What would be the best way of incorporating the weight of water in a filled tank in a 129 transient solution? Non structural mass on the tank shell elements? 0D mass in centre of tank connected to tank with RBE3? Solid mesh with low stiffness? Or is there another option that I don't know of?
Non-Structural Mass is a reasonable option, but as you are dealing with dynamic analysis then the position of center gravity of mass is important, then the use of one spider of RBE3 elements is also correct. Create both models and compare solution results.
An interesting option exist with NX NASTRAN to consider the fluid-structure interaction: virtual fluid mass (MFLUID Method). I think the MFLUID method would be much more reliable than the RBE3 method.
A virtual fluid volume produces a mass matrix which represents the fluid coupled to a boundary consisting of structural elements and other effects, such as free surfaces, planes of symmetry, and infinite fluids. The incompressible fluid produces a mass matrix defined with full coupling between accelerations and pressures on the flexible structural interfaces.
Although free surfaces are allowed, no gravity effects are included directly. Because the fluid is represented by a coupled mass matrix attached directly to the structural points, this capability is allowed in all dynamic solution sequences. This capability may be used to model a wide variety of fluid-structure interaction problems. Some examples are fuel tanks, nuclear fluid containers, drilling platforms, and underwater devices.
A free surface is defined as an x-y plane in any local rectangular coordinate system. The user simply specifies a value of z as the upper limit of the fluid volume. Different fluid volumes may have different levels and orientation. It is not required that the surface coincide with the element properties since partially wetted elements are allowed.
The virtual mass fluid option may be used in all NX Nastran dynamics solutions, including the following special approaches:
The virtual mass method (MFLUID) is well suited for the following problem types:
Note that compressibility and surface gravity effects are neglected. It is assumed that the important frequency range for the structural modes is above the gravity sloshing frequencies and below the compressible acoustic frequencies. It is further assumed that the density within a volume is constant and no viscous (rotational flow) or aerodynamic (high velocity) effects are present. In other words, a ship traveling at a high rate through a thick oil patch would require some extra modeling effort.
The MFLUID method is supported since years in FEMAP, but I do not see direct support of this feature in the GUI of NX Advanced Simulation V9.0, so this could be an important limitation.
In FEMAP the Fluid Region command is very similar to the Connection Region command. The difference is that instead of creating regions for Contact purposes, this command creates individual segments representing incompressible fluid volume regions used for the purpose of generating a virtual mass matrix (MFLUID entry in Nastran input files). This capability is available in FEMAP supported Nastran Solution Sequences 103 (Modal Analysis), 107 through 112 (Complex Modal Analysis and Dynamic Analyses), 129 (Nonlinear Transient Analysis), and 200 (Optimization).
When a Fluid Region is present in your model, FEMAP provides an Output Request (Nastran only) called “Fluid Pressure” which will return an elemental “fluid-structure pressure” along with any other requested results.
Thank you for your reply Blas,
It's a pity that the MFLUID method is not available in the NX GUI. I have started experminting with options that are available and so far an option that is giving ok results is to simply use a solid mesh representing the water. I think this might be the best way to do it for tanks that are completely full. I have given the water isotropic material properties. A youngs modulus as low as possible without heavily distorting elements (0.01 MPa) and a high Poisson's Ratio (0.49). I am still to experiment with the influence of the structural damping coefficient in dynamic runs. It would be nice to hear is someone else has experimented with this approach before, what isotropic properties worked out best and what their conclusions are.
See the Simulation SIG "NX Open Routines" library on the PLM World web site:
Note that you have to be a PLM World "Citizen" and a member of the Simulation SIG for the above link to work.
Please read the note in my original post:
"Note that you have to be a PLM World "Citizen" and a member of the Simulation SIG for the above link to work"
Using the PLM World member search tool, I don't see you listed as a Citizen, let alone a member of the SIG.
Dear Lionel & Georg,
I am checking with PLM world if the NX Open app that Jim referenced is something I can provide here. I'm hoping it is, if so it and other tidbits will be available on the community site