Setting basal pressure til fraction of overburden

Extension of Elmer in computational glaciology
Post Reply
ChristianB
Posts: 10
Joined: 28 Apr 2014, 22:02
Antispam: Yes

Setting basal pressure til fraction of overburden

Post by ChristianB »

Hi,

I am trying to set the basal pressure equal to 60% of my ice overburden pressure. My domain contains both ice-covered and ice-free areas. So far I have tried setting external pressure on the basal boundary everywhere like this,

Code: Select all

  External Pressure = Variable ZsDEM, bedrockDEM
    Real MATC "(tx(0)-tx(1))*910.0*9.82*0.6"
and only setting it in areas where the ice is more then 5 meters,

Code: Select all

  External Pressure = Variable ZsDEM, bedrockDEM
    Real MATC "if (tx(0)>tx(1)+5.0) {(tx(0)-tx(1))*910.0*9.82*0.6} else {0.0}"
But nothing has worked. The velocities are always to high (i.e. ~1e12).

Is there anyone who has experience with something similar?
tzwinger
Site Admin
Posts: 99
Joined: 24 Aug 2009, 12:20
Antispam: Yes

Re: Setting basal pressure til fraction of overburden

Post by tzwinger »

Hi Christian,
I guess you are using this 0.6*overburden as the effective normal pressure in some sliding law, right?
Did you check that your values are dimensional consistent with the rest of your material parameters? Simply, if you do your computations in MPa-a-m system (which usually is the case in glaciology) you should multiply your pressure values with a factor of 1.0E-06 to end up at MPa.

Best Regards,

Thomas
ChristianB
Posts: 10
Joined: 28 Apr 2014, 22:02
Antispam: Yes

Re: Setting basal pressure til fraction of overburden

Post by ChristianB »

Yes, I am trying the weertman sliding law.

I had the same though with units because my output is in MPa. But when I apply the External Pressure in MPa my basal pressures are on the order of ice overburden and not 0.6*overburden. So it seems that the pressure BC is not enforced in that case.

I am using the ParStokes with P as my pressure in the Pressure Preconditioning solver and Flow Solution 4 as pressure in my ParStokes solver. Should I set External Pressure on one of these variables in the BC?

My basal BC is,

Code: Select all

Boundary Condition 3
  Name = "bedrock"
  Height = Real 0.0
  Bottom Surface = Equals bedrockDEM

  External Pressure = Variable ZsDEM, bedrockDEM
    Real MATC "if (tx(0)>tx(1)+5.0) {(tx(0)-tx(1))*910.0*9.82*0.6*1e-6} else {0.0}"

  Normal-Tangential V = Logical True
  Flow Force BC = Logical True

  V 1 = Real 0.0
  Flow Solution 1 = Real 0.0

  Slip Coefficient 2 =  Variable Coordinate 1
    Real Procedure "ElmerIceUSF" "Sliding_Weertman"
  Slip Coefficient 3 =  Variable Coordinate 1
    Real Procedure "ElmerIceUSF" "Sliding_Weertman"

!!! Sliding only in areas with icethickness > 5 m
  Weertman Friction Coefficient = Variable ZsDEM, bedrockDEM
    Real MATC "depthwise(tx)"
  Weertman Exponent = Real 0.5
  Weertman Linear Velocity = Real 0.00001

End
gagliar
Posts: 79
Joined: 04 Sep 2009, 16:34
Location: LGGE - Grenoble
Contact:

Re: Setting basal pressure til fraction of overburden

Post by gagliar »

Hi Christian,

The Weertman friction law doesn't read the 'External Pressure' keyword. This friction law doesn't dependent on the water pressure. If you want to have the friction modulated by the water pressure you will have to use the Coulomb friction law. Have a look on the wiki: http://elmerice.elmerfem.org/wiki/doku. ... ns:coulomb.

You can also write a user function that will give the 'Weertman Friction Coefficient' as a function of the water pressure or effective pressure.

Note that in the Coulomb friction law, the overburden pressure is directly taken as the normal stress exerted by the ice on the bed (which is more correct than the cryospheric pressure approximation \rho g h)

I confirm the remark by Thomas: the water pressure passed through the 'External Pressure' keyword should be in stress units. So, if you are using the MPa - m - a units system, it should be MPa.

Hope it helps
Gag
ChristianB
Posts: 10
Joined: 28 Apr 2014, 22:02
Antispam: Yes

Re: Setting basal pressure til fraction of overburden

Post by ChristianB »

Hi Gag,

Thanks for the reply.
I have been playing with both weertman and coulomb friction which is the reason for why I would like to add a simple basal pressure. Although weertman does not depend on basal pressure it is quite simple compared to coulomb friction. I therefore started with setting pressure in weertman because then I know that the pressure works before moving on.

When I look at the basal pressures (i.e. Flow Var 4 from parstokes) they are very close to ice overburden which I find strange because the basal condition should be 0.6*ice overburden. But now I know pressure should be in MPa I will try the coulomb friction law and see if it will work with the external pressure keyword.

Thanks,
Christian
tzwinger
Site Admin
Posts: 99
Joined: 24 Aug 2009, 12:20
Antispam: Yes

Re: Setting basal pressure til fraction of overburden

Post by tzwinger »

Hi,
perhaps the name "External Pressure" is misleading. It is basically the normal component of the Cauchy stress vector, which is the Cauchy stress tensor dot product with surface normal. It follows from the weak formulation of the Stokes equation. It is a Neumann condition to the system. If you now at the same time prescribe a Dirichlet condition (zero normal velocity) at the same boundary, the Neumann condition will be overruled. In other words, in the hydrostatic limit your pressure at the bedrock always will be the ice overburden as long as you block the normal velocity at the bottom.

Regards,

Thomas
Post Reply