Time for a new scratch built boat.

A couple of years ago I built 'Scullduggery' a scratch built rowing boat manned by Bionic Bill and Ben.

https://model-boats.com/blogs/68795

Scullduggery is always well received when Bill and Ben are out rowing on the lake, so I've decided to have a go at building something similar, this time a kayak.

After a bit of research, I came across a book by Nick Schade which describes in great detail how to build a full sized wooden strip kayak. There are several instructional videos by Nick on YouTube. There is also a company in Cumbria (https://www.fyneboatkits.co.uk/ ) who offer kits to build full sized kayaks based on the designs in the book. The photo of the full sized kayak is from their website.

The book includes tables of offsets for three different kayak designs. Using these offsets it is possible to draw up a set of plans. Entering all the figures from these offset tables into a spreadsheet makes it relatively easy to change the scale of the plans, but what scale to use?

The scale for Scullduggery was determined by the scale of the readily available Action man figures used for the crew i.e. 1/6th scale. Why not use the same scale for the kayak?

One concern is the load carrying capacity of the model kayak as well as it's stability on the water. The full sized kayak has a maximum loaded all up weight specification of 160kgs. This equates to only 740g for a 1/6th scale model. Not very much for the complete model including the paddler and radio gear, servos, battery, etc. Using a larger scale of 1/5th for the model improves the all up weight target to 1.28kg which may be more achievable.

At 1/5th scale, the completed model will be 39" (986mm) long with a beam of 4.8" (120mm). Using a 1/6th scale figure as the paddler should not look out of place.

As I did with the rowing boat, before starting construction of the hull, I want to be confident that I can build a mechanism that will drive the kayak, provide realistic looking paddle movement, fit inside the hull, and not weigh too much. So that is the next step.

The aim is to achieve movement of the paddler that is as realistic as possible. Steering should also be by means of the paddle, rather than solely relying on a separate rudder.

The rowing boat uses simple mechanics with software programmed into a microchip to achieve the required motion. That way the mechanics are easier to build, and adjustments to the movements can mostly be made by editing the software. The same approach is planned for the kayak.

Two small stepper motors were purchased and a test rig assembled. The two motors can be seen, back to back in the first photo. They each have a brass disk mounted on the output shaft. Pushrods link the disks to the paddle via fixed pivots. The second photo shows the overall arrangement and the third gives an idea of how it will fit into the kayak with a cardboard template showing the internal cross section.

The software controls and syncronises the two stepper motors allowing the paddler to paddle forwards, backwards or to use just one blade in the water to steer the craft. A short video showing paddling forwards and backwards:

Paddling is deliberately slow in the video so that I can work out what is happening and adapt the software as necessary.

A second video showing turning to starboard by just using the port blade of the paddle:

A bit more confusing that one🤔. The final photograph shows an additional pushrod between one of the rotating disks and the shaft of the paddle. This feathers the blades as the paddle tips from side to side. Also in this photo a pushrod can just be seen below the paddler's torso. This connects to a separate servo which operates in sync with the paddle movement and twists the paddler from side to side as he paddles.

After several months messing with this mechanism I scrapped it! The movement isn't what I was hoping to achieve, and I'm not convinced that the stepper motors will have enough torque to propel the model.

Back to the drawing board!.....

The project stalled for several months while I got on with the rest of my life! However I did find a short video clip on YouTube which looked promising so I had a go at building a replica.

A small geared motor at the left rotates a brass disk to which are attached two pushrods. The brass plate in the middle pivots on a screw. The other end of the plate will connect to a servo to move the plate from side to side. This movement of the plate will provide steering as it will tilt the vertical shaft on which the paddle is mounted.

A short video clip to show it in operation:

In the video, there is an additional vertical pushrod which feathers the paddle blades as the enter and leave the water.

I think this shows promise, so I will rebuild it in a form that will fit in the planned kayak.

Before going any further, I needed to check that the mechanism will fit inside the proposed kayak as there isn't a lot of room. A test rig was set up by cutting cardboard frames with holes representing the internal size of the hull. This jig only represents the middle section of the kayak. A cardboard template to represent the cockpit was added as everything will have to be installed through the cockpit hole in the hull.

The Mk 2 mechanism just fitted in the planned position, but could not be fitted into that position through the cockpit. Based on the experience testing the Mk 2 I also had concerns that the geared motor I was using had limited torque and may prove to be inadequate to drive the craft. Hence, I decided to build the Mk 3 mechanism which is a rebuild of Mk2 with a different drive motor..

The geared motor was replaced with a metal geared servo. The servo was dismantled and the electronics removed. Wires were added to the motor so that it could be connected to a standard ESC. A metal pin fitted to the final gear was also removed to enable continuous rotation.

The Mk2 mechanism was rebuilt onto both sides of a plywood panel. Most of the mechanism was reused, although some parts were remade (several times!). The steering servo was also mounted onto the panel (a couple of times).

This new mechanism fits into the test rig through the cockpit.

Wanting to keep the total weight down the rower underwent some drastic surgery. Action man figures are made from thick plastic so that they are very durable, but it also makes them relatively heavy. I found that the joints on the original arms were quite stiff, even after surgery, so the arms have been replaced with a couple of linked springs. A mini servo mounted in the 'abdomen' is linked to the head with a flexible drive shaft. This will enable the paddler to look where he is going! Once he has his waterproof on, he looks OK to me. The arms of his waterproof will be lightly stuffed with soft toy polyester stuffing to bulk them out a little.

A short video showing it in operation driven from an ESC which gives forward and backward paddling:

(There are detail differences between some of the photos as they were taken at different stages in the iterative design of the Mk3 mechanism.)

I've ended up with four controls for the kayak:
1. Paddling - speed and fwd/reverse
2. Steering - tilting the paddle to port or starboard
3. Rudder - traditional rudder on the stern of the kayak
4. Paddler head turning - so he can see where he is going!

These could simply be allocated to separate channels on the transmitter with the paddling speed and head turning on the left stick, and the steering/rudder on the right stick. However, I like to add custom electronics to my builds where this helps with the realism of the model and/or it makes it easier to control.

The first photo shows a block diagram of the controller. A single connection from the receiver carries all the channels received from the transmitter. The controller then processes these and feeds out separate signals to the ESC and the three servos. With suitable software programmed into to PIC microchip the controller will provide the following functions:

1. Pulse stretching for the head turning servo to give more rotation than the standard 90 degree servo output.
2. Driving both the steering and rudder servos from a single Tx channel as this is more logical. The range of movement of each servo can be independently set as they need to be different.
3. When stopping, automatically position the paddle horizontal with the blades out of the water. Tricky to achieve this manually from the Tx.
4. When turning, the paddle is tilted to one side, such that one blade is in use, and the other is held clear of the water. When the turn request from the Tx reaches maximum turn, the paddle stops moving and switches sides to place the opposite blade in the water thus acting as a pivot point for turning.
5. Using the rudder to compensate for the turning effect of the paddle stroke. On each stroke of the paddle, the kayak will tend to turn away from the stroke. On full sized kayaks, it is normal to use the rudder to correct for this turning movement on each stroke.

A lash up of the controller was put together to test the circuit and for initial software development. Once this was working, it was rebuilt in its final form and fitted into an ABS enclosure.

To sense the position of the drive motor (necessary for functions 3, 4 and 5 above) a small magnet is fitted to the brass disk on the drive motor shaft and two Hall effect magnetic sensors are mounted either side of the drive motor. These provide the controller with reference points for the paddle position.

The video shows the operation of different functions of the controller.

There is more work to be done on the software, but this can be carried out while the hull is being built.

Having gained enough confidence that the drive mechanism has a fair chance of working, it's time to start the build. The blog posts so far have been catching up with a build that started almost a year ago. The remainder of this blog will be close to real-time.

The book by Nick Schade contains 'tables of offsets' for three different kayaks. The book also describes in detail how to use these tables to draw plans. This is new to me and I struggled with understanding the tables and the terminology, but worked it out eventually with the guidance from the book.

The nice thing about having a plan made up of lots of numbers in a table (of offsets) is that it is relatively easy to scale the dimensions to any desired size. This can be done manually but as there are several hundred figures to be scaled, I opted to use a spreadsheet.

All the numbers were typed into a spreadsheet and then scaled and converted from imperial to metric using a simple formula. The graph function of the spreadsheet was then used to draw a half profile of each form. These graphs were copied and pasted into Photoshop. In Photoshop, each half profile was mirrored to produce a full profile for each of the 16 forms.

With the plans drawn, at last it is time to start making sawdust 👍

The kayak will be built from wood strips glued together over a temporary formwork.

Templates for each of the 16 forms plus two endforms were printed, glued to a sheet of 3mm hardboard and then cut out. 20mm square holes were cut into each of the forms to allow them to be threaded and spaced out on a 'strongback'. The strongback is a 20mm square aluminium tube. Sighting along the forms showed any which were not quite in line. These were adjusted until the forms sighted true. The end forms are threaded into the ends of the tubs and are screwed in place. The forms are spaced apart with plywood spacers. An adjustable spacer tightened with wedges ensures that everything is rigidly held in place. Finally, the edges of the forms were wrapped with plumbers PTFE joint tape to ensure that when the planks are glued together, they will not also be glued to the formwork.

Western Red Cedar is traditionally used for strip built kayaks as it is strong, durable, flexible and lightweight. The strips used on full sized kayaks are typically 22mm x 7mm. Scaling this down to model size results in a size of 4mm x 1.6mm.

I couldn't find a source of ready made cedar strips but a few years ago I made a pair of garage doors from cedar and had saved a few offcuts. These are around 1m long, 95mm wide and 15mm thick. Ideal for cutting into strips.

Two boards of different colour were selected and cut on the bandsaw into pieces measuring 15 mm x 5 mm. These were then sanded to 4mm thick using a home made thickness sander. The thicknessed pieces were then cut into strips measuring 4mm x 2mm and again sanded to a finished size of around 4mm x 1.7mm

The two boards yielded 100+ strips in two different colours which should be plenty for the hull. I still have several more boards left over from the garage doors so I can make more strips if I need - or perhaps another boat! 👍

The first strips to be fitted run along the shearlines either side of the hull. Temporary short sticks were clamped onto each form to position the top edge of the first strips along the shearline. The position and angle of these sticks was adjusted while sighting along the length of the formwork to ensure that the first strip would lay fair along the length. The top edge of the strip was beveled to match the angle of the temporary sticks. Once happy with everything, the strip was glued to the forms using hot melt glue. I really dislike using this glue, as I find it very messy, and not very good at sticking anything, but in this case I need a temporary fix so hot glue works OK.

Once the first sheer strip was in place, the process was repeated for the other side of the hull. Great care was taken to ensure that the two strips were symmetrical on the formwork, with lots of sighting and measuring.

With the first two strips in place, it was then simply a case of adding more strips alternately on each side of the formwork. Each strip had its edge chamfered where necessary to fit with the previous strip. The strips are glued edge to edge with the previous strip using a waterproof white glue. While the glue dries, each strip is held in place with notched plywood pieces clamped to the forms. The PTFE tape wrapped round the forms should ensure that the strips are not glued to the formwork 🤞.

A small section of the hardboard endforms was removed and replaced with a shaped cedar stem to which the strips are glued.

With the sides of the hull completed, I was able to add two strips along the length of the keel. The spaces either side of the keel were then filled with strips, working from alternate sides for each new strip.

With the hull strips all completed, it's nice to see the slender form of the hull emerging.