Discussion Forum

Hi

my first thoughts would go in the direction mix a fluidic simulation with a solid one, apply a gravity body load on the fluid (and on the solid part). Then link the two by a boundary pressure load.

one could also imagine only a solid simulation and apply the pressures as equations to model the surrounding "perfect" fluid. This is probably easier to get to converge

--
Good luck
Ivar

Thanks for your reply

Is the first choice FLUID-STRUCTURE INTERACTION model?!

Hi
The FSi examples I know about are set-up for a flowing liquide, but why not for a static case, it's at least a good example to use to copy for the pressure boundary load distribution. I would start like that, and select a cae sufficient simple that one could compare with simple analytical equations.

But I would also find it worthy to try too to use only equations for the pressure and a structural case, and compare the two models

--
Good luck
Ivar

Ivar,

This is my model using FSI. but it doen't converge. I think there is something wrong with it. Would you please take a look and give me some hints.

Thanks for your attention

Hi

Indeed FSI is jmixing many things, so one need to start in steps. You have made a good start but there are a few things to consider. Indeed you need to apply a body load to the structural part and a volume force to the fluid, both can be writen out "-g_const*fsi.rho" in N/m^3 this is to be seen as the elementary load applied to a single element of size dx*dy*dz, and that is integrated over the domain to get total forces in [N], not to forget also summed over the selected domains (in your case 1 per item, but there could be several). Which means that the "fsi.rho" could be domain dependetn and should be read "fsi.rhi(i)" where i run ove the selected domains. this is part of the "implicit" notaton of COMSOL important to understand, but mainly "hidde" in the doc.

Then as your aluminium shell is light (hollow I assume) it will go straight up and skeeze rapidly the mesh, So I would start wit a density closer to the one of water to start "slow.

Then your initial conditions, you should add a pressure field of the type -g_const*fsi.rho*(Y-1.25[m])/(2.5[m])*1[m]
as initial conditions then I believe the solver would be happeir ;)

I'm still not fully convinced that everything is in the FSI case, I do not seethe pressure loading, as the surface load fluid-structure is based on fluid displacement and velocity. I have to undertand where the pressure is coming in, as the fluid pressure is giving the boyance force, while the fluid viscuosity is retaining its motion

I still have some issues with your 4.0 model, mostly the loading of a 4.0 version into 4.1 is not ideal, so I must build it from scratch
--
Good luck
Ivar

Then your initial conditions, you should add a pressure field of the type -g_const*fsi.rho*(Y-1.25[m])/(2.5[m])*1[m]
as initial conditions then I believe the solver would be happeir ;)

Really thanks Ivar :)

Where should I Add this pressure field? Since I have Added Volume force and Body force to fluid and solid, is it mandatory?

Regards

Hi

well that is exactly it. I had expected that the "p" was on by default for the continuity along the boundary, I'm looking into it, as I need to find it out myself ;).

By the way you had also almost incompressible material turned on for the aluminium, that I do not believe is requred, unchecking it will removes the structural pw variable

--
Good luck
Ivar

But I have added PRESSURE POINT CONSTRAINT in top of the model. It seems enough. isn't it?

Hi

the pressure point is indeed useful to fix a "gauge pressure" value so that is, for me, correct indeed.
But the pressure from the fluid volume load must also be linked to the solid. That I expected to be part of the fluid-structure interaction of the boundary condition, but I have some trouble to identify it in the internal COMSOl equations ;)

--
Good luck
Ivar

is there any progress about pressure ? :)

I haven't find any thing yet :(

Hi

indeed I havent managed to look closer to your issue (I'm a user of COMSOl with a job to do too, and these days I'm sharing my licence with a colleague working on his PhD, s prioritiesare 1) my bosses work, 2) my colleagues work, 3) the Forum, sorry for that. Bit its intersting, the boyance force, such a "simple" problem, how quickly it get complex if you want to implement it in COMSOL

--
Good luck
Ivar

I found a webinar that talks about how to implement buoyant forces it is at R&D magaizine with David Kahn and Rita martin

event.on24.com/clients/default/presentation/default.html?titlecolor=000000&eventid=279197&sessionid=1&username=&partnerref=&format=wmaudio&mobile=false&flashsupportedmobiledevice=false&helpcenter=false&key=A1137B547075F8D75C99F19059C74E7D&text_language_id=en&playerwidth=1000&playerheight=650&silverlight=true&eventuserid=46161531&contenttype=A&mediametricsessionid=44278982&mediametricid=671310&usercd=46161531&mode=launch

Hi Ivar,

I am new to COMSOL, could you explain this more in detail

one could also imagine only a solid simulation and apply the pressures as equations to model the surrounding "perfect" fluid. This is probably easier to get to converge

--
Good luck
Ivar

I have an iceberg in water and it melts, thus draft is changed with the time. What would be the most easiest way to model buoyancy and water level?

Hello;

Is there any progress about finding pressure load in comsol fsi? I am also in need of buoyancy
Please can somebody help me with solving this problem?