Any formulae for calculating dimensions of plywood surfaces ?

[rwatson]

Originally Posted by wardd

couldn't a clay model be made with a straight edge and curves sculpted by pivoting the straight edge around one point?

of course, but it would be even easier to use cardboard or thin ply on a scale model, but thats a lot of work.

The theoretical calculation of the true shape using math or, as I have found, something like Delftship, is far and away easier

[ancient kayaker]

Originally Posted by rwatson

I still cant make any sense of Freeship or Delftship ... the explanation of the logic in its basic methodology reads like a cryptic crossword ...

I have to agree with you, the explanation of its inner mathematical and topological philosophy and inner workings are best avoided IMHO. There was a usefuk PDF manual on the net sometime ago, which got me started. You might want to reread my post #26 for useful hints if you want to give it another try. I believe the effort is worthwhile but if you having success using Rhino you may prefer to persist with that.

Other than the layout of the toolbar, there is little difference between FreeShip and DelftShip, but FreeShip does plank developments and free DelftShip doesn’t.

I find it useful to think of the hull lines that FreeShip creates as flexible battens connected to the control points by rubber bands. The trick is to use the minimum control points; too many makes the task more difficult or even virtually impossible.

This is the sequence for starting a new design inFreeShip:
File, New, enter number of horizontal control points (minimum 3, no more than 4 or 5 for first attempt!), number of vertical control points = sheer + chines + keel (e.g., for a 5-plank flat-bottom design enter 3), enter length/beam/draft (they can all be changed later), select imperial or metric units, click OK
In Project Settings (Main Dimensions) - enter water density, note the descriptive stuff you can input but just click OK for now
You will get FreeShip's idea of typical hull shape.
I usually shape the stem profile first
Adjust the stern: for example, to get a double ender set the Y-coordinates of the transom to zero then adjust the stern stem profile
Set Visibility to "Show both sides"
In the Perspective view try moving the round hull view around, also check out the Right Click menu; I prefer to set mode to Shade for general viewing.
Add chines: in any view CTRL-Click all the horizontal lines linking the control points, one each for the sheer, keel and chines. Get them all highlighted then Edge, Crease
Prepare for plank developments: Visibility, Interior edges; Layer, Dialog, click Developable for each layer; Layer, Autogroup
Display Developments: Tools, Develop Plates
Bidda-bang, Bidda-boom!
Now fool with the control points; there less there are the easier it is!

Hull data can also be imported although I do not use that capability. Many folk do that to get an existing design into FreeShip, to access the plank development, hydrostatics or other capabilities. Unfortunately the software seems to create a control point for every data point whether it is needed or not, which usually far too many to permit convenient manual changes.

[rxcomposite]

As Ad hoc has suggested "geosiming" it would be a way to go. The late Mike D posted a comprehensive article on geosims sometime in Sept 2002 in this forum. He also wrote an article about Wetted Surface Area.

But you are interested in developable surface area.

There was an obscure program I have come across before (I think Plywood Pro) that optimizes the layout on a sheet of plywood. It is a DOS based program so don't expect it to be around anymore.

You can also use the good old fashioned "Simpsons Rule" the Naval Architect uses. You need an Excell spreadsheet to do this.

As our Naval Archie says, nobody uses Simpsons anymore. With a click on the ACAD, the surface area of the defined boundaries was given in a fraction of a second.

[rwatson]

Yeah, all that stuff would have been of interest a few months ago, but its all redundant now.

Just in case we get a future new user, I have included some file entries at the bottom of this post. If you paste the figures into a text file ( kayak.txt ) , then "Import Chines" in Delftship, you get a sort of a kayak hull.

Then select "View", "Plate Developments", you get plate dimensions like the ones attached.

Actually, I noticed I cant print these plate developments in my free version of Delftship. I guess you have to pay for that.




==============================================

... start of file , import figures below this line, up to and including the words EOF

0


.05 0 .06
1.5 0 .04
2.5 0 0
4.5 0 .055


.03 .01 .09
1.5 .2 .026
2.5 .2 .03
4.75 0 .1

-.015 .018 .18
1.5 .27 .08
2.5 .29 .08
4.9 0 .19

-.03 .02 .32
1.5 .29 .24
2.5 .3 .23
5.03 0 .34


EOF


... end of file

[Wayne Grabow]

Originally Posted by rwatson

I was planning on building a small model of a stitch and glue boat to take scaled panel dimensions from, but was curious to know how easy it would be to calculate them mathematically

Its easy to do a design in three dimensions, but 'unrolling' the surface to get patterns for the plywood is a bit of a challenge without a 3d physical model.

I know it is can be done by various computer software, but I wondered if anyone had formulae or methods to do these calculations, thereby avoiding physical modelling.

Sounds fun ?

Apparently quite a few people have pondered this same situation, and many good ideas have been suggested. For complex surfaces with poly-conic developments linked by common ruling lines, the power of a computer program is required to accurately unroll the surface and provide dimensions. It isn't good enough to find points on the surface and calculate the straight-line distance between the points. The distance must be calculated within the curvature of the surface. I would guess that the Dutch used sticks bent to the proper curvature to subdivide the surface into triangles and provide an accurate pattern.

For my purposes, building boats of less than 20' length, I have been able to mathematically design the hull at full-size; scale the dimensions down to create a model if desired for visualization; and get my sheathing panel patterns only after constructing the actual frame or mold setup. Panel dimensions can be rather easily calculated if the panels have a single curvature with no twist, but that limits us to somewhat simple shapes.

Good luck with your model.

[ancient kayaker]

Originally Posted by Wayne Grabow

... It isn't good enough to find points on the surface and calculate the straight-line distance between the points. The distance must be calculated within the curvature of the surface ...

If there are enough points the striaght line approximations can be close enough for practical purposes.

Originally Posted by Wayne Grabow

... Panel dimensions can be rather easily calculated if the panels have a single curvature with no twist, but that limits us to somewhat simple shapes ...

I have gone even further on one or two boats and used a straight-line plank development, the sheer curve in the plan view and the flare angle to define the bottom rocker and sheer profile. Although limiting to the boat designer, this has produced useful and attractive boats.

[Wayne Grabow]

Originally Posted by ancient kayaker

If there are enough points the striaght line approximations can be close enough for practical purposes.

That is probably what computer programs do, but it would be extremely tedious to do manually if there is significant curvature (forcing closely spaced points).



I have gone even further on one or two boats and used a straight-line plank development, the sheer curve in the plan view and the flare angle to define the bottom rocker and sheer profile. Although limiting to the boat designer, this has produced useful and attractive boats.

Agreed. It limits, but does not preclude a desirable result. I used similar methods on the first boats I built and was happy with the result, but eventually you want more options.

[ancient kayaker]

Originally Posted by Wayne Grabow

Agreed. It limits, but does not preclude a desirable result. I used similar methods on the first boats I built and was happy with the result, but eventually you want more options.

- Boat Design 101 ... gotta my boating education start someplace - ... I used that approach for several of my early boats, starting with some awful flat-bottom canoes.

Later - as I progressed to Boat Design 102 - I moved on to 5-plankers with bilge planks which led to Dora at Challenge: The 100$ boat! which is pretty enough for anyone IMHO. Dora’s bilge planks have a lot of twist which creates the deep forefoot; it made her quick she tracks like she’s on rails, but can be is hard to turn and tends to windsock.

For the next canoe - Boat Design 201 - I again kept the flare angle constant but allowed the sheer line in the sheer developments to curve up at the stem to neutralize windsocking. For both these hulls I cut the bilge planks to fit the holes between the sheers and the bottom plank rather than using plank developments. These were not S&G; note the chine logs in the image. This time I traced the bottom shape from the chine edges of the sheer planks: as a result the bilge planks have no twist and were much easier to glue in place. It produced a boat with much more rake in the stem foot; it is much easier to turn though a hair slower. My canoe designs are still evolving.

That approach doesn’t work for S&G where the plank developments determine the hull shape during the build.

I suppose I will move on to BD-202 and allow the flare angle to vary to introduce twist into the sheers. If I were drafting instead of using FreeShip I would plot a line through the middle of the sheer that would unroll to a straight line in the development - not too difficult - then measure vertically from the line to sheer and chine at each station in the profile view, divide by sine(flare_angle) and transfer these values to the plank development drawing.

Did I just figure out the problem for Rwatson? No! It should work for sheer planks but not for more complicated stuff like bilge planks and vee-bottom planks. However, if they are conic projections it is not difficult, but once we abandon the simple 5-planker and attempt a multi-chine hull I get lost. We need help from someone who took Boat design 301 or later

[nordvindcrew]

Graph paper?

[xuxuki]

Freeship has this capability for example .
victory motorcars

[rwatson]

Thanks for that xuxuki. As it happens, I bought a copy of Freeship some time ago, and almost forgot I owned it. I tried the 'import chines', and found I could develop the hull sides, AND print them, and create DXF files from them too.

The ability to specify lines as figures, and them have them drawn and exported for development makes the purchase worthwhile, whereas I initially had discarded the software as too awkward and complex to use.

[dskira]

Originally Posted by rwatson

I was planning on building a small model of a stitch and glue boat to take scaled panel dimensions from, but was curious to know how easy it would be to calculate them mathematically

Its easy to do a design in three dimensions, but 'unrolling' the surface to get patterns for the plywood is a bit of a challenge without a 3d physical model.

I know it is can be done by various computer software, but I wondered if anyone had formulae or methods to do these calculations, thereby avoiding physical modelling.

Sounds fun ?

As always as did on the floor BC. (before computer) of steel ship, by triangulation.