Strip-Building on a Cylinder Mold

Haven't moved this project much further yet but I have refined the approach a bit.

Referring to my first post, I see no advantage in stapling the sheer strip to each station, it can be stapled to the center mold only like all the others, provided there is some feature on the mold to keep it in place.

The attached figure shows a modified mold which allows predictable rocker to be added at the same time as the sheer plan curve during the process of bending. The plank attached to the lower front edges of the station molds permits the flexible mold to bend only in the horizontal plane, being much stiffer than the strips which are not glued together at this stage; it also keeps the sheer strip in place. Because the bottom edge of the hull half is not in the same plane the total curve is shared between the rocker and sheer plan curve.

After bending, the half hull is fixed with glue, cut along the centerline and joined to a second half hull. The bottom and stems are added at this time. It occurs to me that the stems can be added before the bending step but I haven't shown it here.

I understand that the whole thing is highly theoretical at this stage and needs to be demonstrated (or disproved) by experiment. Hopefully I will have time to do that during the Winter.

 

I haven't moved this project into the demonstration stage yet as the unfinished boats are beginning to clog up the workspace. I want a larger workshop so I can have more boatbuilding projects under way but the reaction to that suggestion from the Missus was a bit dramatic and - when I think about it - predictable.

However I have done some refining of the idea.

When the strips are laid flat and taped together ready to drape over the straight mold, the tape can go on the inside of the bilge so there is no need to bevel the strips.

I was worrying about flat spots forming when the mold and strips are bent to the sheer plan curve, but in practice the strips will be under compression, so they are more likely to bulge. That is much easier to deal than flats, I can just use a soft roller to fair them. That also eliminates the need for bead and cove strips.

How to plot the profile for cutting the hull halves before joining them has been solved. I simply put the design into a FreeShip file and use half the lengths of the plank developments to mark off the strip ends from the midships station. That can be done while the strips are still flat before they are on the mold. Once on the mold and bent, they should fall into a nice smooth curve ate each stem, showing the bending is complete. The hull half outline can be cut after the strips are glued together.

 

Terry,

I'm not fully sure I understand what you are proposing here, but it spurred me to ponder the concept. In addition to your battens, you could run a central "camber" member down a fitted cutout in each mold. It could be rectangular to limit camber about a single plane or it could be square to allow flex in two planes. This camber controller would extend in either direction beyond the molds and would allow leverage to be applied and a bending force created about a central (midships) fixed restaint. A canted rectangular member control a predetermined sheerline. A planwise sheer shape could be controlled by either a single, fixed "fulcrum" or multiple fulcrums to spead the bending moment over a larger area and spread more width to the ends. Perhaps your levers on the battens is the same concept. My thought about using a single continuous member was perhaps an increased chance of developing a fair curve.

An interesting and possibly complex endeavour.

 

Sounds like you have a good grasp of the idea. The center camber member that you are proposing corresponds to the rectangular section plank shown attached to the lower front edges of the section molds in post #16. It keeps the curve in a single plane.

A central location would even out the strip-to-strip motion during bending but there would be tension in some strips which could result in flat spots, however. The location on the outside edge of the mold is easier to do and will result in a neutral or compressive force on the strips which will be easier to fair out, I believe.

The curve doesn’t have to be in the horizontal plane - it can be changed to create rocker as well as the planwise sheer curve. The rocker curve is simply a shallower copy of the planwise curve. It is a consequence of straight line plank developments and a constant flare angle in some of my boats - it looks and works fine.

As far as extending it off to the ends, that is a neat idea - you are moving ahead of me on this. It would cause the curve to be hyperbolic rather than circular, and intermediate curve types could be made too. The hyperbolic curve exhibits sharper midships curvature like many modern canoes, especially solo ones, with finer stems. One could even have an asymetrical hull although that would require the bending arrangement to be mirrored for the second hull half.

My levers generate a circular arc. I will put them a little way in from the ends, so the strips will be straight out to the stems - keeps it simple and I can predict things like stem shape and hull half cutting profile mathematically.

Perhaps not so complex; we will see when I build a demo boat. BTW I informed Kurt Hughes of my concept and I think he is interested in the outcome. I think the idea has some advantages over his cylindrical mold concept that it is based on, especially hull shape predictability which might allow the building of class boats to his designs.

 

Starting to do some tests. Bending forces for the dry strips does not seem excessive, the frictional forces between the strips do not increase the bending force significantly, probably because most of the pressure is at the center of the hull where there is little movement between adjacent strips.

I did a small test panel using PVA glue injected into the gaps between the strips using a hypo. I need to try a smaller needle though; the gaps are less than 1/2 mm.

Result was neat with no leaks to the inside which is nice as sanding the inside is the hardest IMHO. Although I tried to completely fill the gaps there seems to be less hardened glue than when it was wet. Perhaps the glue sets slower than when it is in a well-fitted joint and has more time to soak into the grain, in which case an option might be to pre-moisten the wood before applying the glue. I may try a polyurethane glue - I am trying to avoid epoxy for the demo unit.

 

Ancient Guy,

You lost me a long time ago, but I'm still interested.

Can you post a picture of your test and the tool?

Marc

 

It's good to know there's still some interest . . .

I don’t have any pictures at present, at this point I am just fooling around with a few strips left over from a recent build, getting an idea how they will behave on a mold.

I don't have enough material or space for an entire boat but I don’t want to wait until I finish the boats that the workshop is cluttered up with. Six months is long enough for the head-scratching stage, so I plan to do a scale model to prove the principle and work out any bugs while the other boats continue to grow. I will post pictures as I go along.

The glue test was plenty strong enough to hold the hull halves together during hull assembly which is all I needed to know so the next stage is to build a small mold and try it out.

Rev: the model will be 2/3 scale of a half-hull (midships to one stem) of my 12' hybrid stripper canoe design, Rose Lee.

 

Some pics of the mold; it needs a little more work before I can do bending tests with strips but these may illustrate more clearly than words what I am doing. Note that it only deals with the turn of the bilge not the flat bottom (in the design used for for this exercise) or sheer planks.

The first pic is the straight mold ready (nearly) for loading with strips. The stem is nearest the camera.

The second pic shows the mold bent, it is a quarter of a boat with two centerlines drawn beneath. There is rocker but it is only about 0.4" (the model is 48" long) so it doesn't show up well even though I let the batten spring up above the midships mold.

In the third pic I have attached a batten to show the approximate position of the keel and stem.

Now I have do the math to predict the plank lengths and stem shape, since predictability of hull shape is one of the things I am looking for. To save time I will use a spreadsheet rather than a hull design program like FreeShip. To simplify the math I have used combinations of straight lines and circular arcs in the boat design.

 

Ancient kayaker,

Thanks for the pictures. So far it matches what I imagined.
It appears that the bow will turn into something rather flat in section instead of more vertical. The other thing I notice is that the sheer will not have a continuous board - all the planks will curve up to the sheer with a fairly large angle cutoff line.
This will be really interesting to see the sheer line from the top. I assume the deck will be rather blunt at the bow. Is this what you see now?
I imagine this hull to look a lot like the early America's cup boats, long overhangs fore and aft.

It will be really interesting to see the final hull, but don't let me push you, much.

So 2/3's of 1/2 of 12' = 4' if I can still do the figurin. This will be similar to a test piece I did for a kayak (I call it a "nose") so now I can understand what I am seeing.

Marc

 

These images are for Rose Lee, a similar design, also a stripper. The stems are long and gradually curved up rather than the knife-like plumb stems you mostly see. Each plank maintains the same angle for its entire length so there is no twist, this shows up well in the last (bodyplan) view. I am building her with ribs rather than glassing the hull, and I am really liking the way she is turning out.

Rose Lee's bottom and sheer planks are plywood, so she is only partly stripped. Hence the name; Gypsy Rose Lee never took much off so she was only a semi-stripper. For the demo I will probably make a stripped "football" shaped bottom with fore-and-aft planking to see how it looks as that is something I want to do later.

I probably won't bother with sheer planking for the demo since it is more-or-less vertical so should present the least resistance to bending on the mold.

There’s no need for you to do that anyway, the voices in my head are doing a great job! I don’t know how long the math will take, I’m a bit rusty, and I may have to prep some strips, so please stop holding your breath as I don’t have malpractice insurance!

Right on!

 

Terry,

I had forgotten the plan for Gipsy Rose. To match the same shape you would have to change from a flat table to something that matched the football in your middle picture above, I beleive.

Looks like you are going to need about 10 monkeys to get the mold to gracefully bend in the two directions.

I like this more now that I really understand. The only issue is being limited to the fore foot shape that falls out.

Don't worry about the malpractice, you are getting more done than I am.

I wonder if Constant Camber molds would give you more of a narrow fore foot?

Carry on,

Marc

 

I’m not sure I understood that. For this model the bottom is flat apart from a little rocker, but it’s NOT horizontal the way the mold is resting in the previous photos. In the first picture in post #23, note the way the inboard tops of the station molds are sloping upwards, this shows the plane of the bottom plank. When the mold is bent this angle imparts rocker to the bottom at the same time as curving the sheerplan.

See last para for explanation of fiendishly cunning rocker trick . . . generally I avoid using monkeys!

There are design changes I can make to get a different forefoot, for this demo - as for the Rose Lee design - I am subverting these design elements to an understandable desire for simplicity. Actually, I quite like the look and am looking forward to trying out Rose Lee in the water next season. I have another home built/design 12' canoe that has a more conventional forefoot (Dora: see link below), which goes very well, and I want to compare the two.

http://www.boatdesign.net/forums/projects-proposals/challenge-100-boat-12176-7.html#post195863


- the math for the plank lengths is done, wasn't as difficult as I expected. I am plotting out the stem profile so I can see if it's expected shape coincides with the predicted plank ends.

 

Using strips to make a cylinder is going about it the hard way. Plywood will work better. It is faster, cheaper and easier.

 

Gonzo,

In a small size boat it will be hard to beat strips for price. I get mine from Home Depot in typical 8' lengths. And I can't get marine ply in the middle of Texas without paying lots of money, especially less than 1/4".

The kayak I'm building (seems like forever) uses 3/16" strips. Add to the fact that any plywood is a lot heavier than cedar, I don't really see how you are correct.

Marc

 

Gonzo: I doubt there is much cost saving with two layers of plywood plus the cold molding adhesive compared with a single layer whatever reasonably light wood is available in a given location, assuming glassing costs are the same in each case.

I have been following some cylinder mold building blogs and it is clear that some first time builders have problems with the vacuum bagging stage at least, and are nervous about the forces required to join the hull halves.

Then there is the potential to have predictable hulls that come out exactly "as designed" with build-to-build repeatability allowing cylinder molding to be used on class boats.

AFAIK this is a new method and anything new needs time and testing to determine its pros and cons. It still seems a worthwhile endeavour to me, in fact it has so far turned out easier than I expected.

Marc: if you think 3mm ply is expensive in your location check out the price of the 1 or 1.5 mm ply that you would need for cylinder molding - if you can find it . . .