Without sacrificing the legendary performance of the traditional guideboat, I have designed a low cost and easy to build high performance rowboat.
Guideboats are considered to be the fastest of the fixed seat rowboats—combining the carrying capacity of a canoe with good rough weather ability. A remarkable combination.
A well regarded guideboat of its day – the 1880′s Grant ‘Ghost’ – was accurately modeled using marine CAD software for drag and stability analysis. This was used for guidance and comparison.
Many styles of construction were weighed and several software applications employed to arrive at the proposed design.
This boat is a symmetrically hulled stitch and tape rowboat modeled on traditional guideboat performance and layout. The hull was simplified as much as possible to speed construction without compromising performance. Just 5 parts make up the hull: two identical side panels, one bottom, and two identical chine panels. Fewer parts also reduces the amount of epoxy and glass required to glue the hull together – better for your pocketbook, and your health. Also, because glass and epoxy don’t float, it is felt that less of the stuff makes for a lighter boat.
It would have been easier to use more panels to create a mirror likeness of a traditional guideboat, but this would have increased the work, expense and weight, a bad trade off.
The first step in this project was to acquire traditional plans for a guideboat. These are available from the Adirondack Museum store in Blue Mountain Lake, New York. I chose to model an 1880′s boat ‘the Ghost’ by Dwight Grant, based on his reputation as a premier builder among Adirondack guides. I also picked up plans for a Rushton ‘Saranac Laker’ for comparison. Modeling the Grant Ghost was first attempted with Carlson Hulls, freeware available at Carlsondesign.
Carlson Hulls is a great program for modeling simple hull shapes. However I found it somewhat limited in that I could not accurately model the complex curves of a guideboat to my satisfaction (see top left). Given that guideboat hulls are usually symmetric, a half model (see top right) was prepared that better used the resources available in depicting the shape. While Carlson does measure stability, there is no way to calculate drag.
Another model was created with the freeware Hulform. This software is well adapted to inputting existing plans, however again, there is no function for exploring drag. (The updated paid version does not have that limitation) This software was developed for an Australian attempt to capture the America’s Cup.
Next, I used the freeware Bearboat (classic version) to create a series of proposed ‘neo’ guideboat plans, (see below) chine less and slightly asymmetric (swedeform in plan, with slightly more rocker in the front) following the conventional thinking of today. Bearboat quickly creates hull shapes that have a minimum wetted surface, the idea being minimum surface will give a minimum drag. Drag calculations can be done with a spreadsheet (from Mariner Kayaks) using the Kaper algorithm (by John Winters of Redwing Designs). Unfortunately this version of the program does not calculate stability, so I had no way to compare my guideboat designs with my models of the Ghost. Dr Livingston has a new version of Bearboat out that does address the stability issue – worth checking out.
Models of the Ghost (see below right) were created with the Prosurf demo as it showed promise (It turned out to be an excellent program). I also modeled 2 of my Bearboat creations with Prosurf.
Modeling with ProSurf Prosurf allows drag (Kaper algorithm among others) and stability calculations so comparison was at last possible. The results were stunning. The Grant 1882 design had a similar stability profile but had consistently 5% less resistance through the normal speed range than the best of my Bearboat creations. I was expecting the reverse. Even allowing for my possible incompetence with Bearboat and the extra drag of the Ghost chine, I did not see any clear advantage in further refining my ‘neo’ guideboat. This was later confirmed with results from Leo Lazauskas’ Michlet software. The late William Atkin wrote that there is little point in trying to better old designs, as they are so well refined. This was proving very true in this case!
After taking a few months off to recover from the shock, I decided the boat would be chine hulled (a guideboat is chine hulled underwater) to simplify building. It would also be symmetric in plan, like the Grant’s with their Ghost and following the lead of Lazauskas (with his Michlet and Godzilla software which indicates the swedeform plan is slightly slower). Ignoring Bolger’s warnings about flat bottomed chine hulls (inducing vortices which create drag) I designed and built prototype #1 simplifying the hull to matching sides, matching chines and a bottom.
I used the plan and side views of Ghost for outlining it, and tried to match the roll and resistance of Ghost while using Lazauskas’ suggestions about parabolic waterlines and flat bottomed hulls (the bottom was flat with curved up ends to eliminate what Winters calls ’deadwood’). This turned out to be a very fast and decent boat (see pictures), though difficult to build because of the plywood twisting involved. I can easily pass canoes and pace kayaks in it, the stability is good (my 79 year old father was quite comfortable taking it for a row), and it handles rough water very well. Annoyingly, the antique Grant boat still had a slightly better drag/stability solution. Prototype #1 had an overall length of 16’1/2″ which required the side panels to be nearly 16.5 feet. An extra 1/2 foot had to be added to the 16’ panels (two butt jointed 4’x8’ sheets) which was a nuisance and the joint interfered with the considerable bending of the ply at the bow and stern.
1880 Grant Ghost vs. Prototype #1
|250 lbs displacement||Prototype #1||Grant|
|length overall||15′ 8-3/4″||16′ 4-3/8″|
The nose bending difficulties of my first prototype prompted me to change the sheer lines in plan view to be a little more like the Rushton ‘Saranac Laker’. Rushton’s boats were not favored by the guides and a overlay of plan view outline, Grant vrs Rushtonlook at a plan view overlay (to scale from the Adirondack museum plans) might explain why. The Grant boat’s sheer (shown in red) gains volume right away, allowing the guide and sport (his paying passenger) to sit slightly farther apart. This gives extra space amidships, perhaps enough for some dogs in a densely packed boat. The Rushton boat (shown in green) forced the occupants closer together with its narrow bow and stern. As well, in side view it did not have as high ends, both features making me think the Rushton boat was not designed to carry heavy loads, more for sporting rather than professional use. The Grant boat also had a hollow entry on the waterline, Rushton’s was straighter, which could indicate Rushton was planning for lighter loads. Note the Rushton boat features a gentle sheer curve (easier to build) that is wider (to fair the curve and regain the lost stability from the narrow nose). Since my boat was intended for sport I modified my plans accordingly. The Ghost’s high bow (apart from being harder to build ) also would obstruct the view from a car with one of these boats inverted above on a roof rack – the only thing I could find that the Grants did not plan for in 1882!
The next prototype was made of just 4 sheets of ply, the bow and stern of the boat’s side panels extending out to the full length of 2 pieces of joined plywood (see below image) to simplify construction and to extend waterline length for maximum performance.
The waterline was extended as far as possible by dropping the bow and stern of the boat as close to the water as possible with one soul on board (see the two images below). This creates a chine that sweeps up (instead of the usual down) as it travels from bow to amidships, but it works well to extend the waterline length, maintain stability and reduce panel stress. With 2 on board the boat sinks in the water and the bow adopts a negative rake, which works well in practice on open water, but could be a bad solution in an area with abundant rocks or weeds.
I added a lot of rocker to this boat after reading a number of Bolger’s interesting books. His small boats have abundant rocker, and he asserts that drag inducing vortices will develop as water travels over the chines, on a flat bottom. I was not happy with the resulting loss of directional stability however, knowing a slightly slower boat that tracks straight will cross the line first. The next boat was planned to have a flat center section with the rocker of a traditional guideboat at the ends.
Of interest to those who like to try alternate products; as an experiment I fastened the gunnels and decks onto this boat with PL Premium (polyurethane) construction glue from a tube, with a caulking gun. The ribs were made of 1″ thick sheet insulating styrofoam, ripped into strips, and glued to the boat with latex contact cement, covered with 6 oz glass and epoxy. 1/8″ thick door skins were used for the hull, glassed in and out with 6 oz cloth. The boat is stored outside, so far so good but I don’t expect it to last and don’t recommend the experiment.
Prototype #2 was fast, but with slightly less primary stability than #1 (no problem), and less directional stability (a problem). With the software and the imposed constraints (16 foot total length for the side panels, 5 piece hull, flat bottom with rockered ends, parabolic waterline, acceptable panel twist) I could make the boat model ‘faster’ than the Grant model, but not with the same primary stability. Matching the stability always gave a slightly slower boat.
The problem, I decided was rooted in my assumptions. An optimal small boat hull (for minimal drag near hull speed) according to Lasaukas would be something with no rocker, a parabolic waterline, and an elliptical hull section, but this was without regard to stability. The Grant boat has rocker, a hollow entry waterline and a mid section that approximates a parabola. Lasaukas was optimizing drag to a particular width however, if a boat was optimized for drag to a particular stability the lines might be different. I noticed if slight rocker was introduced drag might increase by a point but stability (at 5 degrees roll) would increase by four times that. It seems to be that for very narrow hulls, a boat optimized for low drag near hull speed with regard to stability, will have some rocker.
I noticed a similar but less pronounced trend for the waterline when I narrowed the entry from a parabola to a straight line – stability increased minutely, but at a higher rate than the drag. My studies seem to indicate that a hull optimized for low drag, with regard to stability, will have a straight (like the Rushton) or even hollow (like the Ghost) waterline at the entry (as opposed to Lasaukas’s parabola). I attribute this to taking away some of the displacement where it has no moment (near the centerline) and re distributing it where it has a moment (away from the centerline).
Holtrop writes what seems to be the opposite: that drag and stability reduce with increased rocker. A possible explanation might be that Holtrop was dealing with canoe hulls that are 50% wider (on the waterline) than what I am dealing with. Increasing rocker on a wider hull might reduce displacement where it does have a moment – that is, away from the centerline. Certainly this would be the case in a wide rectangular sectioned hull with a transom. The unexplained deviation in Holtrop’s data, (where stability increases with rocker in his ‘design #2′) can be explained by his reducing displacement where it does not have a moment. He gave the boat rocker in just the ends, where the hull is narrow and does not contribute to stability. Some of this displacement was transferred further from the centerline, where it does contribute to stability.
All this to say that some rocker appears to be warranted. My Prototype #2 had too much rocker, so #3 was a return to what was used in the old days. Otherwise the basic hull shape of #2 is maintained, to allow for tolerable material bending.
My research with marine design software was pushing me from the ‘ideal’ of today toward a solution similar to an antique design. Amazing! For the boat on offer here, with the above in mind, I added some rocker, and made the entry straight (not hollow or parabolic). It would have been helpful to have started with a boat less full in the ends, like Rushton’s. I increased the prismatic coefficient slightly over the Grant’s ( to give me a theoretical top end advantage, with the speed variation of a rowboat between strokes, one should optimize for a speed higher than the anticipated cruising speed). Also, a little depth to the stern end to gain directional stability, this is to counter the Ghost’s slightly better block coefficient – block coefficient being a sometimes faulty indicator of directional stability. The other indicator of directional stability is waterline length over waterline beam which I have the better numbers at.
The Grant guideboat’s mission was for a higher load (two men, a deer, dogs and gear) but it balances the stability speed equation at low loads (one person) very well. It is a longer boat by a few inches (giving it higher hull speed) and is extremely well designed making it tough to beat. It’s symmetric but hollow entry waterline and rocker should put it at a speed disadvantage (hollow entry is usually helpful in the low end of the speed range) but that does not pan out in this informal study, when stability is also factored in. A versatile design in a longer package give it the edge.
If cost is factored in the picture is a little different. New guideboats retail for about $13,000.00 USD, kits sell for about $3,800.00 USD. I am building mine for around $500 USD.