[aspect-devel] strain-rate vs viscosity

Mohamed Gouiza M.Gouiza at leeds.ac.uk
Wed Jun 29 10:43:04 PDT 2016


Hi John,

Thank you for the quick reply and the explanations.
Indeed the next step is to run a setup from a published example.

I look forward to the cookbook addition.

Mohamed

Sent from a mobile device

---- John Naliboff wrote ----

Hi Mohamed,

If the initial viscosity is high, you can certainly get a few orders of magnitude decrease in the viscosity with a strain-rate dependent rheology.  It all depends on the initial viscosity profile and the strain-rate.

In general, I do not think about matching a velocity field to a certain viscosity profile.  The question one should ask is whether the initial conditions (temperature, density, etc), constitutive relationships (viscosity flow laws, plasticity, etc) and boundary conditions (velocity magnitude) are “reasonable".

My suggestion is to try and reproduce a classic extension paper from the literature.  Use the same initial conditions, boundary conditions and rheological laws and see if you get similar first-order behavior.

FYI, at some point soon there will be a cookbook for a continental extension problem added to the aspect github repository.

Cheers,
John

*************************************************
Assistant Project Scientist, CIG
Earth & Planetary Sciences Dept., UC Davis



On Jun 29, 2016, at 9:09 AM, Mohamed Gouiza <M.Gouiza at leeds.ac.uk<mailto:M.Gouiza at leeds.ac.uk>> wrote:

Hi John,

I tried several velocities but the viscosity still decreases dramatically.
What is the best way to find the viscosity profile that is appropriate to a certain velocity field?

I do understand the fact that viscosity is dependent on strain-rate but what I don’t understand is the dramatic decrease of viscosity, by 4 to 6 orders of magnitude.

Thank you.
Mohamed

From: Aspect-devel [mailto:aspect-devel-bounces at geodynamics.org] On Behalf Of John Naliboff
Sent: 23 June 2016 18:15
To: aspect-devel at geodynamics.org<mailto:aspect-devel at geodynamics.org>
Subject: Re: [aspect-devel] Mass balance

Hi Mohamed,

Brief answer As the viscosity is strain-rate dependent,  the velocity boundary conditions you applied will exert a first-order control on the viscosity distribution with the model.

In the original parameter file, I used an ‘artificially’ imposed uniform strain-rate while in your model the strain-rate field arises  from the velocity field.

So, my question is whether your new viscosity field is in fact inconsistent with the strain rate field?

Cheers,
John

*************************************************
Assistant Project Scientist, CIG
Earth & Planetary Sciences Dept., UC Davis



On Jun 21, 2016, at 4:45 PM, Mohamed Gouiza <M.Gouiza at leeds.ac.uk<mailto:M.Gouiza at leeds.ac.uk>> wrote:

Hi John,

Thank you for the script. I did a test run and it seems working fine.
I will look into it in details tomorrow.

Now I have another problem with the visco_plastic material model.
I built a parameter file using your test input file and when I run it without prescribed velocity the obtained viscosity profile looks as it should, but when I run it with prescribed velocity (on the left and right walls) the obtained viscosity profile doesn’t look right. I have attached both input files and screen captures of the two results. Any idea what’s happening?

Cheers,
Mohamed



<visco_plastic_no_prescribed_velocity.prm>
<No Prescribed Velocity.png>
<visco_plastic_with_prescribed_velocity.prm>
<With Prescribed Velocity.png>

On Jun 22, 2016, at 12:08 AM, John Naliboff <jbnaliboff at ucdavis.edu<mailto:jbnaliboff at ucdavis.edu>> wrote:

Hi Mohamed,

The code will always conserve mass according to the prescribed boundary conditions .. the question is whether your boundary conditions are balancing the influx/outflux.

If done carefully, you can for all intensive purposes match the influx/outflux with prescribed boundary velocities … a good diagnostic for this is whether your average free surface elevation is changing with time.

I’ve attached a script where I computed boundary velocities for an extension problem using a different code.  Run the script with the command “python boundaryflux.py” from the terminal.  You will need python and numpy installed.

Let me know if you have any questions about what the code is doing, although it should be fairly clear from the comments.

Cheers,
John

*************************************************
Assistant Project Scientist, CIG
Earth & Planetary Sciences Dept., UC Davis


<boundaryflux.py>


On Jun 21, 2016, at 12:20 AM, Mohamed Gouiza <M.Gouiza at leeds.ac.uk<mailto:M.Gouiza at leeds.ac.uk>> wrote:

-- why don't you just use no-stress boundary conditions at the bottom?
This way, material will simply flow through the bottom as necessary to ensure mass conservation.

Wolfgang: This is what I am doing now, but later I am going to extend my model to the lower mantle (y>650km) and I want to try mass conversation by introducing only lower mantle material, only upper mantle material, and both to see how does that impact the model.


If you want to use prescribed velocities only to ensure mass conservation, you can write out an equation where the horizontal velocity on the side >wall varies by some function and sums to zero.


For example, in the past I’ve had outflow down to 120 km, a transition from outward to inward velocity between 120 and 180 km, fixed inflow >down to 240 km and inflow at the model base.  It is a just a matter of calculating the flux (integrating velocity over length of boundary) over >different sections and setting velocities accordingly.  You can also combine the sidewall inflow/outflow with inflow at the model base.
I>f you would like, I can send the formula I used.

John: That should work too, although it wouldn't ensure total mass conservation. I would appreciate if you can send me the formula.

If I use the Box with lithosphere boundary indicators, and use zero traction boundary for the left and the right (in addition to the bottom), and prescribed velocity for left lithosphere and right lithosphere, would that work?

Cheers,
Mohamed

-----Original Message-----
From: Aspect-devel [mailto:aspect-devel-bounces at geodynamics.org] On Behalf Of Wolfgang Bangerth
Sent: 21 June 2016 07:27
To: aspect-devel at geodynamics.org<mailto:aspect-devel at geodynamics.org>
Subject: Re: [aspect-devel] Mass balance

On 06/21/2016 01:20 AM, Mohamed Gouiza wrote:



I am runing a 2d model (box with x=500km; y=600km) with prescribed
velocity
(ve) on left and right boundaries (if y>=120km & x<250km,
Ve=-1cm/year, if
y>=120km & x>=250km, Ve=1cm/year).

So I am loosing lithospheric material with time and I would like to
replace it by astenospheric matrial. This is usually achieved by
imposing a prescribed velocity (Vm) to the sublithospheric mantle
bounderies (opposite to the one imposed to the lithospheric
boundaries) to achieve masse balance with time. I can run Several
simulations and try to find an empirical relationship between Ve and Vm, but I want to know if this can be done in more clever way?!

Mohamed -- why don't you just use no-stress boundary conditions at the bottom?
This way, material will simply flow through the bottom as necessary to ensure mass conservation.

Best
 WB

--
------------------------------------------------------------------------
Wolfgang Bangerth               email:            bangerth at math.tamu.edu<mailto:bangerth at math.tamu.edu>
                                www: http://www.math.tamu.edu/~bangerth/

_______________________________________________
Aspect-devel mailing list
Aspect-devel at geodynamics.org<mailto:Aspect-devel at geodynamics.org>
http://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel
_______________________________________________
Aspect-devel mailing list
Aspect-devel at geodynamics.org<mailto:Aspect-devel at geodynamics.org>
http://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel

_______________________________________________
Aspect-devel mailing list
Aspect-devel at geodynamics.org<mailto:Aspect-devel at geodynamics.org>
http://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel

_______________________________________________
Aspect-devel mailing list
Aspect-devel at geodynamics.org<mailto:Aspect-devel at geodynamics.org>
http://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel

_______________________________________________
Aspect-devel mailing list
Aspect-devel at geodynamics.org<mailto:Aspect-devel at geodynamics.org>
http://lists.geodynamics.org/cgi-bin/mailman/listinfo/aspect-devel

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.geodynamics.org/pipermail/aspect-devel/attachments/20160629/c6ba62eb/attachment-0001.html>


More information about the Aspect-devel mailing list