rhino unroll prob's!?

Discussion in 'Software' started by tugboat, Sep 20, 2013.

  1. tugboat

    tugboat Previous Member

    I am just checking in to see if anyone else has had this issue with rhino- when I use the unroll command on a developed surface- even though I might have gone over my basis curves till my head is spinning- I still cant get most surfaces I do to unroll to what v.5 states as "within tolerance"
    usually it tells me
    " the surface area is (.15%) larger when unrolled...
    or conversely "the surface area is (0.02%) smaller when unrolled...


    of course I've used some arbitrary figures in the exanmples but those figures have come out in my models...

    :confused::confused::confused:

    I don't get - it unrolls?

    ok- so what's the tolerance?


    does it have to be within 0.01% or can an object be developable if it is within a higher tolerance like 10%???
    I have talked to a couple people at Rhinocad but no one seems to understand the issue well.

    I have gotten different answers on this from Rhino-
    - some say it will build fine others at rhino say it won't...most at Rhino don't seem to not even know what to make of it...

    anyone else having difficulties getting perfectly developable hulls in rhino v.5 in say steel or plywood? i.e. single chine or multi chine designs?

    my big question is what is going on here when it unrolls???:confused::confused::confused::confused::confused:
     
  2. DCockey
    Joined: Oct 2009
    Posts: 5,228
    Likes: 632, Points: 113, Legacy Rep: 1485
    Location: Midcoast Maine

    DCockey Senior Member

    Link to a discussion about UnRollSrf on the McNeel Rhino3D forum: http://discourse.mcneel.com/t/unrollsrf-broken/493

    How are you creating the surfaces you are trying to unroll?

    I do not have any problems with surfaces unrolling but that may be the result of how I build surfaces. I usually use DevSrf with the least twist option, and extend the defining curves where needed so that the entire region of interest is covered by the surface. Then I trim the surface as needed.
     
    1 person likes this.
  3. tugboat

    tugboat Previous Member

    Hi Dcockey- you are a far better draftsman than I. I think I've seen some of your work and its good.

    Ill have to look into devsrf. never heard of it- but there are so many functions in rhino that I know nothing about...
    how I was taught, was how MacNaughton does it, by building basis curves.
    Im sure there are quicker methods.

    I was trying to unroll a hard chine steel design...no matter what I do, I cant get it lower on the bottom section than .02% I am wondering- what the heck the tolerance is set to?

    does it have machinists tolerances built in ?
     
  4. rwatson
    Joined: Aug 2007
    Posts: 6,160
    Likes: 494, Points: 83, Legacy Rep: 1749
    Location: Tasmania,Australia

    rwatson Senior Member

    Do you FAIR your hull lines before creating your hull shell ? So many times a perfectly good looking line is unfair, and throws a spanner into the works.
     
  5. tugboat

    tugboat Previous Member

    absolutely...I have faired them so much that my eyes are buggy.

    its has no trouble unrolling and even in Gaussian analysis it shows good curvature on the mean and auto values. but it reads 0.2% difference from the model to unrolled surfaces...im wondering where the line of margin is for tolerance? is it 10% or .1%?? to make the surface developable perfectly. apparently this is a bug in rhino- they increased tolerances to allow curvature to unroll which normally would not be developable.

    so Im wondering what the tolerance is set to. Im pretty sure my surface would work but I like that proof in the unroll...and nothing I do- not changing the lines- fairing them more -nothing seems to give. I even had a rhino cad tech guy fair my lines

    even HIS model taken from my lines and corrected still was out by 0.07%. (scratching head...)
     
  6. rwatson
    Joined: Aug 2007
    Posts: 6,160
    Likes: 494, Points: 83, Legacy Rep: 1749
    Location: Tasmania,Australia

    rwatson Senior Member

    The tolerance is discussed in the discussion link in the previous posts. It even has the developer discuss the logic behind the process.

    Its not a 'bug', its just a case of what logic they applied to the process.

    In any event .02% is .0002 of whatever he length is, which is .0052 of a foot over 26 feet for example.

    This is not much bigger than a wide weld in steel, and not a problem for your glass application.
     
  7. DCockey
    Joined: Oct 2009
    Posts: 5,228
    Likes: 632, Points: 113, Legacy Rep: 1485
    Location: Midcoast Maine

    DCockey Senior Member

    DevSrf is a "Lab Plug-In" which is available from McNeel (free) here: http://wiki.mcneel.com/labs/devsrf

    The interface is somewhat different than most Rhino commands so read the instructions. Unlike Loft with Developable option DevSrf doesn't create triangular surfaces at the ends of curves. Instead, it stops the surface at the locations along the curves where a developable surface is no longer defined by the pair of curves. The solution is to extend the curves and then trim the surface.

    A difference in area of 0.02% is good. But if you are using Rhino 5 and getting a warning message with 0.02% difference in area then there is probably another problem with your surface. Post the curves and surface here or send a copy to me and I'll see what I can do.
     
  8. tugboat

    tugboat Previous Member

    I owe you one amigo!

    the bottom plates at the bow from amidships are .12% but the sides are of course 0.03% not 0.02.- I made the curves more fair and it actually went out .01% more...

    ...it still should plate up- using templates.

    but it would be nice to have cnc cut panels for this...p.s. would it be worth it to insert a small radius at the chine and (since I know you are a capable draftsman) offset the radius so it ends towards the stern?..I would pay you for this service IF its worth doing?
     

    Attached Files:

  9. SukiSolo
    Joined: Dec 2012
    Posts: 1,269
    Likes: 27, Points: 0, Legacy Rep: 271
    Location: Hampshire UK

    SukiSolo Senior Member

    Rhino's developable surface is pretty accurate in my experience. Having unfolded aluminium aircraft wings it even shows with Gaussian garphics the bits that are stretching. when I have designed 2 chine hulls ie 3 planks per side it has been very good and allowed me to accurately cut panels. A very small ammount of deviation ie 0.02 is at least tolerated with notification. Some programs would simply say not possible. Ply is pretty tolerant and I would expect 0.01 to 0.03 maybe more depending on shape to be achievable in the real world. When I modelled the decks of one of the last dinghies it could not unfold them yet I was able to physically make them. Go figure! in US speak. I knew I could from a previous build but Rhino could not work that out in Ver 4.0.

    A big advantage in timber is it is a bit more tolerant of a small ammount of 3D curvature. The trick is knowing when it is too much, and I'm afraid only real world experience is your best bet. Malleable and stretchable materials like aluminium will also accomodate a certain ammount of tolerance. Steel is not quite as forgiving but I bet if you anneal it you can get a fair bit more bend in all 3 planes. We used to put light scalpel cuts in veneer to get 3D curvature and the glue would fill it.

    Try designing a sail in Rhino - that's quite a good exercise. Mostly it will unfold it.... but sometimes it won't. Now you will understand North Sails and others modes of creating 3D sails.

    I still think you should get a better result from the modelling ie it should unfold perfectly with a single or double chine hull unless you are distorting it grossly in the bow area. If so and you want to build in timber be prepared to build up the ply in veneers. A bit time consuming but it allows pretty difficult shapes to be achieved.
     
    1 person likes this.
  10. DCockey
    Joined: Oct 2009
    Posts: 5,228
    Likes: 632, Points: 113, Legacy Rep: 1485
    Location: Midcoast Maine

    DCockey Senior Member

    tugboat, please post the curves you used to create the surfaces if possible.

    The recent Rhino 5 Service Release 6 candidate has a modified UnRollSrf command with explicit setting of the tolerance in the command.
     
  11. tugboat

    tugboat Previous Member

    Dcockey, sadly I only created the aft sections- Nathan, MacNaughton's teacher in drafting, helped create the striped sections forward, I simply took what he did and redid the topsides and took out 4 inch flare in the topsides( flared topsides on a tugboat is a repugnant thing, an absolute "do not do" on a tugboat!!)

    but as for lines
    the simple way I did this was to just go to curve from objects, DupEdge, and did the the edges to form the sheer, the chine and the hornbeam stem and keel. I then manipulated the curves to form my desired shape and spliced it into the bottom section which I really liked. They are my lines from a file I sent for him to "look over"

    he just sent me a return corrected version that I did and could not quite get.

    I used "normal" for lofting. It simply does not "develop" using the developable feature...

    btw- I tried to use the devsrf program and the first time I tried it did create a surface- 0.03%,
    but in trying to further refine my lines to get "within tolerance"
    (after some hours) I got only 0.08% but nothing close to tolerance.

    how Nathan did the forward section is a enigma to me but of course he is a master draftsman.

    also im learning that I am a builder- not a draftsman. and more a sailor than a builder.

    Its funny since I DO NOT want to learn any more of this than is necessary to get on with my build- I had planned to actually get my ticket in design later but I actually find this very frustrating.

    getting things that don't unroll properly after hours of trying and retrying to get fair lines and doing my best I have to admit Im just not talented in the ways of drafting and design...but Im very OK with this understanding.

    are you able to just DupEdge it? it would be easier.

    let me know...if not that's how I would be sending you the lines...:)
     
  12. tugboat

    tugboat Previous Member

    when is the service release due out?
     
  13. rwatson
    Joined: Aug 2007
    Posts: 6,160
    Likes: 494, Points: 83, Legacy Rep: 1749
    Location: Tasmania,Australia

    rwatson Senior Member

    Hey, Tuggie If you want to email me the lines ( pm me for the email address ), I will have a play.

    I am about the stage in Rhino where I might luck out and spot something. I can certainly use the practice.
     
    1 person likes this.
  14. MikeJohns
    Joined: Aug 2004
    Posts: 3,192
    Likes: 208, Points: 63, Legacy Rep: 2054
    Location: Australia

    MikeJohns Senior Member

    Try rebuilding the surface, and also try splitting the surface into say 3 parts before unrolling it then join the unrolled parts together.
    I project the sheer and chine curves well past the center line and keep them the same length before lofting the surface, then trim to the mirrored intersects.

    Don't get bogged down with accurate tolerances if your surface are 20mm different in a few parts after rebuild for example they are still functionally identical and building tolerance is no more accurate anyway.

    What material are you using ?
     

  15. DCockey
    Joined: Oct 2009
    Posts: 5,228
    Likes: 632, Points: 113, Legacy Rep: 1485
    Location: Midcoast Maine

    DCockey Senior Member

    tugboat attached the .3dm file with the surfaces in post #8.
     
Loading...
Forum posts represent the experience, opinion, and view of individual users. Boat Design Net does not necessarily endorse nor share the view of each individual post.
When making potentially dangerous or financial decisions, always employ and consult appropriate professionals. Your circumstances or experience may be different.