Some time ago I developed a process for Sundance Balloons involving a lot of scripting and some modeling procedures for patterning hot-air balloons. It handles distorting artwork to look correct when inflated and the details needed to output for CNC cutting like seam allowances, labels, alignment marks, and how that differs depending on whether the artwork is made by splitting up a panel into pieces or with a second(or third) layer of fabric sewn on top.
The system enables me to crank out patterns for a simple graphic(for a balloon model I already have set up)in as little as an hour, replacing a fantastically tedious manual process that took weeks. This one took a bit longer than an hour. A typical balloon pattern might wind up with about 600 or 800 pieces to ultimately sew together, this has 2000, with random colors that made each one unique. Heavy use of Rhino’s Universal Deformation tools was made to adjust parts of the design to more or less maintain their proportions as the circumference of the balloon varies.
Very impressive! You allude to things that the average person wouldn’t even think about when seeing the finished product and which even designers would need to be reminded of (like you just did in this post).
Maybe a detailed discussion of the issues and the (manual) techniques to solve them could be another Rhino class or tutorial?
Oh I haven’t worked on any ‘unusual’ shapes, there are people here who do different sorts of inflatables.
My guess from looking at “wacky” balloons is that most of them are actually a regular balloon with extra added-on envelopes. I would model such shapes for production as a series of flat panels, knowing what’s gonna happen when inflated.
Well what I’m doing is putting graphics on an existing balloon design, not making a new one. The 2D patterns are a given, and every one I’ve seen follows a general pattern of how the panels are arched at the top and bottom (or not) to create the particular shape–more or less, it’s not exactly rocket surgery. So I have to make an approximate 3D shape from those in order to correctly apply the graphics, then transfer those back to the 2D patterns, and do the patterning tasks. I came up with an elaborate proprietary procedure in 2010, it’s probably possible to do a lot more simply today.
My impression of how more complex 3D shapes are made is that…you do prototyping! There are others with more experience with those.