Accruate Developable Panels

Hi All,

I posted this in the Boat Design section, but there was a general lack of response, so I hope you do not mind if I re-introduce it here.

I would like hear from people who have successfully used any application to, not only design developable hull panels, but to also flatten the design into accurately sized panels. Ply, steel, or aluminum does not matter, but I recognize they have different characteristics, and so experience with one may not apply fully to the other.

I have browsed this site and the web a fair amount, and while there is a lot about drawing developable panels, there is less about turning them into accurate flat panels and CNC data, and very little about the experiences, good and bad, of people who have actually built their own design, in particular by first cutting and fitting the panels (instead of using an oversized panel and triming it afterwards).

What I have heard, and even read on the McNeely's site, is that Rhino cannot reliably (and from what I know, easily) generate developable shapes typical boats, and even less reliably flatten them. Some hint that Freeship can, and Formsys says that workshop can do it to within 2mm.

Thanks in advance for any replies!

 

From what I know many people have used Rhino to create developable surfaces using the unrollsrf command. There is a plug-in, Developable Surface Creation, available at http://wiki.mcneel.com/labs/devsrf, the RhinoLabs site. I haven't heard anyone from this area complaining about developable surface creation in Rhino. What is the McNeely site? I'd like to know more.

 

By McNeely's did you mean McNeel's? I should have figured that out.

 

No, no, totally my fault... yes, McNeel

I must say I am suprised that there are no responses to the building experience side of this question... I am not complaining, its just that I assumed there would be more people on this forum with practical building experience based on CAD output... but then I am asking for experiences for a specific type of design and contruction...

Cheers,

Grant

 

All right Grant here my view, since you have been begging for comments.......:

Started some 25 years ago with Aerohydro's Fairline FL2B. Could'nt afford the full rollout version back then, but the lean version had a "Rulings test" facility, and I made a simple add-on plate expansion program, based on triangulation. With this, I could lift the rolled-out points into a CAD program (still using CadKey). Connecting the points via the CK's standard spline generates the cutting profile. I work exclusively with aluminium, one-off workboats for commercial customers.

For some panels, I don't even worry about "going CAD", but just print the list of the triangle legs for each panel. This way, the cutting profile is generated by simply marking all the consecutive triangles, one after the other, onto the plate; no X-Y transformation, simple and "exact enough".

Later, I have been using Hullform and others for plating rollout, but I always check a few critical dimensions "by hand" before cutting, as I have found that the algorithms integrated in the programmes sometimes select odd, or non-existing rulings for the calculation process. Generally, when everything is OK, both the classical triangulation and the integrated rollout sequences give a tolerance band of ~+/- one mm, which is accurate enough, since there are other disturbing influences during the building process, that generate greater faults; welding distortion to name one.

I also designed a few boats for a local boatbuilder/collegue, who used one of the other (don't remember the name) hull design programs. He ordered all panels ready-cut from his materials supplier; perfect fit, no trimming. Hope this gives you an idea of how it can be done in small scale, specialized production. Frankly speaking, the hull cost is not a high enough percentage in our low volume production; more advanced rollout-marking-cutting procedures do not pay off, instead we may loose control and contact with the real world.......So: keep it simple what is simple and focus on rational, efficient design!!

 

I dont know what size boats you are designing but I use Carene (from www.epoxy-resins.co.uk ) and then move the data to Freeship ( not the free Delftship which will not print ) and then I print out the developed panels which have all the mesurements marked on them . Everything has fitted well so far but I only go up to about 16 ft boats

 

I use Delftship for sheet development and it is accurate to my ability to cut it. Typically +/- 1mm. Usually good enough.

The only machine cutting I have done from Delftship files is milling foam blocks for male plug and a couple of aluminium propellers. These all gave precise results.

Rick W

 

Freeship/Delftship will do it easily...provided the design is viable for such a method. It is specifically designed to provide panels from steel so the program is rather strict in its interpretation of developable...but it is actually less strict than Carene 2008 in which there is no leeway given. Most CAD programs will provide flat sheet panels from a developable surfaces.

 

Seems that we generally have good experience from the various transfer methods, then. Must say that I am very interested in learning more about the Freeship after Apex1's eminent tutorial, but haven't had the time yet.

PandB: do you print your panels in full scale, or did I misunderstand you somewhere? What's your material and what kind of boats do you build?

One observation on precision then; in those programs, where you work with splines "flexing" over a few nodes (like battens), it is very important to have a GOOD smoothing, or fairing tool. You may very well have a longitudinally smooth surface or chine as long as only the nodes are dictating the shape, but the line may be corrupted when you add a set of transverse sections! Fairness has to be checked as local radius all along longitudinal strakes; any change to be smooth all the way.

Even small non-fairness will show up as a bad fitting panel; we blame the rollout procedure, but the origin lies in lack of fairness.

 

Excellent everybody, thanks a lot. Good to hear that these tools, and work arounds seem to work.

It sounds to me like the key to good flattening of panels is to have a good developable panel.. the more strick you contol the rule lines the better.

Keep the experience coming, I am sure we are all benefiting from hearing from each other. I am at least!

Thanks!

 

I use Maxsurf & Workshop.

Workshop shows the amount of strain in the panel so if the panel is not 100% developable, ie slightly compound, you can still flatten it and get a feel for how much stretch or compression will be needed.

The full version also allows you to run stringers, flatten them, and also generate the cutouts in the frames. Havent found anything else that will do this.

 

The ability to recover strains from a flat panel to be bent into a non-developable surface is very useful. If you know the yield strain of the material, you can tell if the material will buckle before you try and form it to the desired shape.

If the strain predicts buckling, you can use a thinner panel, or change the degree of curvature until the panel does not buckle. This can also be done with fabrics as well as metals, in the case of fabric, often "lock angle" is used instead of strain. The lock angle of a bi-directional fabric can be measured in a parallelogram type of Instron fixture, and pull tested until the fabric ceases to elongate.

In this fashion, composite layups over compound tool surfaces can be checked for whether or not they will lay nicely over the contour, or whether they will require darting or thinner layer thicknesses. Nice to know in advance of a wet layup.

The strain functionality is better, as it opens the door to more materials, such as metals or even rigid/semi-rigid core materials.

 

Rhino's unroll command works perfectly if the plate is developable. I usually run the command as I go just to test if my design is actually developable. The smash command that was introduced in Rhino 4 also works well if you absolutely have to flatten an undevelopable surface (it works better the closer the surface is to being developable). The percent changed is listed when you complete the command is very handy. I find in general if it is less than 1% (occasionally 2%) then you shouldn't have any trouble fitting the parts.

I'm a professional designer using Rhino at a firm to produce real parts. We have done very large projects with many different shipyards. Wood, steel, and aluminum. I've never heard complaints related to parts not fitting that were software related problems. User related problems - yes. I was told by the builders on the last project I did cut files for that they were amazed with how easy the pieces went together. Like lego was their description.

The most common user related problem is not allowing room for plate thickness or offsetting the plate thickness in the wrong direction. You always have to remember when using Rhino that it only does surfaces with no thickness. Real life plates of course have thickness. It takes some fiddling and thinking to make sure you trim the surfaces to the exact shape they need to be and you need to make sure you know which way the plates are going to go. Are you building from moulded lines? Are the plates joined on the inner edge or the outer edge? Will they be plug welded? How is it going to be faired? How much allowance is there for fairing? Etc. If you design it well and pay attention to all of the details Rhino will produce a perfect part. I always make a separate layer that I convert to solids (extrude surfaces, offset surface as solid, etc) and I check it by zooming in close and taking various section cuts. If the parts are meeting properly in the model they will meet well in real life.

 

Yes it does but what happens when the plate is not developable?

The beauty of Maxsurf/workshop is that even if the plate is not developable you can still flatten it and know exactly where there strain is, and what direction it is.

 

If the plate isn't developable you can use the "smash" command in Rhino.

Rhino is no replacement for a higher end engineering package. I understand that and we do use other software in our office as well. For the price I know of no other program that can do all Rhino can do.