>> CIG-SHORT mailing list
>> CIG-SHORT at geodynamics.org
>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>
> Charles A. Williams
> Scientist
> GNS Science
> 1 Fairway Drive, Avalon
> PO Box 30368
> Lower Hutt 5040
> New Zealand
> ph (office): 0064-4570-4566
> fax (office): 0064-4570-4600
> C.Williams at gns.cri.nz
>
> _______________________________________________
> CIG-SHORT mailing list
> CIG-SHORT at geodynamics.org
> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
More information about the CIG-SHORT
mailing list