Review of W17 Hull Forms November 21, 2010 (updated Jan 2025)
Builder: Is there anything really very different in the Design Approach behind the W17 hull forms compared to other boats out there?
Designer Yes, I think it's fair to say there is … several things in fact.
In the first instance, I am trying to optimize the basically simple construction of a box with a flat bottom, into forms that will be efficient in their passage through the water—not at first glance, such as easy task. They can look so bad in sectional view, but remember, it's the elevation and plan view that will indicate FAR better how they will travel through the water! After all, a typical Stealth vehicle or plane, does not look so streamlined in section either. So let's go through what has been done for the W17 and why.
Speedwise, there's nothing so wrong with a boat that has a flat bottom. It's been shown in tests to actually be THE most efficient planing surface, and round-bilge hulls are a compromise away from that ideal in order to lower resistance at their lower speed or to improve sea-kindliness. So how can we use these flat surfaces to advantage?
Ama Design: First of all, the design of the amas are quite unique. While we need a flat surface aft to potentially slide or 'ski' on, this can be a source of pounding up forward, so we need a finer entry there. Also, when a multihull heels, the windward ama lifts out above the waves. No problem when it's a large boat but for a small boat of 16 ft (average hull length of the three W17 hulls), that ama is still not far up. So wave tops will repeatedly impact the ama-bottom and for that reason, the bottom of the W17 amas are warped as they go forward, to present a Vee shape to the waves coming up under them and eventually running out forward to a fine bow that does not pound but slices. Contrary to several other amas out there, I have chosen to incline the outer bottom face, rather than the inner. This shows two advantages. It not only gives a better Vee shape for the waves but it eases the outer edge from being sharp, so that when hit sideways by a strong squall, the boat can 'give a little' and not trip over the more vertical edge that occurs when you incline the bottom on the inner side.
A deep forefoot also adds to waterline length and compared to a 'banana hull' profile, keeps the center of buoyancy consistently forward as the boat advances through waves. As the boat heels and the ama is pressed in further, you want this volume to move even more forward to keep the bows up, so a slowly lifting keel line running aft with less and less buoyancy, will naturally help to achieve this. After all, for an ama nearly underwater, its hard to argue with nature and the shape evolution gave to fish ;) Fat transoms might be needed when the load is aft, but for an ama, their tendency to lift the stern on heeling is not a good one for a bow-up trim, and can also seriously affect steering control if the rudder is partially lifted out by an ama with too much buoyancy aft.
Another unique feature is that these amas are built with a small 'toe-in'. This helps both the climb to windward and also for turning. It's not much, but it's better than the toe-out at the waterline that many other trimarans carry—perhaps unwittingly! Just follow me here.
Most amas are set with some inward inclination to the vertical—10–12° is quite common. Their bows are also almost always higher than their sterns above the water. So if an ama is built to be parallel to the main hull (as most are) and that is measured at the deck level, then if you project down the inclined line of the stem and stern, the intersection with the waterline will be further from the main hull at the bow, than at the stern, due to the greater height at the bow. This means that at the waterline of the ama, there would effectively be some toe-out! So on all my tri designs, I build with some toe-in to compensate, and even add a small amount of genuine toe-in that I find helps as noted earlier. Not something typically found on other tris. This combines with the ama bottom, that has a small curl to windward like a carviong ski, to help push the ama upwind as well as making turning easier.
January 2025: One question raised was: "why are the amas shorter than the main hull if length permits a higher speed?"
There are several reasons, but first, length is not generally taken off the front end, as we need buoyancy up forward to add to diagonal stability, reduce pitching and lower pitch-poling risk. But what about the stern? First, if there is risk of an ama being pushed under (true on all small trimarans), a wide buoyant stern will cause very high drag when submerged, so the ama of a small trimaran (under say 10m) needs to be tapered down more like a fish body, so that flow can readily reconnect without excessive vortices & suction. If an ama is extended aft, its buoyancy will resist heeling which now lifts the stern of the main hull and if there is only a central rudder, steering control can be lost. [This happened on "Magic Hempel", when another designer decided to empower the boat with more aft ama buoyancy for a faster speed downwind. At 20 kts 'she now rebelled', as the stern lifted out 1/2 the spade rudder and steerage control disappeared. Fortunately, the boat still sailed straight but one had to then slow the boat to regain control ... hardly a good solution. It should be noted that with the more tapered original amas, the boat had once been timed at 25kts!]. Ending the amas somewhat forward of the boats' stern helps to avoid that major issue. Also, when the boat heels, we want the buoyancy to move forward with added heel as this has several benefits. It not only prevents the ama bow from dropping in deep and cause a pitch-pole, but it also increases the separation of the center of ama buoyancy from the main hull buoyancy, creating a more spread-out buoyancy-base that serves to dampen pitching .., which is one of the most annoying attributes of many boats that not only adds significant hull resistance but also destroys sail drive.
Some may now ask "well what about the shortened waterline ? " While we know that longer waterlines typically allow more speed, tests have shown there are limits. Once the L/B ratio exceeds about 17:1, the added skin friction will typically start to outweigh any gains due to lower wavemaking which at that ratio has now dropped quite low. Nature tends to bear this out as we do not see any fast fish with bodies slimmer than about a 14:1 ratio (if we justifiably exclude the long 'spear tusks' of marlins and sailfish etc). Tank tests also show that we really need an L/B ratio of about 12 to get a very low wave-making hull, giving us a practical L/B design range of 12-17. [Parallel sided hulls also lower wave making, by not pushing surface water out horizontally]. With small boats under 10m, too much space is given up in a main hull of L/B =12, so one has to compromise a bit. I try to not go below 10 though (which matches the W17 at the waterline). For the ama, it's much easier. While we need a decent beam for walking on, structural strength and buoyancy, I typically try to get an ama L/B of 13 or more at the deck with the ratio growing higher as the waterline narrows lower down. As there seems no advantage to exceed an L/B ratio of 17, there is just no need to make the ama longer at the stern, especially when the other factors above are factored in. Also, in tri's where ama rudders are justified (or even needed), there is a practical reason why you do not want those rudders to stick out aft of the main hull. So if you really want to make a hull longer for more speed ... add your length to the Main hull first .. it will almost always benefit from the added length. An excessively long ama is vulnerable to damage and even dynamic load failure in waves, as some earlier ocean racing trimarans found out when they broke off forward of the main beam. Even breaking side waves can do this if an ama is made too narrow with a resulting low beam strength.
So what is the end result re Ama to Main Hull Length ? I generally end up with a ratio around 90% but would accept to drop that to 85% for a swing-arm design, just to reduce the folded length. While one can regain a little length by pushing the ama bow forward of the main hull, I recommend keeping this offset below 3%L or the bow becomes too vulnerable when docking and the boat more difficult to turn. A motor tri, not subject to sail-load heeling, will typically have a much shorter ama of more like 50% as its buoyancy can be significantly less.
So now to the Main Hull. Even if it's basically of three flat panels, there were still options to choose from. They could be set to be quite Vee'd as for a dory and some cats are built like this. 
This gives more width (and hence stability) as it's heeled and also supports more load as it sinks—both ideal for a dory. But for the hull of a multihull, we just do not need that type of stability gain from heeling, as the amas are FAR more effective for that. What we need is low resistance to waves at the relatively higher speeds this boat is capable of. If we incline the sides, what happens to the passing waves? They are progressively pushed out as they rise up the sloped side and attempt to lift the hull. This is a lot of 'added work' on the boat and takes energy that adds resistance to the shape running through the water. Now if the sides were vertical, the waves would simply rise and fall, with minimal added work and resistance. So on the W17, the sides are kept as vertical as practical, with just enough slope to make sitting in the cockpit comfortable.
And what about that sharp lower corner of the main hull? Should we not round it off? I don't think so! While you'd have a fraction less surface friction, you'd lose an important element of this design. That square corner gives an important resistance to the transverse flow of water under the hull, that would allow side slip. The fact that a tri hull is so long and slim, allows that corner to work positively to help progress to windward and keep flow going more fore and aft than would occur on a wider monohull and numerous comparisons with other boats have indeed shown leeway on this W17 to be very low, contributing to an excellent performnce to windward for a small multihull, something quite incomparable to the many kayak-style trimarans that struggle upwind. In addition, that sharp corner appears to lower residuary resistance by keeping flowlines parallel.
(see Pics and Notes added below - 2024)
As far as the bottom is concerned, we ideally need this to work more as a ski and be as flat as practical so that as much of the surface can run with a very low angle to the flow. Having lots of rocker in the keel line would not help top speed at all. But this creates a conflict, as we need the keel below the waterline to provide buoyancy to support the boat weight. If it were indeed flat like a barge, then the transom would also end up deep under the water. At the bow, we can keep it fairly low, as it will be fine in width and as long as the boat still turns ok, this is a good way to assure the bow stays in the water, keeps the waterline as long as possible and starts to separate the water to either side as far forward as possible. (Think Sailfish ;). But clearly, designing with the transom underwater is not going to work. But for minimal rocker, we also want to raise it the least possible! So what is that amount? Well, not accepting to compromise top end speed more than necessary, has been my approach for the W17. It's clear that as boat speed picks up, this wonderful flat bottom surface can certainly provide some dynamic lift, so we can use this lift to help keep the rocker to a minimum. For the W17, I have tried to design the boat so that the transom is just afloat with no crew in the boat. "Yikes!" some might say, "what happens when you add the crew?" Well, a boat trims about its centre of flotation and with this hull, that is well aft of amidships, as the stern is much wider than the bow. If the crew are initially added forward of that 'pivot point', then most if not all the sinkage will be by the bow and not by the stern—so keeping the transom at roughly the same level.
Now, as the boat picks up speed, the flat stern sections start to become more supportive, just as a plank (with virtually no depth) that we call 'a ski', can support more and more weight as its speed is increased. We can now use this added dynamic support to carry some weight and slowly bring the crew more aft as speed picks up. This I believe will give the best overall performance for what is basically, a flat-bottom boat. If I had provided the depth under the aft end of the cockpit to initially support the crew with the transom still riding high, then I'd have automatically designed in an underwater 'bustle' that would have added more rocker to the keel line and inevitably limiting 'the skiing' that this boat will otherwise do. When a boat is designed for cruising at low speeds and sitting at anchor with a party on board, that 'bustle' might then just be necessary (see most other designs) … but if you then push that hull form to try and reach higher speeds, that adverse rocker will suck the bottom down, add to the trim and create too much resistance to get the best speed. This will be visible in the stern wake. The flatter the wake, the less energy being wasted. So constantly check your wake and if it starts to lift at all, move forward for less drag.
As in ALL designs, one has to choose ones compromise and I hope this helps to understand how I chose mine. The fact that the first W17 can run at 6–8 kt with the transom lifted to the waterline while carrying a significant excess of weight and then press on to reach over 14 kt in higher wind, shows the potential of what I am describing here.
For a boat built to the design weight, it should always be possible to run a clean transom.
But if you really WANT to ride a higher stern all the time AND can accept the slower top speed that will go with it, then the simple basic central-hull shape of the W17 could certainly be deepened to achieve that. Although this might be a required compromise for those much above average weight, it's just not the compromise I have personally chosen for the base design.
Finally, while the above explains some of the main design features of the two hulls (vaka and ama), the fact that they are both interconnected by beams (akas) has further important design effects. One is the effect of diagonal stability (see also this article) and another is pitch control. Both of these critical aspects can be improved by pushing the volume forward in the amas, being careful to still maintain a fine entrance of course. Clearly, more volume forward will add resistance to being pitchpoled, but even more important for daily sailing is its help to reduce pitching. The more fore and aft distance one can create between the center of buoyancy of the ama compared to that of the main hull, the more quickly pitching will be dampened out, and pitching not only adds to hull resistance, it also significantly affects the airflow past the sails very negatively.
Each design will have specific needs, but the above should help the reader understand this specific example.
Mike, November 2010 ... updated 2024/2025
In October 2010, I had an email from someone who'd been sailing the very first W17 and he reported this:
"I just had a really good sailor show me how to crank up speed and then turn on the back of a wave to let this boat really go. And GO it did. In fact I had the impression I was sliding on oil, it felt so frictionless by comparison to what I was used to."
[This intrepid sailor has since reported hitting 14.9kts on his GPS. This is a Speed/Length Ratio of 3.63, the design equivalent of a 50 footer doing 25.4 kts, or a 100 footer, doing 36.]
Then in Feb 2014, another sailor who had sailed the W17 wrote in to say this:
"Hi Mike. I was thinking about that sail I had with you last fall and recently told a friend, "I cannot remember ever sailing on a boat that felt 'just so damned efficient' ... it was really a very unique feeling".
After Professional Boatbuilding invited me (2017) to write an article to explain why the simple W17 hull shapes work so well (as reported by their sister publication WoodenBoat Magazine in their independent test), this article is now available on this website under 'Published Articles' .... "W17 Design Brief - Can Simple Shapes be justified by Science?"\
.
NOTE added May 2024: Although I have had this image somewhere on file for many years, I just relocated it and I think its worth sharing. It shows the result of a ship model test with flow lines from ink ports. On a 2nd copy, I have added the tangent lines of the bilge radius, hi-lighting how the flow crosses the radiius when the bilge radius is large but seems to run far more parallel when the bilge radius is sharp .. appearing to resist running around the sharp corner. It would be great if a tank test facility would do more tests on this phenomena, as this would indicate lower residuary resistance with the sharp chine corner ..... perhaps more than making up for the small amount of extra wetted surface. ie: flow taking the 'line of least resistance'. The square corner also offers fractionally more displacement low down which is also a positive thing. This clean fore-and-aft flow may help to explain the seemingly low forward drag of the W17, as well as explain why she has such little leeway. So those who think she would be better with a larger bilge radius, might want to reconsider why I have been resisting that ;) I might agree if this were not a sailboat that needs to go to windward well .... MW
.
Feel free to send in specific questions or comments via my Questions Form, that I may select to answer through this webpage if considered of broad interest.
"See the Copyright Information & Legal Disclaimer page for copyright info and use of ANY part of this text or article"