[CIG-SHORT] Question about normal slip of points along the fault

Wed Aug 23 12:28:31 PDT 2017

Teng,

My guess is that your nondimensional scales are too large for your 
simulation. In the nondimensionalization (pylithapp.problem.normalizer 
in the parameter hierarchy), you use a shear wave speed of 1 km/s and a 
minimum period of 1 s, which means I would expect your discretization 
size to be about 100 m. From your fault file, it looks like your 
discretization size if about 12.5m.

My first suggestion is to adjust the minimum period and shear wave speed 
parameters used in the nondimensionalization to better match your problem.

If you have severely distorted cells, then you should improve the mesh 
quality as severely distorted cells will introduce very stiff cells and 
stress concentrations (and increase the discretization error) in your 
solution.

Additionally, the zero_tolerance for all of your FaultCohesiveDyn faults 
is the default value of 1.0e-10. This nondimensional value is compared 
against the nondimensional slip and slip rate to determine if the fault 
is locked or sliding. If your nondimensional scales (e.g., minimum 
period) are not appropriate for your problem, PyLith may have difficulty 
resolving when to lock/unlock the fault (due to roundoff errors, we 
can't use a value of 0.0). For example, with your nondimensional scales 
of a minimum period of 1.0 s and shear wave speed of 1 km/s, the 
nondimensional length scale is 1.0 km (wavelength of the shear wave).

Finally, you are using the default value of 0.1 for the normalized 
viscosity (under pylithapp.problem.formulation in the parameter 
hierarchy). If you see high frequency oscillations in your solution that 
cannot be resolved by the discretization size you are using, then you 
should increase this value. You might need a value closer to 0.4 or 0.5.

I recommend using the PyLith parameter viewer to carefully review all of 
the parameters to make sure the values used (such as the default values) 
are appropriate for your problem.

Regards,
Brad

On 08/22/2017 03:33 PM, Li, Teng wrote:
> Hi,
> 
> In this case, the fault is the dynamic fault with slip-weakening friction model. Please find the attachment for the JSON parameters file.
> 
> And please see the fault.xml file.
> 
> Best,
> Teng
> 
> Teng Li
> 
> Master Candidate in Structures
> 
> Department of Civil and Environmental Engineering
> 
> University of Illinois at Urbana-Champaign
> 
> 205 North Mathews Ave, Urbana, IL. 61801
> 
> Phone:(217)8196210, Email: tengli2 at illinois.edu
> 
> 
> 
> ________________________________________
> From: CIG-SHORT [cig-short-bounces at geodynamics.org] on behalf of Brad Aagaard [baagaard at usgs.gov]
> Sent: Tuesday, August 22, 2017 3:22 PM
> To: cig-short at geodynamics.org
> Subject: Re: [CIG-SHORT] Question about normal slip of points along the fault
> 
> Teng,
> 
> Is this a quasi-static simulation? If so, is the solution converging at
> every time step?
> 
> If you are using PyLith v2.2.0, please send the JSON parameters file
> that is generated during the run.
> 
> Regards,
> Brad
> 
> 
> On 8/22/17 12:52 PM, Li, Teng wrote:
>> Dear Charles,
>>
>> We have a 2d model in my problem. And in this region, where I have
>> normal slip, both normal traction and shear traction become zero. In the
>> beginning, all the points in this region have positive normal slip, and
>> the maximum slip is 0.004. Then, the points have negative normal slip,
>> the values are between -0.006 to -0.003.
>>
>> And I will use Paraview to see the possible normal slip of the points
>> along the fault.
>>
>> Best,
>> Teng
>>
>> Teng Li
>>
>> Master Candidate in Structures
>>
>> Department of Civil and Environmental Engineering
>>
>> University of Illinois at Urbana-Champaign
>>
>> 205 North Mathews Ave, Urbana, IL. 61801
>>
>> Phone:(217)8196210, Email: tengli2 at illinois.edu
>>
>>
>>
>> ------------------------------------------------------------------------
>> *From:* CIG-SHORT [cig-short-bounces at geodynamics.org] on behalf of
>> Charles Williams [willic3 at gmail.com]
>> *Sent:* Monday, August 21, 2017 11:04 PM
>> *To:* cig-short at geodynamics.org
>> *Subject:* Re: [CIG-SHORT] Question about normal slip of points along
>> the fault
>>
>> Dear Teng,
>>
>> Is this a 2D or 3D problem?  If it’s 2D then slip(:,2) should be the
>> normal slip.  If it’s 3D I think it should be slip(:,3) (I’m not a
>> Matlab user, but I believe it uses 1-based indexing).  If there is
>> actually normal slip, how large is it?
>>
>> You should be able to look at the VTK files in Paraview.  Just open the
>> fault VTK file and view ‘slip’.  Then select the component you want to view.
>>
>> Cheers,
>> Charles
>>
>>
>>> On 22/08/2017, at 3:02 PM, Li, Teng <tengli2 at illinois.edu
>>> <mailto:tengli2 at illinois.edu>> wrote:
>>>
>>> Hi,
>>>
>>> I have some questions about the normal slip of the points along the
>>> fault. Please find the attachments.
>>>
>>> I output the slip(:,2) in matlab to obtain the normal slip of all the
>>> points along the fault in my model. And I use open_free_surface=False
>>> to make sure the initial traction still exists when the fault opens.
>>> However, I find certain points have normal slip. In some time steps,
>>> they have positive normal slip, while in some other times, they have
>>> negative normal slip. I am wondering why they have the normal slip in
>>> 2 different directions?
>>>
>>> And the second picture is the slip convention in pylith-2.2.0 manual.
>>> It is in the page 119/268. I think the meaning is that the positive
>>> normal slip is the fault opening direction.
>>>
>>> And in the third picture, I see negative values of fault opening
>>> implys penetration.
>>>
>>> Since the penetration is never allowed, I wonder why we have both
>>> positive and negative slip results? Is it due to the incomplete or
>>> wrong .vtk output?
>>>
>>> And is there a way to see the fault opening using Paraview? I think
>>> using Paraview to visualize those points can help me figure out the
>>> meaning of their normal slips.
>>>
>>> Hope for your reply!
>>>
>>> Best,
>>> Teng
>>>
>>>
>>>
>>> Teng Li
>>>
>>> Master Candidate in Structures
>>>
>>> Department of Civil and Environmental Engineering
>>>
>>> University of Illinois at Urbana-Champaign
>>>
>>> 205 North Mathews Ave, Urbana, IL. 61801
>>>
>>> Phone:(217)8196210, Email: tengli2 at illinois.edu
>>> <mailto:tengli2 at illinois.edu>
>>>
>>>
>>>
>>> <Positive_values.png><Negative.png>_______________________________________________
>>> CIG-SHORT mailing list
>>> CIG-SHORT at geodynamics.org <mailto:CIG-SHORT at geodynamics.org>
>>> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>>
>> *Charles Williams I Geodynamic Modeler
>> GNS Science **I** Te Pῡ Ao*
>> 1 Fairway Drive, Avalon 5010, PO Box 30368, Lower Hutt 5040, New Zealand
>> *Ph* 0064-4-570-4566 I *Mob* 0064-22-350-7326 I *Fax* 0064-4-570-4600*_
>> _**http://www.gns.cri.nz/* *I* *Email: **C.Williams at gns.cri.nz*
>> <mailto:your.email at gns.cri.nz>
>>
> 
> _______________________________________________
> CIG-SHORT mailing list
> CIG-SHORT at geodynamics.org
> http://lists.geodynamics.org/cgi-bin/mailman/listinfo/cig-short
> 