You are here
Boat simulation-how to define the buoyancy? Can it be done in ABAQUS
Sun, 2009-02-22 12:50 - xjwang
I am trying to model a boat sitting in the water. I want to see the stress distribution after the boat is put in the water. If I only consider the solid boat model, I need define the weight of the boat and the buoyance otherwise the boat would not be in a equilibrium. My question is that how can I define the buoyance in ABAQUS, or I have to find an alternative to define the floating force? Since the water pressure is function of depth, I think we might be able to define the pressure at different level. But I don't know if it will works fine.Any comment will be appreciated.Thanks a lot.
Forums:
First of all, I have seen
First of all, I have seen that there is an Abaqus version called
Abaqus Aqua, that does buoyancy and wave loading simulations. I have
not used this myself, but it looks nice ( I do some work with marine
structures).
Perhaps you don't have access to Abaqus Aqua so I
will mention that I had a similar situation in the past, and I didn't
have access to Abaqus but the method could be applied with Abaqus as
well I would imagine. Anyway, I used a homogeneous mesh and applied
spring elements at the bottom nodes. The spring stiffness in this case
would just be (gamma x A) with gamma being the water density and A the
tributary area of the node in question (hence the need for a
homogeneous mesh, though I suppose you could write some routine to
compute and write the stiffness of each spring from the tributary area
of its boat node). Thus the force in each spring would correspond to
the buoyancy acting on the area tributary to the node. This is a rather
crude method and severely limited, but it works. With an explicit code,
you could even impose sinusoidal displacements at the bottom of the
springs to simulate the vertical component of wave action.
Thanks a lot. I will try
Thanks a lot. I will try the way. If I don't have homogenous mesh, so the stiffness of springs will be a problem. Maybe we could just fix the nodes at bottom where the prings will be applied, then we apply the weight of boat to get the reactions at the fixed nodes. The reactions might be used to inverse the stiffness of springs. I am not sure that is the way to define the stiffness of springs.
why not just apply variable pressure?
First of all hello,
I don't have any experience with bounyacy in FEM, but this post caught my eye and I decided to check it out. While reading gpekmezi's reply, I couldn't help but think that his solution, despite might work, sounded a bit too complicated. However I don't have any experience with Abacus (I've only worked with ANSYS) and I don'tr mean to criticise the sugestion that was made. I only wanted to know if simply applying pressure along the hull (varying of course along it's height) wouldn't do the trick. Remenber that pressure at a certain point in a fluid is the same in any direction, dpending only on the depth.
Thanks
Hello Filipe, What you
Hello Filipe,
What you suggest is perfectly reasonable and I should have mentioned it in my post. If you have a flat or triangular hull the distribution is fairly straight-forward. For a contoured hull (e.g. NURBS) it is not quite as simple, since there might not be a mathematical function to define the surface area with respect to hull height (for balance of forces) or the volume with respect to hul height (for Archimedes' principle). What I would do would be to find the total buoyancy at a number of floating depths. The total buoyancy would depend on the horizontal projection of the submerged surface area and the center of the vertical projection of the submerged surface area. With a good 3D CAD program you could query both for different submerged depths. Alternatively you could query the volume and use Archimedes' principle. Match a buoyancy number with the weight of the boat and then you can apply pressure as a function of the depth below water. Once you start dealing with mooring, berthing, and unbalanced cargo forces in addition to a contoured hull, the buoyancy pressure gets to be a messy deal.
I should mention that there is a problem with my spring approach as described above. Tying the spring elements to the full bottom of the hull would give you erroneous results, since the buoyancy is zero for hull points above the water. In this case you'd still need to find the floating depth through the matching approach in the previous paragraph and apply springs only to those nodes below the waterline. This is still rather cumbersome.
There are packages available for buoyancy calcs, and there may be free or trial versions. The only strictly Abaqus approach I can think of, would be to model the water level with 2D contact elements supported on springs with stiffness of (gamma*A) and lower the boat onto these elements. This would likely result in long computation times.
Thanks for all of your
Thanks for all of your help.
If I apply the variable pressure along the depth, how do we get the pressure number? gpekmezi mentioned that good 3D cad program can do that, what is that software? Or we could get another way to find out the buoyancy force which is balanced by the weight? By the way, how do you guys think abou the boundary conditions? If I apply variable pressure as a function of the depth below water, then where and how should I add boundary condition? Since we know the buoyancy force will balance the weight of boat, can we just calculate the applied pressure distribution as a function of depth below water using the weight of boat if we know the area below water? Let Ai(di) is the area at di depth, Ph is the pressure at the bottom of boat, h is the total depth below water, pi is the pressure at di depth. Then we have sum(Ai*Ph*di/h)= W, then we could have Ph = W*h/sum(Ai*di), then Pi can be caluated as Pi = Wdi/sum(Ai*di).
Modeling buoynacy
I am also trying to model hydrostatic force in providing buoyancy to a house for lifting. I did by contributory area method and secondly by aplying uniform area load in kn/m2 with additional boundary conditions by restraiing 4 joints. There is not much difference observed in the results. also I wanted to clarify whether there is any impact of bulk modulus of water in calculating spring constant for buoyancy. Also stiffness of structure will have effect on spring constant. how do we take care of it?
I love making models of
I love making models of boats. If you are able to do these, then you are a very artistic person. I sometimes get impatient and then end up not finishing. The other day I ended up doing a huge Boat Donation because I am unable to do these models anymore. If you are interested in some let me know.