Aluminium Boat Design

[saintA]

I Am Doing A Final Year Project In Boat Design At University And Use Solidworks For Model Drawing. I Want To Carry Out Fea And Apply Slamming Pressure To The Bottom Section, If I Restrain The Aft Section, It Will Act As A Cantilever Beam And Not Give Accurate Result.
Any Ideas Please On How To Test The Model Using Solidworks?

[Brandon]

The biggest questions in FEA sometimes do not have a totally straight forward answer. First I hope that you will perform at least some hand calculations to at the very least establish an order of magnitude for the types of numbers you might me getting for bending moment and for stress/strain, your Soliworks model will give you moments of interia for various cross-sections. I remember fondly some of my uni teachers saying with regards to computers in all of their glory : crap in-->crap out. Being that dynamics of a model at sea are far from predictable I would suggest trying to constrain the model about the CG as a worst case and applying forces near the bow and the stern to provide static equilibrium, you would be in essence treating your boat as a simply supported beam with the largest bending moment occuring perpendicular to the centerline of the boat, at the CG.
This is just an idea, but modeling is just that, an idea about how the model will experience what you foresee, play with it...maybe constrain the boat in various locations along the centerline and apply distributed forces that will provide you with static equilibrium and locate areas where failure might be a possibility and start changing the parameters of the model to drive it to failure and then look at how realistic your model was that caused failure...come from the front and then turn around and look at it from the back.

[saintA]

Hi Brandon,

Thanks so much for the suggestions, very well appreciated.

Rgds,

saintA

[ClarkT]

The way I've seen such a problem modeled in FEA is to decide upon the pressure values you will use for each plate element on your hull shell. This is the step requiring the most 'voodoo', you can guess, CFD, model test, or ask a medium, but you must start with this pressure distribution.

Next I would fix my FEA model against translation only, no nodal moment fixity. Your best bet would be to fix against heave along the entire gunwale, then fix against surge and sway with only two or so nodes. Now solve the FEA. From this result, you will know your net heave, surge and sway forces imparted by the slam, and you now must apply forces to counterbalance these forces. For heave, you could adjust the density of your aluminum until the gravatational force on your model equals the heave force. Similarly, a propulsor force and rudder or keel force will counteract the other slam forces.

It may take several passes, but this will give you a model without inordinate forces at the points of fixity, and a reasonable solution of the plate elements loading under the slam.

Good luck with it! Send us some pictures when you have it running!

[Brandon]

Good call Clark, you sound like you know better than me how exactly FEA on a marine vessel should work! Is it reasonable though to treat your model as a simply supported beam? I am picturing in my mind a small craft in med-short period seas traveling such that the vessel is on plane and with the waves passing beneath the vessel in a sinusoidal form with a period and wavelength which can be known (at least for modelings sake) and characterized with those parameters petaining to short-med. seas with occasional spikes in wave amplitude causing your vessel to "slam". At each point along the vessel that it contacts our sinusoidal waveform the vessel sits deep enough "in" each "peak" to provide the bouyant forces necessary to float the vessel...thinking on this i guess that one would need to know the speed of the boat and consider the planing characteristics as that would change the pressure magnitude and distribution along the bottom of the vessel ( admitting I am just beginning to dabble in all things marine I am just letting my thoughts flow here so bear with me)...unless the boat is a displacement hull...anyways, it seems to me that were the boat to fail catastrophically it would be at a location farthest away from the neutral axis ( and we can assume linear elastic material props when considering aluminum) so the location of failure would be along the top rail of the boat (the gunnel?) or along the deepest part of the vee in the hull. Ok, i have to stop rambling. I am curious what your opinion of this is. Thanks!!

[ClarkT]

Sounds like you are thinking about a mechanical event simulation (that's what Algor would call it). Virtually crashing the boat over the waves, solving the moment by moment stress values throughout the hull structure. As you spring off one wave, the compressed bottom panel would spring out and we would record the stress induced by that reaction.

There may be people in the world solving this sort of thing, I don't know them. I imagine solving a 30 second interval to the level of detail described above would need Los Alamos' top computers for a couple of days, if you've written the code.

I think most engineers would select several 'worst case' instances of slamming, based upon their experience and intuition. They would then set up a restraint condition as I described above for each of those 'worst case' load conditions, then let the computer solve the FEA.