Constrain a SHIP FEA

[TheMan008]

Question:

Full Global model of a ship. How can a ship be properly constrained in finite element analysis software. A fixed constraint at any node does not properly simulate hydrostatic buoyancy forces.

How can buoyancy forces be properly modeled in FEA?

 

[rb1957]

apply balanced loads and reactions; then constrain the 6 rigid body motions. one way to determine the correct balancingreaction is to estimate it, then examine the rigid body reactions (which should be zero) and see the loads you need to apply to achieve the correct balance.

water pressure is proportional to depth.

 

[johnhors]

as rb1957 says apply a balanced set of forces and moments, then apply 3-2-1 minimal supports to remove the rigid body motions.

as discussed before in ET :-

thread727-221271

thread727-215433

http://www.Roshaz.com

 

[TheMan008]

Yeah, constrain 6 degrees of freedom for a 3dimension model. Thats FE 101 for fixed structures. However a ship isn't fixed, its floating. So how does one constrain something that moves, like a ball?

Question: Where and how would you apply constraints to a ships hull?

 

[amorrison]

The same problem applies to aircraft design.

See thread404-188619

 

[johnhors]

TheMan008, amorrison please read the attachment that describes 3-2-1 minimal supports. A prerequisite is that the sum of all applied loads and moments is zero.


www.Roshaz.com

 

[corus]

I would consider the ship subject to a hydrostatic pressure loading which varies in depth with the weight of the ship. In order to balance the loads you'd need to calculate the linearly varying (with depth) pressure distribution around the hull so that the vertical component balances the weight of the ship. I'm not sure how easy that is to do as you'd be integrating around a curved surface, but presumably shipwrights have been doing that for years. Once you have that pressure distribution and apply the self weight of the ship to oppose that load then you should be able to restrain the ship, as others suggest, such that the reaction force at the restraint will be zero. Try and use symmetry restraints where ever possible to take a lot of the hassle out.

Tata

 

[rb1957]

pls read the posts ...
apply balanced load and reactions. this means you have to figure out where the waterline is before you try to determine the internal loads.

or talk to your prof.

 

[johnhors]

Corus - there are pre-processors that will correctly summate varying pressure distributions regardless of the complexity involved.

http://www.Roshaz.com

 

[crisb]

Am I missing something? Is there any need to calculate the sum of buoyancy pressure distributions, which will be zero laterally, and equal to the weight of the ship vertically.

 

[johnhors]

Cris, if I understand it correctly the sum of the buoyancy pressure distributions must be equal and opposite to the gravitational loading. Applying the gavity load is simple and the pre-processor should tell you what it is. You then need to define the correct distributed varying buoyancy pressure to achieve a load balance, that as I see it, is the interesting bit.

http://www.Roshaz.com

 

[crisb]

Archimedes' principle?

 

[rb1957]

it's easy to figure out a static total load (ie = the weight) and so the volume which will lead to the waterline (not a stunningly simple sum, but also not stunningly difficult). the element pressures are somewhat harder to derive but still a little less than rocket science; remember the surface is quite complex. i expect that there is a preprocessor that would spit out the three components of the element normal, the co-ordinates of the element centroid.

it would be somewhat harder to solve for a pitching, heaving, yawing, etc loadcase; but again i'm sure there's a preprocessor for nav archs.

 

[GregLocock]

yes, but bear in mind the deflected shape of the hull will affect the displacement.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?