Capsize Ratio's

[barnakiel]

Originally Posted by TeddyDiver

and twice the structure to withstand it.. No seriously, both hulls would disintegrate without a doupt (there's no thumb down smilie). However this totally irrelevant subject or do you know such pancake occasions?

Sabray, this haven't been about capsize ratio for looong time

Seen a Lord Nelson slide from the truck and then another boat of similar construction toppled in the boatyard. There was some damage and both boats got fixed.

Seen a B****** beached on a sandy shore then pounded by the oncoming waves. Not much left over in the morning.

This much for lighter/stronger in respect of how a mass market boat is built today.

b.

[kefroeschner]

Yes, way off 'Capsize Ratios,' which, I think we all agree, are relatively uselessly simplistic.

On the wave vs. wind causing capsizes -- seems to me that where modern designs get into trouble (Fastnet et al) is that the staic stability curve peaks at about 60 degrees and drops rapidly as 90 degrees is approached. So, a gust of wind heels the boat down to spreaders in the water: in an old classic design, with the point of maximum righting moment at around 90 degrees, this is not a problem -- the wind force has been spilled and the righting moment is maximum -- she stands up. However in the newer designs, the righting moment at this sort of knockdown, while still somewhat positive, is pretty damned small and this is where the wave comes in. How big of a wave does it take to overcome that last bit of the stability curve and turn the thing entirely over? Not much. Look at the curves.

To look at it another way any boat can be rolled by a big enough wave. The size of the wave required to do this is vastly smaller if you are at 70 degrees and only have about 20 percent of your righting moment integral left.

And yes roll moment of inertia is important and they are static curves, but the dynamics are not that dominant. The roll moment of inertia tells you more or less how fast things will happen, not whether they will or not. Or to look at it another way, the moments of inertia (in any axis) of a floating vessel are generally far smaller than that of the mass of water immediately surrounding, so the inertial effect of the boat trying to resist what the water wants to do to it is usually futile.

The impact discussion is right on, glad to see you guys thinking on these lines. However it's way more complicated than just weight or strength or strength to weight ratio. It is essentially a ballistic impact problem. I did the calculations for a trimaran I designed long ago -- the essence of it is that you have to consider the hull panel by panel, between bulkhead, and etc. then use Timoshenko's good book to estimate the bending loads on the panels, when subject to a uniform load over the area, ie.e. the stagnation pressure of seawater at whatever impact velocity, sy 20 knots. If the hull were simply a thin unsupported shell, like an egg, the only option for analysis would be finite element methods. Do-able, but not by amateurs.

I should be able to do a simple model -- pressure vessel for example, same problem but inside out. The stress, S, in the wall of a spherioidal vessel is Pr/2t. Tensile in a pressure vessel, compressive in our simple model of a hull being compressed by stagnation pressure. We can re-arrange this to find what pressure can be sustained, P = 2tS/r. The weight of the vessel, m, is proportional to t, to r^2, and to the density of the hull material, d. ( I'll get back to this later, the mate has dinner on.)

The model proposed by ________, in which we consider the stresses which the middle of the hull structure must transmit to the far side in order to make it match the motion of the side which has just slammed into the wave, thus not 'pancake.' A complex calculation of bending stresses in the hull skin if it is essentially an egg, otherwise, the bulkheads do the work, purely in compression. Most boats work on the bulkhead, not the egg-shell principle, to get the stresses from one side to the other, so unless the bulkheads fail, it is not an issue. And ...

We are back to the stresses in an edge-supported panel.

On that light, The Gudegeon brothers presented test data on stiffness measurements of various structures. Others as well. The surprising fact is that fiberglas is not all that good as a stiff panel. Wood is actually better -- although the uncertainty issue is there as ____ pointed out. The very best was wood coated both sides with f'glas. Aluminum is terrible. These data can be seen in actual boats -- aluminum hulls with depressions between bulkheads, and the fact that some of the fastest, lightest and strongest hulls ever made were multiple diagonal wood veneer with f'glas over. Too expensive for most, and you can get almost as good with other cores like balsa or foam. And now graphite is even better than glas, but the principles remain.

Great discussion Thanks for all the thoughts. Almost as much fun as sailing. Or painting and varnishing.

[Jim Cate]

Originally Posted by kefroeschner

We are back to the stresses in an edge-supported panel.

On that light, The Gudegeon brothers presented test data on stiffness measurements of various structures. Others as well. The surprising fact is that fiberglas is not all that good as a stiff panel. Wood is actually better -- although the uncertainty issue is there as ____ pointed out. The very best was wood coated both sides with f'glas. Aluminum is terrible. These data can be seen in actual boats -- aluminum hulls with depressions between bulkheads, and the fact that some of the fastest, lightest and strongest hulls ever made were multiple diagonal wood veneer with f'glas over. Too expensive for most, and you can get almost as good with other cores like balsa or foam. And now graphite is even better than glas, but the principles remain.

Great discussion Thanks for all the thoughts. Almost as much fun as sailing. Or painting and varnishing.

Here is some anecdotal and subjective support for the above comments:

We have owned and cruised in three types of hull: solid frp, very heavy Airex cored frp, and strip plank Western Red Cedar w/ frp inside and out.

The latter construction is subjectively much stiffer. There are far less distortions observed under sailing loads, ie, doors/drawers continue to operate normally, there are no creaks or groans from hull movements, the cabin sole doesn't squeek, the rig stays tighter, no stress cracks in paint or the frp sheathing near point load areas, no oilcanning. Good stuff!

I was initially concerned that this rather light hull would be like an eggshell, but have come to the conclusion (after 9 years and 42000 miles at sea) that it is bloody strong! Still don't want to test it by free-falling 20 feet off of a wave front, though...

Cheers,

Jim

[barnakiel]

It does not matter if the max moment is at 60, 90 or elsewhere. It is the moment that counts, not where its peak falls.

If we want to think about the max/min values of the moment, I'd rather concentrate on where the zero moment is as this dictates how the boat will behave once the roller passed by as well as what she will do if we get slammed at the peak and end up flat on the slope - which is how plenty of capsize accidents happen.

Plenty out of context pieces of information in this thread. I bet at least one of posters is not an engineer.

Smiley!

b.

[cwyckham]

Originally Posted by kefroeschner

I should be able to do a simple model -- pressure vessel for example, same problem but inside out. The stress, S, in the wall of a spherioidal vessel is Pr/2t. Tensile in a pressure vessel, compressive in our simple model of a hull being compressed by stagnation pressure. We can re-arrange this to find what pressure can be sustained, P = 2tS/r. The weight of the vessel, m, is proportional to t, to r^2, and to the density of the hull material, d. ( I'll get back to this later, the mate has dinner on.)

.

Thanks for your great thoughts, you obviously know of what you speak. Interesting about the wood sheathed in fiberglass. I think they call that "cold molded?" Supposed to make great boats. Jim's sounds like a dream.

I don't think the inside out pressure vessel will work, because structures in compression will fail by buckling much sooner than by any other means. I think your panel method is the only way to go (or finite element analysis).

[cwyckham]

Originally Posted by TeddyDiver

and twice the structure to withstand it.. No seriously, both hulls would disintegrate without a doupt (there's no thumb down smilie). However this totally irrelevant subject or do you know such pancake occasions?

Sabray, this haven't been about capsize ratio for looong time

If there is "twice the structure to withstand it", then you are saying it' twice as strong, so the strength to weight ratio is the same and the boats will behave the same. If the weight of one is twice the other but the strength is the same (material, design, workmanship differences), but the shape is identical, then the heavier boat with the same strength will break earlier in such a drop.

Garrett again: "That this scenario is basically correct has been verified by numerous reports and sailors have often spoken of a roaring noise just before capsize. K. Adlard Coles in his classic book Heavy Weather Sailing says: 'When a yacht suffers damage in a gale it is usually because it is struck by a sea and literally thrown down in the trough so the doghouse, or coachroof, is stove in on the lee side.'"

[kefroeschner]

cywyckham:

I agree that the pressure vessel model is not really appropriate and that buckling or some other mode of failure in an edge supported panel is the right model.

But just as a simple model note that the pressure supported by a shell in external compression is completely independent of the weight (density) of that shell and depends only on thickness to size (t/r) ratio as well as the strength (S) of the material. Pretty much the same thing comes out of the edge supported panel analysis -- no dependence on density or weight.

The arguement that the loads are proportional to the weight of the hull (or it's momentum) I find unconvincing. If it is a wave slamming into a stationary hull, then it's only the water velocity that matters -- stagnation pressure = 1/2 rho v^2. Same if it's the hull falling off a crest and hitting that trough below. (Tower of Pisa experiment) You could argue that the duration of the impulsive load on the hull panels will be proportional to the momentum of the hull and I would agree. But I don't think it is an important effect -- if the panel is going to fail it will do so the instant the yield stress is exceeded, and if that stress is not exceeded it will hold for a long time. Let us not get into crack propagation, creep and all the rest.

Your quote from Coles is a propos, as is a look at all the wreckage from Katrina et al. The failures in the hulls are usually at the panel edges, i.e the bulkheads, deck beams or other supporting points.

[sneuman]

Originally Posted by Bash

Round the buoys? Even beer-can racers don't race "round the beer cans."

It's a nice image all the same.

[sneuman]

Originally Posted by alba

constant pounding does damage the structure and if this happens again I will try and head a bit more off the swell to avoid the slamming you get.

Glad you made it through in one piece, but I would have to disagree that structural damage is inevitable or even likely in these conditions. What kind of a boat?

I have been in similar (for a similar amount of time) and my boat was dismasted but never sustained structural damage to the hull. I don't think it's likely at all in a well designed and built boat.

[TeddyDiver]

Originally Posted by cwyckham

If the weight of one is twice the other but the strength is the same (material, design, workmanship differences), but the shape is identical, then the heavier boat with the same strength will break earlier in such a drop.

But there are no two boats such comparison could be possible. As I said earlier boats are build up to meet some rule or standard. These have a bad habit (from your point of view) of taking into account also the displacement/weight of the boat so that the heavier boat must have stronger members.
What we could look into more is the dynamic stability in breaking seas
BR Teddy

[Polux]

Originally Posted by s/v Jedi

Exactly. Sundeers have all that: very light and very strong. But there are only a few production builders who do the dry layup / infusion process correctly to achieve the goal. TPI Rhode Island is one of them, using the SCRIMP process.

ps, I was reading back a bit... wind can't capsize, waves do. This is true for monohulls but not multihulls, which can theoretically capsize from wind force alone.

cheers,
Nick.

Humm!!!

Wauquiez, X yachts, GrandSoleil, Comar, Luffe yachts, J boats, Dehler, Pogo and even Jeanneau use vacuum infusion and are just some of the production manufacturers that use it. Most performance cruisers are built that way.

[cwyckham]

Originally Posted by kefroeschner

cywyckham:

I agree that the pressure vessel model is not really appropriate and that buckling or some other mode of failure in an edge supported panel is the right model.

But just as a simple model note that the pressure supported by a shell in external compression is completely independent of the weight (density) of that shell and depends only on thickness to size (t/r) ratio as well as the strength (S) of the material. Pretty much the same thing comes out of the edge supported panel analysis -- no dependence on density or weight.

The arguement that the loads are proportional to the weight of the hull (or it's momentum) I find unconvincing. If it is a wave slamming into a stationary hull, then it's only the water velocity that matters -- stagnation pressure = 1/2 rho v^2. Same if it's the hull falling off a crest and hitting that trough below. (Tower of Pisa experiment) You could argue that the duration of the impulsive load on the hull panels will be proportional to the momentum of the hull and I would agree. But I don't think it is an important effect -- if the panel is going to fail it will do so the instant the yield stress is exceeded, and if that stress is not exceeded it will hold for a long time. Let us not get into crack propagation, creep and all the rest.

Your quote from Coles is a propos, as is a look at all the wreckage from Katrina et al. The failures in the hulls are usually at the panel edges, i.e the bulkheads, deck beams or other supporting points.

Agree that weight has nothing to do with strength (not directly, anyways). I've been a bit puzzled myself over the metod of calculating forces in a drop. Easy to do onto a solid. Not so sure about a liquid. Ross Garrett is a physics professor, and every item in his book that I have specific knowledge of has been exactly correct. I'm just not sure I can wrap my head around why in this case. He has an appendix for each chapter with more calculations, so I'll look into that.

[s/v Jedi]

Originally Posted by Polux

Humm!!!

Wauquiez, X yachts, GrandSoleil, Comar, Luffe yachts, J boats, Dehler, Pogo and even Jeanneau use vacuum infusion and are just some of the production manufacturers that use it. Most performance cruisers are built that way.

I believe many of those, if not all nowadays, do. That is part of it, making sure the implementation of the laminate is good. But it doesn't say anything about how strong it is yet.. you have to examine the layers of glass or other fibers for that, the ribs, stringers, core etc. Even the resin; for example: TPI used vinylester instead of polyester resin. I know they built the Lagoons like that, while I'm not sure how they were built in La Rochelle.

It all comes down to the fact that everything needs to be done right to get a strong yet light boat. You also want a stiff boat for performance.

cheers,
Nick.

[TeddyDiver]

Originally Posted by cwyckham

I've been a bit puzzled myself over the metod of calculating forces in a drop. Easy to do onto a solid. Not so sure about a liquid.

You are not the only one. Me thinks nobody has the ultimate truth of wave induced capsizes and the forces involved. Not that it could be possible in a simplified scenario in a tow tank, but there's not two similar real world cases at sea.
Remember also that the crest of the wave is also dropping with the boat so it's actually btw two masses of water crashing together.
BR Teddy

[cwyckham]

Well, there's nothing more in Garrett's book on calculating the impact force, so I went looking for Jordan's stuff. I forgot how great it is. Especially the Coast Guard report should be required reading. Find it all here: Jordan Series Drogue

In his calculation on the Winston Churchill (on the site) he says that the impact was equivalent to be dropped from 70' onto green water and the force is 200,000 lbs. Unfortunately, he didn't say how he calculated that.

Interestingly, in another part he points out that the impact from the actual breaking wave part onto a boat isn't nearly as bad because the water is quite aerated.