[rupole1185]

Is it correct this in the D file:

wall_name
{
type tractionDisplacement;
traction uniform ( 0 0 0 );
pressure uniform 600000;
value uniform (0 0 0);
}

Application of the pressure that act of a particular wall?

[bigphil]

Originally Posted by rupole1185

Is it correct this in the D file:

wall_name
{
type tractionDisplacement;
traction uniform ( 0 0 0 );
pressure uniform 600000;
value uniform (0 0 0);
}

Application of the pressure that act of a particular wall?

Yes, this will apply a surface pressure of 0.6 MPa to the wall.
The surface pressure is: Tn = -n & (n & sigma),
where 'n' is the surface unit normals and 'sigma' is the stress tensor.

Philip

[rupole1185]

Thanks a lot for your reply!

[tomdylan]

Originally Posted by bigphil

Yes, this will apply a surface pressure of 0.6 MPa to the wall.
The surface pressure is: Tn = -n & (n & sigma),
where 'n' is the surface unit normals and 'sigma' is the stress tensor.

Philip

Could you kindly explain the third parameter vector "value (0 0 0)"?

Does it have to do something with a pressure gradient?

My problem consists of pressure variations on a vertical (or inclined) wall in response to water level fluctuations. In that case the pressure has a hydrostatic vertical distribution. I am seeking the right boundary condition for such a case.

Any hints warmly appreciated!

Tom

[rupole1185]

Originally Posted by tomdylan

Could you kindly explain the third parameter vector "value (0 0 0)"?

Does it have to do something with a pressure gradient?

My problem consists of pressure variations on a vertical (or inclined) wall in response to water level fluctuations. In that case the pressure has a hydrostatic vertical distribution. I am seeking the right boundary condition for such a case.

Any hints warmly appreciated!

Tom

No the third term is just a common thing in openfoam. It's the variable value on the surface (I.e. the displacement is zero at first iteration)

[bigphil]

Originally Posted by rupole1185

No the third term is just a common thing in openfoam. It's the variable value on the surface (I.e. the displacement is zero at first iteration)

Hi Tom,

To add to the previous comment, I noticed from the other thread that you are interested in poro-elasticity; you should be careful to not to confuse the pore-pressure with the "pressure" in the solidTraction boundary condition, which is the normal component of the effective traction (or total traction depending on your particular implementation).

Philip

[tomdylan]

Dear Foamers,

first, Philip, let me thank you again for your clarification on the difference between pore-pressure and pressure. to be honest I was at first sight a little bit confused but after a while I got things clear.

Allow me to present some pictures to illustrate the analysis I am into right now. I want to apply a representative (in respect to the water level in the chamber) tractionBC on a vertical (or inclined) soil boundary patch.

At the present I am not considering the rigid gravity wall structure of
the ship lock, since I have not managed to deal with heterogeneity/multiregions.

As you can figure out form the pictures the filling and emptying of the chamber (more the 6 m head difference)

leads to a time varying hydrostatic pressure distribution along the horizontal sole and the chamber walls. At the bottom the traction BC is straight forward. However I am insecure as to how to represent a time varying traction BC at the vertical (or inclined) boundary patches resulting form a fluctuating hydrostatic pressure in the chamber.

Interestingly, our groundwater measurements in a distance from the lock detected a clear and almost in-phase poroelastic response to the filling and emptying of the chamber in a range of up to 10 kN/m2 (yes the ground can react quite elastic). These head fluctuations are solely induced by stress variations in the chamber.

I guess that I am not the first to deal with head fluctuations and their effect on solid traction and therefore suppose that a solution exists, but currently I have no clue for this kind of problem.

Ana hints warmly appreciated!

Tom