[CIG-LONG] A question about Gale

Walter Landry walter at geodynamics.org
Fri Sep 25 00:10:11 PDT 2009


Taichi SATO <taichix at ori.u-tokyo.ac.jp> wrote:
> Hello,
> 
> I previously asked how to simulate oblique spreading ridge and I have
> three questions.
> 
> 1)Dr. Walter Landry gave me input file "sato.xml". However, this works
> pretty good, the temperature field does not fit the diking direction, I
> Would you tell me how to rotate temperature field parallel to the
> oblique ridge axis?

The only variation in temperature is in the vertical direction.  If
you see some horizontal variation that is pointing the wrong way, it
is coming out of the simulation.  You may want to change the size of
the box (e.g. change maxX and maxZ) and see if that changes things.

> 2)Would you tell me about the initial condition of "rifting.xml"?
> Especially, I do not understand why the force_value of divergenceForce
> is -0.8e9.

I do not remember how that particular number was chosen.  The comment
next to it is unhelpful.  In any case, the left and right sides are
both moving at 4e-10 m/s.  The dike is 2000 m across.  If the
divergence was completely accounted for in the dike, then we get a
divergence of

  2*4e-10/2000 = 4e-13

The input file scales time by 1e22, so in the scaled units that becomes

  4e9

So in this example, the divergence would only account for 20% of the
expansion.

> 3)Would you teach me the usage of Velocity_SolidBodyRotaion,
> Velocity_PartialRotationX and Velocity_PartialRotationY ?
> I think these standard condition functions are specified in velocity
> bounary condtion, but I do not understand how to specify initial
> velocity when using these functions.

I do not think you want to use these functions for this problem.  But
if you do, Velocity_SolidBodyRotation computes velocities for a solid
rotating cylinder.  Velocity_PartialRotationX and Y give the x and y
components of those velocities.  So in the velocity boundary
conditions, you would use a construct like

      <struct>
        <param name="type">WallVC</param>
        <param name="wall">right</param>
        <list name="variables">
          <struct>
            <param name="name">vx</param>
            <param name="type">func</param>
            <param name="value">Velocity_PartialRotationX</param>
          </struct>
          <struct>
            <param name="name">vy</param>
            <param name="type">func</param>
            <param name="value">Velocity_PartialRotationY</param>
          </struct>
        </list>
      </struct>

Cheers,
Walter Landry
walter at geodynamics.org


More information about the CIG-LONG mailing list