The Black Swan Project
The black swan theory or theory of black swan events is a metaphor that describes an event that is a surprise (to the observer), has a major impact, and after the fact is often inappropriately rationalized with the benefit of hindsight.
I’ve been re-reading Design For Fast Sailing by Edmond Bruce and Harry Morss, AYRS 82, 1976.
Chapter XXII: Running Resistance Versus Speed of Sailing Multi-Hulls by Edmond Bruce, first published AYRS 45, 1963; AYRS 78, 1972. In which Edmond gives results from the AYRS home-made laminar flow towing tank. IMHO, Edmond Bruce is one of those unsung heroes who exemplify the best of British “garage band” creativity, other notable examples being Colin Chapman, Geoffrey de Havilland and Alan Turing.
British creativity thrives in near anonymity, and once exposed to the light of public and government attention appears to wither on the vine. I believe this is because amateur inventors are free to ask themselves their own questions, and not “what would be fastest under this or that rule”. Edmond was free of racing rules, and he asked a very basic question: given a fixed displacement, what is the best over all Length to Beam ratio for a sailing multihull?
He compared hulls of 3:1, 5:1, 8:1, 12:1 and 16:1. I won’t belabor his results, you can read them yourself, but Edmond was quite satisfied with the 8:1 hull as an ideal compromise - what I would call the the “Goldilocks ratio”. Neither to fat for speed, or too thin for load carrying, the 8:1 ratio hull holds great promise for multihulls, and especially proas. I’ve attached a drawing of the 8:1 hull in various sizes.
Perhaps this ratio will prove itself, with the benefit of hindsight. This is part of one of a series.
Very interesting!
I have been thinking a lot about the same problem. 8:1 makes it a lot easier to build a not to long proa for cruising.
My small Advanced Sharpie Proa is 10:1 and i think it sails very well.
There are several aspects of this though. Speed and a low wet surface is important, but behaviour in waves is just as important. It should not loose to much speed slaming through the waves, and it should not shake the wind out of the sails, disturbing the airflow.
Johannes.
8:1 has been considered the minimum for a long time for the center hull of a trimaran. Most load they can get and still have acceptable performance.
On a “3 sheet” proa that ends up with a waterline 20’-23’ depending on the hull form, that gives an inside berth width of 30”-36” without resorting to a pod, and without accounting for shrinkage due to skin thickness/framing. Way comfy 😉
Tom
Very interesting,
perhaps we now get away from this 16:1 fetishism found here all around. I allways looked at the resistance curves of Bruce and compared it to my sailing experiences onboard the Dragonfly 800 Trimaran.(9,5:1 mainhull, 15:1 amas)
Of course there is a small speed barrier at hull speed (around 7 knots), but a puff or a short sprint off to leeward mostly brought us over the hump ... speeding up. (Exept light wind (1-2 Bft).)
Every time when I sketch up a multihull I end up in the 9.8 - 11:1 range (mainhulls), for my feelings a good ratio between resistance, draft and load carrying capacity. But I’am always sketching cruiser racers being stingily with payload. For example my 30’ tri currently build has a payload of only 400kg, just enough for three crew.
For a cruiser, including the upcoming “three sheets long” proas here at the forum, a ratio under 9- or 10:1 would be better to get a more balanced boat.
Greetings from the North Sea Coast, Michel
Edmond was quite satisfied with the 8:1 hull as an ideal compromise
I gave the book to a friend 25 years ago, so I can’t look up the detailed procedure of the tow tests. I am guessing that Bruce didn’t tow from a mast. Assuming equal displacement and aiming for roughly equal speed, all boats with different length to beam, (or displacement to length) ratios would have the same rig. So if you were to compare two hulls of the same displacement, one with an 8:1 ratio, the other 16:1, then the 8:1 hull would have half the length. That means if you look at the ratio between hull length and rig height, the shorter hull has a proportionally twice as high rig, which might give you a longitudinal stability problem.
The designer of the Windrigger series of proas and cats studied longitudinal stability and concluded that making the hull longer was by far the best way to increase stability, better than increasing prismatic coefficient (I don’t remember whether he studied flare or raised bows). Just for the sake of longitudinal stability, I would want a long hull. I do remember my Hobie 14 having more transverse stability than in any other direction, especially diagonally over the stern. A short multihull could easily end up with the same problem. The stability curve of a multihull should not be plotted only for transverse stability, but as a 3D polar plot showing stability in all directions (180 degrees will do if you have a symmetry axis). Wave making is not the only reason for long hulls.
Regards
Robert Biegler
Let’s not forget that to make a 16:1 hull of a given length into a 8:1 hull requires vastly less than twice the material more.
This one bothered me for a bit, it seems counter intuitive but an open mind is called for 😉
The old Herbert Spencer quotation about “contempt prior to investigation” comes to mind again (it came up recently when finally playing around with greenland style kayak paddles).
For this I think I need to start with given displacement and flush it through Mitchlett a few times and see what comes up. My current AYRS publications don’t go back that far, but one thing that concerns me is the reference to laminar flow; we don’t really exist in that world…chaos rules. The other issue has already been mentioned, there is a lot of goodness to length when it comes to seaworthiness. Still being able to comfortably doodle something out at 8:1 sure opens up possibilities for trailer sailing.
Good topic
Skip
To get a rough feel for the difference that changing hull paramaters makes, you can use my hull resistance spreadsheet at:
http://www.users.on.net/~malcolmandjane/HullCalc/HullCalc.xls . I wrote the spreadsheet primarily for investigating these types of questions and to get a general feel for how a given hull might perform. The spreadsheet may not be not particularly accurate for a single design point, but it’s a good tool for looking at trends.
Statements like “8:1 is a good length to beam ratio for multihulls” can be misleading. For instance, 8:1, but at what displacement to length ratio?. In my opinion there aren’t many general rules for hull design. Each design needs to to be considered individually based on the specific requirements of the intended purpose.
Cheers,
Mal.
Statements like “8:1 is a good length to beam ratio for multihulls” can be misleading. For instance, 8:1, but at what displacement to length ratio?.
All the test hulls had the same displacement (110g). They all had half circle hull sections, and the rocker was also a simple arc. He rejected the “Taylor Standard Series” of ship models as being inappropriate for multihulls. Basic, but I think he made some good assumptions.
The designer of the Windrigger series of proas and cats studied longitudinal stability and concluded that making the hull longer was by far the best way to increase stability, better than increasing prismatic coefficient (I don’t remember whether he studied flare or raised bows). Just for the sake of longitudinal stability, I would want a long hull.
Agreed. Though in comparison to monohulls, 8:1 is “long”.
Intrigued by the idea that a somewhat “fatter” hull might be an ideal compromise, I used the 40’x5’ hull to explore hull sections and accommodations potential. While the extremely narrow hulls seem very slippery when considered on their own, the fact is, a cruiser has a certain amount of volume required, and if you squeeze the hulls so thin that they become virtually uninhabitable, then the accommodations spill out into pods, blisters and bridgedeck cabins, all of which take their own toll on weight, simplicity, efficiency and speed.
In other words, what if we treat the proa as if it were a monohull “canoe” with the ballast slung out to windward instead of underneath. Yes, this requires a fatter hull, but if, at the same time, you lose the pods or bridgedeck cabins, then perhaps the equation swings back in our favor.
Attached are a few sections that explore the 5’ waterline beam.
1. Simple Box: Easiest to build, but wow, that flat bottom looks like it will pound, and those seat backs look sort of upright and uncomfortable.
2. Sharpie: Ahhh, that’s more like it, as I lean back into the comfy hull sides. But still, what about pounding?
3. Shallow V: Comfy and smooth, the only thing I don’t like is the extra draft and sole required, also the hidden potential nastiness of a “bilge”.
4. Pretty much perfect. A flat bottom, wide enough to be the cabin sole and narrow enough not to pound. The extra chine is less trouble than the Shallow V w/ cabin sole.
Statements like “8:1 is a good length to beam ratio for multihulls” can be misleading. For instance, 8:1, but at what displacement to length ratio?.
All the test hulls had the same displacement (110g). They all had half circle hull sections, and the rocker was also a simple arc. He rejected the “Taylor Standard Series” of ship models as being inappropriate for multihulls. Basic, but I think he made some good assumptions.
Should that be 110kg? 110g gives a test model less than 0.5m long, which is quite small and probably not very accurate.
I used my spreadsheet to make some estimates based on a hull of 110kg, which is about 5m long for the 16:1 ratio. I looked at four cases:
1) The parent 16:1 hull with semi-circular hull sections
2) An 8:1 hull with the same draft as the parent hull (gives a waterline length 71% of the parent hull)
3) An 8:1 hull with semi circular sections (gives a waterline length 63% of the parent hull)
4) An 8:1 hull with the same waterline length as the parent hull (shallow draft)
As you can see from the chart below, according to the spreadsheet there is a big penalty for reducing the waterline length because the wavemaking resistance goes up dramatically. However, if you maintain the length, the only price you pay for making the hull wider is the extra wetted surface, which is not quite so dramatic and may be a reasonable tradeoff for greater internal volume. Note also that an 8:1 hull with only 71% of the 16:1 parent hull’s waterline length is only actually 41% wider, not twice as wide.
Hullspeed being a function of length - for the same hull size (wl beam ) - a longer hull wins.
The compromise is ergonomics - how you fit the human into the ideal hull form.
The polynesians didn’t have this problem - accommodations on the beams or deck and best hull (and bailers possible).
If not concerned with windage - i.e. a bigger hull to fit everything “needed” into it - then equal or lesser windage can be achieved by minimal hulls and a deckhouse.
Easier to build a simple hull/s without all the interior fit-outs and a basic deckhouse?
Truer to the original concept too…...
All true Alex, but up here in the PNW, land of 45 degree rainy days, we take our accommodation seriously 😊
Tom
Attached are a few sections that explore the 5’ waterline beam.
1. Simple Box: Easiest to build, but wow, that flat bottom looks like it will pound, and those seat backs look sort of upright and uncomfortable.
I must say a word or two about this. Yes sitting with the back against the hull-side will probably be a little bit more uncomfortable than a sharpie type of hull. A pillow or some type of bag containing soft stuff will make a very nice cushion to lean against and is removable when one wants to lay down to sleep. The bunk will be wider for a given with of the hull, or the aisle in the middle can be wider, whatever one needs/wants. There might be some pounding in a steep chop, but i have not seen anything that worries me, when i test-sailed my small model. I have been sailing it a lot. Probably somewhere around 15 - 20 hours or so, in very different conditions. I have seen some splashes, but nothing excessive, only what i consider normal and have seen from other hull-shapes (Jzerro, Madness and some monohulls).
I dont think one should be too worried about pounding from a box-shaped hull, as long as its heavy and slender enough. Very light and/or little rocker will pound more than heavy and deep rocker hulls.
Johannes.
Have you looked at des jours mellieurs videos on you tube - pounding into a short chop?
I guess it was this you meant?
I made a screenshot from a video on youtube. 31 seconds inte the video.
Johannes.