Rozinante - Tilting at Windmills
Speaking of models and junk sails,
Leave it to Todd to have already done what I’m thinking in a model in the way of a rig. I’d want to go a lower aspect ration than he has hear, but it illustrates the general idea beautifully.
http://www.youtube.com/watch?v=NfmoYQBCa0A
Tom
I’m pretty sure Proadude is James Brett. Nice model, that’s for sure. Speaking of junk rigged proas, have you seen Terho Halme’s Ping Pong? http://youtu.be/QGCGqyLSRWk
You’re asolutely right, I stand corrected.
Yes, I’ve studied the heck out of ping pong. Wonderful boat. I prefer the flying proa configuration as opposed to the harry style though. definitely another great testament to how well a junk can work.
I believe I read somehwere how Gary D started playing with the junk on wa’apa after he “got his ass handed to him” by rett at the NZ proa conference 😉 his 5m boat just smokes along as well.
Tom
I agree. Roz is a purrrty boat!
Very useful post.
I’ve been pondering lifting rigs myself. Then, when I read about the downside, that is, that when a boat heels you get exactly the opposite of what you want, I went, “DOH!”
Nice list of “Do’s and Dont’s.” It’s funny how I keep coming back to tradition and the proven. But it is SO fun to push the edge!
BTW, the link to the construction technique gives a redirect followed a by a 404.
Cheers,
Rick
[Edit: 23:18 PST—Link works now]
Construction: wood/epoxy, using the Cylinder Mold technique.
What is your experience with cylinder molding? Did it work well? Any worries about trapping air between the sheets of plywood?
Regards
Robert Biegler
I used 1/8” luan door skins, which was the cheap route and a bit risky since the quality of said skins can vary. No problem with vacuum bagging, the problem I had was the double-ended shape of the hull did not work well with the system. Cylinder molding likes a transom hull. Achieving hull symmetry was quite a challenge, and I was not able to get the semi-circular hull sections that are possible with a transom hull.
Doing it again, I would strip-plank the bottom or build a chined hull.
No problem with vacuum bagging, the problem I had was the double-ended shape of the hull did not work well with the system. Cylinder molding likes a transom hull. Achieving hull symmetry was quite a challenge, and I was not able to get the semi-circular hull sections that are possible with a transom hull.
Regarding vaccum bagging, I am not sure I expressed my question clearly enough. I remember an article in Multihulls Magazine where someone worried that when you put on the second layer of ply, air bubbles could get trapped between the sheets. The pressure of the vacuum bag would then compress the bubbles, but the ply would still only be joined in some places. I helped someone vacuum bag a Farrier trimaran, and I saw that some of the foam he was using had little holes. I think he used that when he put foam on top of laminate, so that air could be sucked out through the holes and would not get trapped between laminate and foam. So I expect you could do the same with cylinder molding, and if there is a potential problem, that would fix it. But I don’t know whether anything needs fixing in the first place. Air trapped between sheets of ply seems like a plausible concern, but I try not to mistake plausibility for reality. I have experienced the difference a few times.
Regarding transom hulls, is it that cylinder molding likes fore and aft asymmetry? Does that perhaps lead to less longitudinal bending? Did you have trouble with achieving fore and aft symmetry, or trouble giving the two sides identical shape?
Regards
Robert Biegler
I should mention that I hired an experienced builder to fabricate the hull who had several large cylinder molded cats under his belt before Rozinante. He built the cylinder mold and did the vacuum bagging, I may have missed some details in the process.
Regarding transom hulls, is it that cylinder molding likes fore and aft asymmetry? Does that perhaps lead to less longitudinal bending? Did you have trouble with achieving fore and aft symmetry, or trouble giving the two sides identical shape?
Cylinder molding is like a tortured ply construction, except without quite so much torture, so there a forces building up in the panels as they are stitched, spread and forced into shape. a hull with a transom allows the panels to spread as you head toward the mid section and does not require them to come back together, this is much easier to manage than bringing them back together into another bow. The transom hull is working *with* the way the ply wants, the double ender is working *against* it. The builder was surprised how difficult it was to get the hull to behave compared to even much larger catamarans, and we attributed this to the double ended shape.
I have long been fascinated with manipulating flat panels into curved and compound curved shapes. I have experimented over the years with paper, card and fibreglass sheeting both as models and one foray into a full size section of a daysailing beach proa. I found that there is a lot less sheet development or compound curving going on than most people familiar with ‘torturing’ ply’ have assumed. In reviewing my experiments, my view is that a goodly amount of the apparent compound curving is actually conical projection instead.
I finally came to this conclusion after fooling around with double ended shapes such as in Michael’s Rozinante. The shapes always ‘hogged’ along the keel line. The apparently added surface area along the compound curved topsides came not from expanding the panel by curving it in two directions but instead came from the keel line. The keel moved up towards the sheer line and thus allowed for the extra curving in the topsides. In a sense, the surface area was displaced rather than increased overall.
You can see this hogging readily with a one dimensional model. Trace the shape of two curved sides (a la cylinder moulded panels) joined at the keel line in a vee shape. Place a string or cord along that line pinching it at the keel point and pin the tops in place i.e. at the sheer line. Now move the mid topside points outwards and you will see the sting at the keel move upwards. I’m sure most people could readily visualise this.
In a 3D model, there will be forces moving the sheer line downwards in compression at the same time but because the forces are acting along the panel (i.e. vertically along a vertical panel) the panel will resist deformation in this plane (particularly with the help of a sheer clamp) much better than will the panel at the keel line where it is much closer to horizontal and the forces are acting more across the panel i.e. across the thickness which is a much shorter path than along its surface.
In the case of Michael’s Rozinante, by using two curved cylinder moulded panels, I dare say the hogging was masked by the fact that the two panels formed a ‘broadseamed’ shape along the keel line. As with a Tornado hull, material is effectively added where the hogging will tend to take place (midspan) but there will still be massive forces wanting to raise the keel line when the sides are bent up. In the case of the Tornado hull, as opposed to a double ended one, extra material can come from the transom edge.
If you cut a double ended flat keel line shape amidships after it has been bent up (and hogged) and place the two halves on a flat surface so the keel of each half is flat on the horizontal plane, you will find that there is a wedge shaped gap between the tow halves. In my opinion, this is where a lot (but not all) of the added surface area comes from in a ‘developed’ hull. This is not noticed in a transom hull nor it is a problem but it quickly becomes a problem in a double ended hull.
I hope this helps the understanding of the forces at play here.
The transom hull is working *with* the way the ply wants, the double ender is working *against* it. The builder was surprised how difficult it was to get the hull to behave compared to even much larger catamarans, and we attributed this to the double ended shape.
Thanks. That is useful information.
Regards
Robert Biegler
Do you have any information regarding Rozinate’s fate? Do you remember the name of the person you gave her to? She was a very pretty boat and it seems a shame that she should fade away so quietly.
Do you have any pictures or video of her with the rig up, or under sail?
Great thread and the advice is spot on regarding the wisdom of limiting resource exposure and testing concepts at an affordable scale.
Your “Roz” looks very beautiful and well-designed. I can’t say I’m a Junk-rigged junkie - all the good reasons it makes sense are unquestionably right, but I’ve learned the hard way that technically correct and real world correct sometimes are different.
I guess for me, proas are visually far enough from mainstream enough with conventional rigging. Adding in a junk rig seems almost like intentionally provoking negative reactions from the dock rats. Being a source of chatter amongst the armchair critics isn’t exact a resale-enhancing quality. This is strictly a North American yacht club perspective, as things unfamiliar are subject to derision, regardless of their true worth.
I’ve found major projects like your Roz to have the best chance of success if you limit design evolution to only one major variable or risk element at a time. Once a design element proves itself one way or the other, then you move on to the next risky item. Combining more than one risk element at the same time has been a project killer for me. Your Roz project combined several risk elements concurrently:
- Junk Rig
- Windward inclination of junk rig
- Single-sided Ogive sail (wind always on same side, with luff/leach swap during shunt). This kind of rig is always loaded with wind pressure - and it is highest in the middle of a shunt as the rig is square to the wind.
- High speed shunting - perhaps too fast for boat momentum, loading of rig is almost instantaneous, whereas “normal” rig shunting loads sails from crabclaws to Marconi very incrementally and slowly as the wind changes from one side of the sail to the other. Even Brett Design’s junk rig is shunting speed limited as it is not single-sided.
- BI-directional Ogive trunked rudder foils
I’m not criticizing your design choices at all - they are all soundly thought out and logical - but having that many variables in the equation at once can be overwhelming. I’m not so certain that building a scale model helps when there are too many variables in play concurrently.
Advice gained from “expert” hindsight is cheap and can be really annoying. That isn’t my intent here at all. But in the spirit of helping others avoid problems it is important to recognize how many variables are in play at once in a new design. It is often faster developmentally to break a design into staged milestones implementing one and only one variable at a time. I wish I had gained this wisdom by being smart and avoiding the traps I’m talking about, but sadly I’ve learned these lessons through scar tissue and stupid mistakes.
—
Bill
One novelty at a time. That’s good advice Bill. I am certain that had I opted for a more conventional rig then Roz would have been a success.