Boat Design [33]

By John Winters

Any one with a good memory may remember Norm’s article of Sea Kayak design (or was it an add for an Inuit) in January 1997. Norm’s article ended up in the hands of John Winters – a leading North American sea kayak designer. He has been conversing with Norm via the internet and questions some of Norm ideas. At Norms request he has written an article about design and in partictular Norms design philosophy. A word of warning, before reading this article read the rest of the magazine first, go for a paddle, grab a beer then sit down and read it; its rather long. Ed

I can still recall when a dearly held opinion of mine on how sails worked was crushed under the weight of a more accurate explanation. It surprised and perplexed me that I could have sailed successfully for so long without a proper understanding of aerodynamcis. Of course, like many things, it was simply a matter of doing the right things for the wrong reasons thus proving that you can be wrong but still be right – more or less.

Sea kayakers and canoeists also fall prey to erroneous notions of how things work. The boats are simple and aboriginal people made successful boats without knowing anything about hydrodynamics so what could be complicated about them?

The reality is that it is rather difficult not to make a good boat if it is light enough, narrow enough, and pointed at both ends.

That, however, doesn’t mean that the principles of fluid dynamics are simple. It was once said that “one need not understand the principles of digestion to enjoy one’s meal” and that also applies to paddling.

We manage to paddle along successfully even if we believe in magic or its equivalent. So long as what we believe doesn’t drown us we have no need to prove anything. Science on the other hand, is held to a more rigid standard. Every theory must come with a test that can prove the theory wrong. This might seem strange but that is the nature of science. A theory’s value is a function of its ability to withstand efforts to disprove it. To put things in perspective, it is worth noting that some professional naval architects hold rather bizarre opinions that spring from incomplete information or a failure to critically examine a phenomenon or theory. (Some day I will write a book about the ones that I have embraced to my everlasting embarrassment) In this article I will examine just a few ideas that form what might be called “the common wisdom”. Norm Sanders has graciously agreed to “draw fire” for me through his article on sea kayak design in New South Wales Sea Kayaker, No. 29 to demonstrate the point. Now, before anyone draws any hasty conclusions about how “smart” Norm is, let me say that Norm has drawn his conclusions from his personal observations and experience. He is no better or worse off than ship designers of two hundred years ago who, in the absence of more exact knowledge, based their opinions and theories of ship design on their experiences and observations. He is probably better off than the Inuit who built developed the sea kayak. In any case, no one should expect him to know any more about fluid dynamics than I know about politics – which is bloody little. In the article in question Norm says;

“Another design consideration is the location of the greatest width of the boat, the beam. There are three basic choices:

  1. Fish-form-greatest beam forward of centre.
  2. Symmetrical- greatest beam at centre.
  3. Swede- form-greatest beam aft of centre. Boat design is a massive collection of compromises.

Fish-form is fastest, but hard to handle without a rudder. (Have you ever seen a fish without a tail?) Symmetrical is easier to steer and slightly slower. Swede-form sacrifices a little more speed for a lot of directional stability.

The shape tracks well and is easy to turn, the obvious result of sitting further aft with the paddle.”

First let’s discuss geometry. Most people assume that fish form has to do with the location of the greatest beam relative to the centre of the boat. But is this valid? Is it just beam that determines the type or is it also the distribution of volume? Most Greenland style kayaks have the greatest beam forward of midships but their greatest volume aft. (Comment based on the study of twenty-seven traditional Greenland boats done by the author for Eugene Arima’s next volume of “Contributions to Kayak Studies”) Are they fish form or swede form? Most modern sea kayaks that are swede form have both the widest point of the waterline (called the longitudinal center of flotation – LCF for short) and the longitudinal center of buoyancy (LCB for short) aft of midships. Interestingly the location of the LCB has been shown to be more important to performance than the location of the LCF and yet it is the LCF that most paddlers use to define the type. Naval architects stubbornly refuse to label the boat in such a general way and refer to the LCB and the LCF separately thus avoiding the confusion inherent in simplistic terms.

Those who carefully observe boats will know that visually determining the location of the LCF is much easier than the LCB. Perhaps that is why paddlers have adopted the simplistic view – it is too difficult to handle the more complex relationship in casual conversation. Now, let us examine the meat in the statement – that the fish form is the fastest. Is there any reason why it shouldn’t be true if our experience tells us it is true? For the moment, let us imagine two kayaks with the same beam, waterline length, and displacement but one having its LCB and LCF forward of midships (fish form and designated boat “A”) and one having its LCB and LCF aft of amidships (Swede form and designated boat “B”). Now suppose we paddle both boats and “A” proves to be faster. (The method of determining which is faster is of no consequence) Is it not reasonable to assume that the fish form is faster? But suppose everything isn’t equal. Suppose the fish form boat had less wetted surface. In this case we won’t know if it is the volume distribution or the lower wetted surface that made the boat faster. Now, suppose the waterline length of our boats is 14 feet and we are testing the boats at 4 knots. At this Speed/Length ratio a boat with a prismatic coefficient of 0.60 will have approximately 38% more wavemaking resistance than a boat with a prismatic coefficient of 0.53. If boat “A” had the latter Cp then it would definitely be easier to paddle than “B” at our test speeds and now we don’t know if it is the volume distribution or the wetted surface or the prismatic coefficient that makes boat “A” faster. Since the opposite is also true, one can see the problem.

This is precisely the problem that faced naval architects up until the late nineteenth century. It was William Froude who showed that frictional resistance was largely separate from wavemaking resistance, increased at a different rate, and was affected by different factors. This simple discovery ushered in the beginnings of modern hydrodynamics. Until naval architects could separate the two types of resistance they could never know how and to what degree the various differences in shape affected performance. This simple discovery made it possible to isolate the various effects and find out how a change in shape actually affected performance . Unfortunately, one needs either a test tank or a sophisticated Computational Fluid Dynamics program to do the work. The average paddler is no better off than naval architects prior to Froude’s discovery. Incomplete knowledge often led them to erroneous conclusions about hull shape just as paddlers are led to erroneous conclusions from their observations. An excellent documentation of this problem can be found in Howard Chappelle’s “The Search for Speed Under Sail” and reading some of the early texts on hydrodynamics reveals the scientific method at its best as theories were proposed, applied, and cast aside as knowledge increased.

Now, let us return to my earlier statement about proof and the scientific method. If one reaches a conclusion based upon experience how do we prove it is wrong? This is quite simple, we just perform the same experiment with a different paddler and see if the results are the same. If they aren’t we know that, either the test method is flawed or the conclusion is wrong. Unfortunately there are plenty of paddlers who have reached just the opposite conclusion from Norm Sanders. Who is right? Obviously experience can produce much different results and it is apparent that experience and personal observation are badly flawed by prejudice, mood, and superstition. A better way is to do methodical tests under controlled impartial circumstances using a generic hull shape and changing only the parametre of interest. This is precisely what has been done.

Literally thousands of tests have been performed on all types of boats in test tanks around the world and the result is that, for the Speed/Length ratios of sea kayaks when cruising, the LCB is best located slightly aft of midships. The appropriate range appears to be between fifty and fifty -five percent aft of the forward waterline ending2.

Now let us look at the danger in using inappropriate analogies. We know an analogy is best when the conditions are congruent. That is, when all the circumstances that affect one thing are similar to or identical to those that affect the other.

To find out we ask some simple questions. What are the similarities between a fish and a sea kayka? Do sea kayaks travel underwater? Do fish swim at the surface? Are sea kayaks propelled by a fin on the stern? The answer is no to all but why are these questions important? The reason is that kayaks are surface craft and make waves. These waves are indications of lost energy and the shapes that are best for wave reduction are different from those of reducing resistance when fully submerged. We can turn to the U.S. Navy for some valuable instruction here. Attack submarines when submerged are much faster than surface ships and the submarines are fish form. Does this mean that surface ships should be fish form? If so, then the submarines should also be faster on the surface but they aren’t. When attack submarines are on the surface they are pathetically slow. Why? In simple terms, because their hull form is best for underwater travel but not good for surface travel. Now let us add another consideration. The tail fin of a fish is used for propulsion as well as steering. Its shape has evolved to best serve a multiple function. It is worth noting that research is underway to create more efficient propulsion devices that function like a fish’s tail and if you think you know how a tail works you might be surprised at what the research has discovered. Obviously we must be careful when drawing analogies because a fishes tail doesn’t have a lot in common with a paddle or paddler. Let’s move along in the article.

“Yet another consideration is deadrise-the amount of “V” measured upward from the keel. A deep V hull with a lot of deadrise will track well and have less wetted surface, but be initially unstable. Again, compromise is necessary. I find that deadrise of about 12 degrees works well.”

This is an easy one because our good friends at the U.S. Navy put dozens of engineers to work to determine exactly what kind of shape produced the lowest wetted surface. As it happens, for a consistent displacement, a semi-circular hull section has the least wetted surface but things are more complicated. The most critical factor is the ratio of beam to draft and the ideal ratio is 2.8:1 for the lowest total surface area. Why not 2:1? Because any section forward of midships will be of a lower ratio due to the different curvature of the waterlines and keel line. 2.8:1 seems to hit the ight balance overall or so say the Navy engineers. The interesting thing is that a deep “V” hull may track well but not necessarily because it is a deep “V”. It tracks well because the ratio of beam to draft is low and the lower it is in relative terms the better the boat will track (once again in relative terms). Now, suppose our deep “V” hull had lots of rocker fore and aft and we compared it with a round bilge boat had a very straight keel line aft and rocker forward similar to many modern sprint kayaks. Now we would have a situation where the round bottom or low deadrise hull would track straighter that the deep “V”. All of this has to do with something called stability roots (in this case stability is course stability) and the primary factors that affect the stability roots are length, profile coefficients, beam, draft, and LCB but not section shape midships. How do we know? Once again lots of engineers studying course stability to make sure a fully loaded ship doesn’t run amuck in a harbor. The reference in footnote 2 is a good one for an overview of the topic.

Now let us look at another comment.

“There is still another hull characteristic to consider: the keel line from bow to stern. For many years, racing kayakers and canoeists thought that a flat run was the most efficient. Now the trend is for the bow and stern to be raised in the form of rocker. Rocker yields less wetted surface and thus reduces drag. Pronounced rocker also allows the bow to lift more easily over breaking waves. However, the main advantage for sea kayakers is the additional ease of turning.”

Here we have a classic problem. What we know is that the straighter keel line is more efficient in racing kayaks because the net resistance is lower even though the surface area is higher. The problem is that straight keel line hulls don’t handle very well so rocker is added for improved handling. The result is a boat that is “slower” but is actually faster at the end of a race because it is more controllable and the paddler wastes less energy steering. An interesting aspect of this is the role of paddle stroke mechanics. When the racing stroke was straight up and down, directional stability was less important because there was low turning moment with each stroke. With the wider wing paddle stroke the turning moments are greater and directional stability is more of a problem. Now we see recent designs returning to the straighter keel line aft to improve directional stability while the bow remains rockered to allow course corrections for riding waves.

From tank test results we know that course stability improves with:

  • a lower block coefficient (CB)
  • increased L/B ratio
  • stern-down trim
  • increased hull profile aft
  • increased L/H ratio (Length to draft)

Course stability is slightly affected by:

  • location of LCB
  • mid-section shape.
  • waterline shape
  • CP within normal limits.

The next question is whether the rockered bow “…allows the bow to lift more easily over breaking waves”. This is a common observation and one would be perfectly justified in assuming that it was the profile the lifted the boat. But suppose we had a very fine hull with no flare but a rockered profile and compared that to a boat with a straight profile but a flared hull – sort of like a naval ship. The results would be much different. So, is it the profile or the way volume is distributed in the forward sections. As it happens, extensive testing in wave tanks shows that it is how the volume is distributed. So, if some boats with cut-away profiles perform well in waves the rocker is not the cause of the good performance so much as a coincidental factor.

Perhaps the area where opinions differ most is on the topic of weathercocking. Here is what Norm said. “This is particularly noticeable in fish-form kayaks where the long sweep of hull behind the cockpit acts like the tail on a weather vane. Swede-form kayaks behave better, with the bow tending to head more downwind. This is because the paddler’s weight is well aft. Downwind tracking ability can be enhanced in any kayak on extended trips by loading heavier items in the aft compartment. If the kayak is Swede-form and the bow is high due to pronounced sheer, the kayak may actually lay naturally almost downwind which means that the paddler can use far fewer course-correcting sweep strokes and concentrate on forward progress.”

Wind from any direction except dead ahead and dead aft will exert a turning moment (yaw) on the hull. This moment is determined by the relationship between the wind force and the hydrodynamic forces created. Figure 1 shows these forces. For steady motion and in the absence of a rudder the forces must balance so that:

R sin b = W sin a

If C and L coincide there is no turning moment. If L is forward of C the boat will turn into the wind. This is the normal condition but, on some canoes with very full waterlines forward and high ends, L will be aft of C and the canoe will turn away from the wind.

The locations of C and L depend upon the wind direction and the yaw angle. C does not always move aft uniformly and the limits imposed by seating do not always permit effective adjustment of windage or trim for control purposes. This turning moment is offset by the lateral resistance aft due to hull shape, skeg, or rudder action. From the discussion of course stability we can see that whether a boat is swede or fish form is of minor importance but, more importantly, we can’t really know from paddling boats which factor caused what and to what degree. Either type can have weather helm or lee helm depending upon the profile coefficient and the stability roots. If the turning moment aft is high enough, it will offset the turning moment forward and cause the boat to develop lee helm. Some of the most exciting boats to paddle are Greenland style boats like Figure 199 in The Adney – Chappelle book “Bark and Skin Boats of North America. Here is a swede form boat that has a mind of its own. One must assume that either the owner was one hell of a paddler or one hell of a lousy boat builder. The flip side of this was a Chestnut canoe that could only be held into the wind by trimming well down in the bow and paddling only on the lee side. A good friend sums it up nicely – Boats are inscrutable sometimes.” In some boats the tendency towards weather helm is so strong that an inordinate amount of effort is required to maintain course. One method of correcting this is through the use of a retractable skeg. The skeg shifts the center lateral area aft and balances the turning moments at the bow. The farther aft the skeg is located the smaller it needs to be and the lower the added resistance will be. Thick skegs of airfoil shape and high aspect ratio are not always the best answer. In theory they do develop more lateral resistance through lift but this can be offset by the turbulence in the skeg trunk. One must be careful to balance merits of increased lift over increased drag.

The point is that there are so many factors that influence how the boat handles that blaming or crediting a characteristic on one is rarely valid even though ones experience suggests that it is so. Do not, get the idea that everything in Norm’s article is wrong or even that being wrong does a lot of harm. As I mentioned, science is held to more rigid standards than paddlers who can believe anything they want and do so without any visible penalty. What I have done here is show how easy it is to be mistaken.

Does it matter?

One can believe the earth is flat, board a plane in Toronto and fly to Melbourne and your belief that the world is flat won’t change things or endanger your life. You can believe that sea kayaks are propelled by levitating cats and still have loads of fun paddling about with or without your cat. The degree to which anything matters depends upon its importance to you either perceived or real. If you are racing kayaks it is important that you have the fastest boat and an incorrect belief could lead to unwittingly paddling a slower boat. If you are paddling in the open ocean, a mistaken belief might lead to paddling a boat that was less seaworthy than you would like.

If you perceive a boat is superior in some way relative to other boats it may be superior. That fact, however, does not mean you will know “why” it is better. I can tell from reading Norm’s article that he is a keen student of boats and a good observer without ever having met him. The problem is that, no matter how good he or anyone else is, it is not possible to isolate the effect of every different hull form variation through experience. Things are just too complicated. That, of course, is why billions of dollars have been spent building test tanks, testing boats and doing all the mathematical joss that we do to try to understand these very complex issues.

The flip side of this is that one need not understand it to enjoy paddling nor do you have to know it to design good boats. On the wall behind my computer I have a drawing of a seventeenth century West Greenland kayak. The lines were taken of the original and it is one gorgeous boat. My replica is also a heck of a nice paddling boat and I am willing to bet that the builder didn’t even know what a square root from a spruce root.

John Winters
Armour Township
Burk’s Falls, Ontario

Those who are interested in this kind of thing might find an article by the author in the Winter 1996 issue of Kanawa, the magazine of the Canadian Recreational Canoe Association, entertaining. In it I showed how a good boat was almost inevitable so long as the builder does not try to force natural materials to do unnatural things. “The Principles of Naval Architecture” published by the Society of Naval Architects and Marine Engineers is a good source of information on the various theories and tests that have been performed.

(I must thank John for going to the trouble of supplying us with this informative article. It good that we have people with that sort of technical qualification designing kayaks. But it equally as good that we have others such as Norm, Dave Winkworth, Lary Gray, Don Andrews and a whole host of others who are going to the trouble of designing a new kayak or even just modifying an existing design. The kayak world can only benefit from their experiments no matter what the basis of their theories. The kayak world is relatively small and word soon gets around pretty quick if a boat is a flop. So thanks John and all the others for I reckon it great having all these designs available even if there is some healthy arguments between the designers. Ed)