The following notes and diagrams give you the plotting points for producing NACA sections for rudders and centreboards. Once again, a drawing of our portly mate NACA0015 has been used (Fig.1), mainly as an aid to following the table (Fig.2). From the examples in the table you can work out what other slimmer or fatter sections would be if needed. E.g. halve the numbers for 0018 to get 0009, which is a common section for racing boat rudders. As I mentioned in the last article, practicality limits the options for dinghies. The thickest part of a 10-inch wide board would be .3 of an inch with NACA0003, or 3 inches with NACA0030. Enough said.

The first job is to ascertain what the rudder/centreboard's length and width are going to be. The width is the dimension you will need in order to use the table – bear in mind that these numbers are always percentages of the chord, that is, the board's width. You need to know the finished length just to cut suitable pieces of timber. Remember to allow for flattening the trailing edge when considering the board width. The width of this flat can be anything up to 10% of full board thickness without penalty (thickness, not width/chord - check the drawing.) You will see from the table that the thickest part of the NACA four-digit section is always 30% of the overall width from the nose (first column). The temptation to use a 10" wide board is almost overwhelming, with all those decimals and percentages around; by moving the decimal point one place to the left, all of those vertical percentages become measurements in inches for a 10" chord, so at 30% aft the half-width is .75", which is 1½" overall, i.e 15% of chord, hence 0015.

The next difficult decision is deciding on the maximum thickness you want. As the thickness increases, the section can manage steeper angles of attack without stalling, within reason, but you also have increased drag, which does affect small light boats quite quickly as this dimension grows. Racing rudders are generally in the region of 0009/0010, but on a cruising boat they are found up to 0025. As centreboards are wider than rudders, and work at shallower angles of attack, as we saw last time, their NACA sections are usually smaller, say up to 0015 absolute maximum. Do you want a foil which resists stalling at the expense of some extra drag and possible loss of speed, or do you want the last iota of performance out of your racing boat and so would prefer to maintain concentration on your technique to keep the likelihood of stalling to a minimum? Remember also that the fatter foils are easier to carve. And much stronger. Drawing a number of sections before you make up your mind is probably the best approach.

The malleable plastic 'sticks' available as drawing aids (not French curves) are one way of joining most of the dots. A flexible batten of some sort should give a fair curve more easily, but the first four plotting points from the nose present the greatest difficulty. The nose is usually plotted by drawing a circle, but that does not imply that it ends up being a round nose (Fig.3). Having drawn your section, you can then work out the best way to form it.

The obvious way is to produce a laminated blank of sufficient size to accommodate maximum chord and thickness. If this is the way you want to go, do not hold the work horizontally and complete one side first – you won't be able to hold it while you complete the other side. Grip it edgeways in the vice(s) and form first the nose then the run to the trailing edge. This will leave you with a flat down the foil roughly between 25% and 60% along the section on each side. It is possible to shape this part last with the board held horizontally because it is a less complicated curve, and it is a good idea to retain the flat near the upper end of the board anyway, where it is out of the water when in use, so the rudder cheeks or centreboard case can hold the foil more securely.

It may also be possible to build up the blade from strips of varying dimensions, so that glue lines coincide with the right thickness at some points as an aid to shaping, but I have found this promising idea to be of limited use when I have got down to actually trying it. It might work with the fatter sections, but each strip would have to be centred exactly on its neighbour. Fiddly.

Another method is to have a core which is thick enough to contain the dimensions at the ends of the foil section, and then glue strips on to both faces where you need more thickness. I believe this is the way to go. Taking our 0015 as an example, the core blank would, say, be 10" wide by 1" thick, formed of several laminations to inhibit warping and splitting. Before you even begin to glue on extra strips, you can shape the nose back to plot 3, and also the last three inches of the trailing edge from about plot 12. Not only is the job more under control, but it is easier to produce two small templates or profiles in plywood (or less durably in card) to check these edges as you work. This will leave a central section on both sides of about 5⅜" which will need building up.

Preparation and a meticulous approach are the keys to this job. My first attempt years ago was formed from a full-thickness blank of laminations in contrasting timber, with the object of its looking beautiful when varnished. After the lengthy process of selecting the wood, followed by the initial gluing and cramping, then a lot of time with plane and Surform, I made one or two serious errors through not keeping my eye on the ball, which I smoothed out then ignored, being reluctant to scrap all that effort, and reasoning that my unique shape would probably prove to be better than the conventional section anyway. The human mind is infinitely persuadable, especially when the persuading is being done by itself.

It became a thing of beauty, but not a joy forever. The NACA/NASA foil designers should have seen it before it was burned, as they would have recognised instantly that it was the perfect section to operate efficiently in a fluid that none of them had encountered before, but which is no doubt commonly found on the outer planets. The second effort was much more successful, being more carefully fashioned, and it improved the handling of the 17-ft mini cabin cruiser for which it was made, at all speeds. There was far less need for extreme movement of the tiller to steer the boat.

Not everyone regards the NACA sections as holy writ to be followed rigidly, as can be seen from the final word on cruising boat centreboards, which I leave to the New Zealand designer John Welsford, whose dinghy designs, mainly yawls, are proving to be immensely popular and have already been used to achieve some epic dinghy cruises – in both hemispheres. Also, Dave Jennings was sent a copy of this article, and has added his own comments, which are appended below.

"On foils, I have experimented with all sorts of stuff and come back to this simple formula. For a non-racer centreboard, the thickness should be 12% of chord, the leading edge radius should be 1/6th the thickness, the maximum thickness should be at 40% back from the leading edge, the trailing edge should be square across 5% of the maximum thickness. Make a nice fair curve from leading to trailing edge. This gives a high lift section with a very high stalling angle which will assist the boat to tack reliably in very light or very rough conditions when the low-drag sections tend to stall and lose lift. The leading edge is not prone to damage if you hit anything, either. It is a thick section, though, and on some boats would not fit the centrecase, so you might have to scale the proportions down to suit the thickness possible."

— John Welsford

Dave Jennings Writes:

I am in full agreement with Keith on his description of the theory of foils shapes. Most articles that I have read suggest that the best NACA section for dinghies will fall between 0009 and 0012 with the thicker sections being more suited to slower boats; this is borne out by John Welsford's views. I expect that most of our dinghies would be classed as 'slow'. The thicker sections will give more lift (less leeway) and be more resistant to stall with drag only becoming a real issue at higher speeds. However it is very likely that you will limited by the practicalities of things such as the size of the existing centreboard casing or the rudder stock. If you plan to race then you also need to consider the class rules which may limit what can be done to the standard foil shapes. As the thickness of the foil becomes smaller then the exact section shape becomes less critical.

I am currently building a centreboard for my new boat which I plan to shape using a router guided by a template of the desired shape. The technique was well explained in a "Watercraft" magazine article about a year ago. I will be using the NACA section and with the thickness coming out at around 1½ inches on a 12 inch chord it is just below the 0012 section. This still looks very fat compared to most dinghy centreboards. The last thing that I would suggest is worth doing once you have carefully crafted your foil is to sheath it in glassfibre and epoxy resin, particularly along the leading edge. This gives it some much-needed protection against abrasion as you run aground or fail to get the board up quickly enough as you approach the shore.