Fluid domain

[ibry]

Dear everyone,

Recently, I am modelling a gravity dam with FEAP and I found that there is fluid domain acting on the dam.
What the command should I use to make the input file of it?
The fluid domain is assumed to be linearly elastic and it also has data of bulk modulus of 2070 MPa and mass density of 1000 kg/m3 (water).
I am wondering whether it can be modelled as surface load or not, or any idea for implementing it?
Thank you very much.

Regards,
Ryo

[FEAP_Admin]

As a first step start with a boundary load on the dam and do not model the fluid; so just a surface traction.  You only need to model the fluid if you are looking for the coupled interaction.

[ibry]

Thanks for your answer.

So how about the data of bulk modulus and mass density of water?
Can you give me an example for that surface traction?
I am still confused about it.

Many thanks!

[ibry]

Dear admin or Prof. Taylor,

Based on the previous answer by admin, I have to use surface tractions for implementing fluid domain. Meanwhile, I have data of bulk modulus and water density which should be as input file.
I am still confused to apply it since command CSURface just requires coordinates and load (P).
I need an explanation and example of this case as well.

Thank you very much.

[Prof. R.L. Taylor]

It is difficult to understand what problem you are trying to solve.

The suggestion by "admin" was to model only the dam and apply the pressure from the water as a traction (normal pressure) load. This load would indeed depend on the density of the fluid and you would need to include it in the computation of the loading "P".

At present FEAP does not include elements based on fluid mechanics formulations. Trying to model the fluid as an elastic medium with a bulk modulus and density is not the best solution -- first, because the input to models is by E and nu which create both bulk and shear moduli.  You could write your own model but that would lead to the element having many modes of deformation related to the shear modulus that have no stiffness -- leading to a singular stiffness.

Thus, if you want to model fluid-structure problems you must code the fluid parts and its interface to solids yourself -- or couple feap with a fluids code.  There are some suggestions in the finite element text by Zienkiewicz, Taylor and Zhu (6th ed) that describe using a pressure model for the fluid.  Using this form you could use the thermal element in feap and only have to write the surface and interface parts as user modules.

[ibry]

Thanks for your answer, Prof. Taylor.

I am trying to apply water pressure as a surface load (traction).
As for the three dimensional case, how can I write the input file if the surface load is quadratic?
I am still wondering how many nodes should be applied by loads.
Maybe you could give an example in writing that input file.
Thank you.

[Prof. R.L. Taylor]

You should be able to use a CSURface loading in which the patch is a quadratic description so the pressure can change quadraticallyt on it.  FEAP should search for all the nodes that lie on this surface and apply the correct nodal forces.