flat vs convex surfaces

[lucdekeyser]

In the industrialization of building home and work spaces, roads, furniture and other industries going, straight and perpendicular has increased productivity. Questioning how far these principles can be applied to boat building, more specifically 40+ cruising multihulls, is a thought experiment ro which DogCavalry concluded that "the massive efficiency loss in proplusion would outweigh the minor labour savings". Fair enough, but this is the start of a design spiral. All the other business considerations that you have identified are of course valid; this thought experiment only explores how far off the beaten path the design spiral would lead, this is a stepwise refinement of what loss of efficiency vs what savings in labor (over the useful lifetime of a hull), or in other words, how few curves the body can get by to keep the loss of efficiency of flat panels in the rest worth it (given the SOR common to 40+ cruising catamarans currently built, corrected and repairs with lots of skilled labor, as YT videos can attest to)

Rob's latest proof of concepts optimize for lowest cost of ownership using just-enough high tech materials that require only rather primitive tooling and low skills for building. There is an overlap when only optimizing for "square" industrial built (like for using unsophisticated cut, hold and "weld" at constant quality robots).

A hull as a rectangular cross sectioned tube goes a long way. A proa may not need rocker in that tube. The pointy ends of that tube above the waterline could be just two flat panels as a vertical wedge; below the waterline on each side shallowly convex curved surfaces (for shaping the foil) "welded" to the sides of such a wedge may be all that is needed to keep enough streamline...

Lacking a learned sense of proportions across the many parameters that make a boat, I turn to you to point to more realistic directions for the design spiral.

 

[lucdekeyser]

The simplest shape for a catamaran hull is the triangle, commonly called a "deep V hull". Rocker is obtained by cutting the keel profile into the plates. All the required bending usually relies on the natural flexibility of the used materials (some designers do prescribe some additional shaping).
The hull form is known, and you can find all the advantages and disadvantages listed and argued about in the literature or in forums.

Thank you for this thoughtful reply. It crossed my reply just above. As you point out, in every industrial step, the details and their economics matter.

So, the challenge may be to stress and go beyond the legacy hull forms, like, in a next step, open up the deep V in a U, as a proa to eliminate rocker and need only one bow shape, ...

 

[Robert Biegler]

In the industrialization of building home and work spaces, roads, furniture and other industries going, straight and perpendicular has increased productivity.

Those products don't need to be streamlined, though.

this is a stepwise refinement of what loss of efficiency vs what savings in labor (over the useful lifetime of a hull), or in other words, how few curves the body can get by to keep the loss of efficiency of flat panels in the rest worth it

Rob's building methods already rely on industrially-produced flat materials, given simple curvature. I don't think you will get a worthwhile trade-off by removing the last bit of curvature.

A hull as a rectangular cross sectioned tube goes a long way.

How do you get that tube in the first place? If you need to join flat panels, either for the mould or for the skin, how much labour can you save by not giving the flat panels some curve?

If the tube already exists, then you might have a case. See

But if the tube consists of a ductile material, you might as well squeeze the ends and give your proa two smoothly curved axe bows instead of joining one or two flat plates to the tube. Not so much an option for a sailing catamaran that must tack.

Look at Multihull Structure Thoughts https://www.boatdesign.net/threads/multihull-structure-thoughts.62361/page-254#post-956524 for some information on Sundreamer, with 14 parts taken from the same mould.

The hull shape closest to what you ask for is the transonic hull https://onepetro.org/snamefast/proceedings-abstract/FAST15/2-FAST15/D021S005R005/461517 That still needs rounded chines, I think the claimed efficiency may be sensitive to the waterline, and may not respond well to the significant changes in hull waterline and trim that a sailing catamaran experiences, I don't know how well it would tack, and I wonder why I don't see such hulls around.

 

[lucdekeyser]

Very interesting indeed.

you might as well squeeze the ends and give your proa two smoothly curved axe bows instead of joining one or two flat plates to the tube.

I am familiar with this technique as applied to large HDPE pipes for producing island hopping tris and cats.
I had looked at Sundreamer as inspiration for a simplified version with HDPE pipes. As existing industrial pipes are meant for use on or in the ground, stiffness (and weight) remained an issue.

The hull shape closest to what you ask for is the transonic hull https://onepetro.org/snamefast/proceedings-abstract/FAST15/2-FAST15/D021S005R005/461517 That still needs rounded chines, I think the claimed efficiency may be sensitive to the waterline, and may not respond well to the significant changes in hull waterline and trim that a sailing catamaran experiences, I don't know how well it would tack, and I wonder why I don't see such hulls around.

Very exciting find. The bow shape reminds me of Rick Willoughby's record pedal stabilized monos. Rounded chines may be avoided on the lee side of a pacific proa lee hull to help counter some leeway. A proa also does not tack. An informal discussion on the TH is here. As the TH was compared with the Axe bow, the waterline may be crucial. That such inventions are not around is mostly that the patent holder did not raise enough capital to prove its advantages at an industrial scale (when Axe bow was)

Rob's building methods already rely on industrially-produced flat materials, given simple curvature. I don't think you will get a worthwhile trade-off by removing the last bit of curvature.

You are probably right, in the end. Rob is also very hands-on pragmatic. As a thought experiment I am drawn to start the spiral at the very edge and make sure it is enough explored before "giving in" to curves.

 

[Rumars]

You don't "give in to curves" you accept reality, and that is we already know what works and how.
For a 40 ft cat there are only a few materials to consider. Plywood, wood composite, various fiber reinforced and aluminum.

Rob designes for amateurs who build one boat. No factoy will even fold a flat panel to build a rectangular box from fiberglass, simply because it's more work then using a mold. With a mold form isn't an issue, plugs are CNC cut on portal routers and hand finished. It's probably possible to build a machine to do the final fairing but the expense isn't worth it.
Aluminium can actually be formed using molds, but in the big boat market nobody actually builds enough of them to make that investment.

That leaves you with the low volume producer, either professional or amateur. They have completely different financial priorities when dealing with a build. The level of automation they tool up for is limited, a trained worker is more valuable simply because he is more versatile overall. Higher level of robotics is usually outsourced.

In the multichine world bending a shaped panel and glueing or welding it takes the same time as for a straight sided one. There are various chined hull shapes in use today, we have the deep V, shallow V (single chine), dory (double chine), and multichines (three and more). Since today's catamaran fashion is to use straight topsides even for round bilged hulls most designers rarely use more then three chines per side.
In the round bilge department there are several ways to do it. When it comes to molding we have come a long way with automation, but you need to pay to use the machines.
For example on this one you can thermoforme cores, infuse panels and make custom shaped plywood (cold molding).

 

[rob denney]

The simplest (cross section) shape for a catamaran hull is the triangle.

Maybe, but a box cross section would be close. And better (less wetted surface, more room inside, faster).

Rocker is obtained by cutting the keel profile into the plates.

Why bother? What are the benefits of rocker on, for example a boat that floats stern down (ie an outboard powered catamaran) and can be trimmed so that slamming is not a problem? Check out the Candela P12 with flat bottom, no rocker and double ended hulls below the water, the same hull shapes as the 1 hp/6m cat in the video above. The P12's a foiler, but not until 17 knots of boat speed.

Rob designs for amateurs who build one boat. snip That leaves you with the low volume producer, either professional or amateur.

The infused Harryproas are one offs, but the hulls, bottoms and decks are not built from 4 flat panels, nor are they right angle box sections. The moulds are built from flat panels, with filletted corners. The cross sections are channels, one side taller than the other, with radiused edges. The join between the 2 half hull/decks is along the centreline. In this pic from the Cruiser plans everything is built from flat panels except the beams and hulls, which are from channel moulds built from flat panels.

Advantages:
1) Easy to build, and inherently fair.
2) Because they are not compounded, it is possible to infuse the entire laminate in one shot. This is extremely difficult with compound curve panels, even with thermoformed foam. I'd be interested to hear of any cat manufacturers who do a one shot (outer laminate, core, inner laminate) hull infusion?
3) It's easy to lay a mould release material (thin plastic sheet) on the flat mould. This saves the effort, time and cost of polishing and waxing. It also means the mould is easily repaired, so can be abused/used as a work table.
4) Each hull/deck piece has 2 joins, one mid deck, one mid bottom. Because they are on flat surfaces they are simple lap joins which are self aligning and easy to include in the infusion. Rebates are included to minimise additional fairing of additional laminate across the join, and for post infusion reinforcing around the mast bearings or beams.
5) The foam pieces are only joined lengthwise, the gunwhales and chines are solid glass. Shaping the foam is easy and there are no gaps in the foam to fill with resin or to open up when the laminate beneath them is compressed.
6) The mould materials can be reused for other moulds or for a flat table for bulkheads, etc or can be disassembled if storage space is an issue.

I have not done any numbers, but these moulds are so simple that I suspect it would require a run of at least 20 hulls (ie double the industry norm) to justify building, maintaining and storing a conventional compound curve mould. And that hulls taken from it would be at least double the labour and consumables.

The hulls, beams and bridgedecks of the village suitable boats the students are building (fishing canoes, water taxis, mini cargo proas) are 4 flat sheets of recycled foam with glass each side, joined to make boxes with curved sides and pointed ends. They are hand laminated, with no power tools required so moulding them and vac bagging the core is not viable and probably would not be faster to build. Cutting and glassing rectangular pieces of foam and glass is quick and easy as is setting up and filletting of right angles. There is more resin in the laminate than if they were infused, but the laminates are quite light, so this is not a bad thing from an impact point of view and the actual extra resin used is minimal.

The down side of this is that gel coat and polyester can't be used, the hulls need to be painted. Given the weight, smell, solvents required and fire risk of polyester plus the hassle of post mould gel work, this is a small price to pay.

The students have a field trip to one of the production fibre/panga/banana boat builders in a couple of weeks. It will be interesting to compare times, costs and weights of the foam/epoxy box boats vs the gel coat/csm/rovings/polyester of the production boats.

 

[gonzo]

That may be true. However, the OP is asking about flat panels and not about developable surfaces. His intention is to design a hull with facets.

 

[lucdekeyser]

These are all good points. And in this business it must be all in the right proportions. For example, as target production volume, as brought up by Rumars, I look at the very mature Danish Dragonfly tri production company: high end, high quality, long tradition but innovating very carefully, race sailing at heart but cruise sailing in sales, make much of it in house, ... It takes a lot of effort to make that work, because it is all quite complicated to fit technically and market wise.

All else equal, there remains an advantage for ease/cost of construction automation to stay flat and right angled. This means building a "square" frame to the maximum and add curves but only where absolutely necessary. Going for the extreme is testing the envelope of hydrodynamics and building efficiency, to come up with the least of the compromises to make. So, the P12 is pretty much "square" concentrating its hydrodynamic curves in the foils; Rob's infused models are roughly 60/40 flat/curved, the village models about 80/20...

The Transonic Hull (TH) pointed to by Biegler, is a mind opener as it is all unabashedly flat facetted. From the paper I learned that the traditional design cycle for hulls combines commonly accepted, but, to my surprise, not always so well proven rules of thumb (like hull speed) to fit a hull form to the specificities of a SOR. Lots of the discussion in the paper is out of my comfort zone and I would not be surprised that this rebellious inquiry has solicited lots of criticism. And more likely than not, this TH concept would need way more extensive tow tank testing and prototyping before industrial builders like Damen or so, would replace their competing Axe bow concept.
-
But as this concept has not been, as far as I know, laughed away, I am venturing these speculations:
- flat perpendicular sides could generate less waves
- flat bottoms may help absorb wave making
- there is less slamming
- there was no insistence for rounded edges opening the possibility to ease construction and to also use the straight edge to counter some leeway
- THs would work for a catamaran
- having a transonic bow form double ended in a proa hull would give it reverse rocker, whereby the aft transonic bow would flow in reverse: how much more drag and/or even lift at the back? Annihilating the advantage of the front bow?
- a right angled triangled wedge simplified to only one oblique side instead of the TH's isosceles setup may work also despite separating oncoming waves asymmetrically.

There must be gaping holes in this set of suppositions.

 

[gonzo]

flat perpendicular sides could generate less waves

It will make them bigger

flat bottoms may help absorb wave making

It is possible.

- there is less slamming

Flat bottoms slam hard. That is the reason vee bottoms got developed.

there was no insistence for rounded edges opening the possibility to ease construction and to also use the straight edge to counter some leeway

I don't understand what you mean by insistence or by whom.

THs would work for a catamaran

Why?

having a transonic bow form double ended in a proa hull would give it reverse rocker, whereby the aft transonic bow would flow in reverse: how much more drag and/or even lift at the back? Annihilating the advantage of the front bow?

I don't understand what you are saying. Do you mean a hog on the keel by reverse rocker? That creates more drag. As for the flow in reverse, if you mean the water that gets dragged by the boat, that adds to the resistance.

a right angled triangled wedge simplified to only one oblique side instead of the TH's isosceles setup may work also despite separating oncoming waves asymmetrically.

It is really hard to understand what you mean. What would work?

 

[lucdekeyser]

Fair enough, gonzo. I wished I could share the TH paper from FAST 2015 and discuss it at a white board . From experience, the patent may be construed to expand the scope of the claims beyond what the author would believe themselves and defend characteristics that are just distinguishly unique but may not be that crucial for performance. The paper show test results with actual models. TH was also mentioned in some threads in this forum. I see if I can make some drawings.

 

[redreuben]

I love new ideas and torturing envelopes but i can tell you from here, that hull is going to slam fillings out upwind and scare the bejesus out of you down wind. But .......
What might it look and behave like at 15;1 ?
And where do you keep the beer ?
Get an old Tornado and put these hulls under it and see what happens at 20 knots true

 

[lucdekeyser]

I love new ideas and torturing envelopes but i can tell you from here, that hull is going to slam fillings out upwind

The conclusion on slamming from the results of the tests in the paper :
Transonic hull (i) does not slam at any speed, (ii) retains near
zero pitch in waves, and has decreasing accelerations as speed is
increased in a sea. Hence, it has low added resistance due to sea
state, and its operating speed is not limited by slam

scare the bejesus out of you down wind

I am not so sure because it should be safer surfing down a wave and piercing the back of the previous wave, right?

What might it look and behave like at 15;1 ?

The models above vary in my estimation from 5-1 to 7-1

And where do you keep the beer ?

crew in transonic hull boats drink only ad-fundum, leaving no beer out and keeping the beer in, as the lack of pitching mitigates any seasickness.

 

[lucdekeyser]

drawings

 

[redreuben]

The tests in the paper were in a tank not a seaway.

 

[lucdekeyser]

The tests in the paper were in a tank not a seaway.

In open water ...
TH had a speed sufficient to overtake
a following sea, with an intentionally small speed differential
relative to its crest to facilitate a broach. At lower speed in
which a following wave overtakes TH by the stem, or
diagonally, no broach tendencies were encountered in powered
model tests of dynamically scaled TH model with CG at deck
level, shown in Fig. 19. No broach were encountered in a
smaller TH model tests in surf condition that broached a
powered model of a conventional patrol boat.