When I do the same Flow Along Surface operation in Rhino 7 or 8, I get different results than I used to in Rhino 6. I use the same command line options in each one:
Copy = Yes; Rigid = Yes; RigidGroups = Yes; Contstrain Normal = Yes; AutoAdjust = No; PreserveStructure = No.
Wow, that’s a lot of options. And will be difficult to describe in class to students who are just learning Flow Along Surface. Still, I look forward to the update. @Gijs can you please post again when it gets into a Service Release?
Mike, FYI I’d suggest using _Flow TargetSurface for all stone layouts in your courseware. It covers 99% of the use cases with much better fidelity – your fishtail cutters will flow perpendicular to the curve, and the stone spacing uses physical distances (0.25 mm → 0.25 mm).
With _FlowSrf, those are somewhat random – your cutters follow the UV directions, and the stone spacing uses the mapped domain space (0.25mm → -0.25 mm to +0.75 mm).
The only time I use _FlowSrf is for pave, and I hate every second of it.
_Flow measures the distance at the curve, so it’s best if your flow curves are located at the girdle. For my small melee, I target a girdle depth of 0.35 mm. To get it to work with History, I take an extracted iso and use _OffsetNormal -0.35 to get my flow target.
The only difficulty of using _Flow TargetSurface is the surface normal. If you make a surface where the normal is inverted, the stones will flow culet up. The easiest way to fix it is to invert the stones on the base curve.
The harder way is to practice creating right-side-out surfaces from the start. Rhino doesn’t make it easy…or even possible.
My display settings color all backfaces purple, and I pay attention to the order in which I select inputs to surface commands. A general rule of thumb is to start with the curve closest to 3 o’clock in the world XY frame and proceed counterclockwise.
Thank you for the suggestion! I am a big fan of _Flow Target Surface. We have yet to implement it into our curriculum. Maybe this change will help move us along.
I did not know about the OffsetNormal command! That is super useful. I’ve been modifying the spacing on the base curves (or surfaces) to adjust the spacing on the target. Thank you for sharing this!
@EricM How well does OffsetNormal work when you have graduated stones? I’m thinking something like channel setting where you want the tables to be flush but the girdles are at different heights. How do you handle that situation?
You might be overthinking channel-set stones. My setters sand the channels down to the tables making them defacto flush. You could align all of your tables on the base crv and set the base crv and _OffsetNormal to the average girdle depth. Then bump your spacing up a hair to account for the variability.
Even with your girdles aligned to the flow crv, that accurate spacing only exists at the curve. Your girdle to girdle spacing will vary with problem areas. An average depth would increase that variability slightly, but it’s far better than using _Flow/Srf at the surface:
But that only works for rounds with nice makes. The table heights are all over the place with single cuts, native, baguette, princess…
Do you show them how to calculate spacing tolerance? Like my 0.25 mm comes from having very expensive sieve plates ensuring melee is at most +0.10 mm and knowing my spacing can wander ± 0.04 mm with _Flow. That gives me a girdle distance of 0.01 to 0.24 mm.
My setters use brass pushers in the GraverMax so rounds cut their own seats, and I can’t remember the last time they chipped a stone. But I do remember the occasional cracking noise back when I had cheap sieves and used _FlowSrf or _Flow without _OffsetNormal. Using 0.35 mm wasn’t enough. Even with everyone being hypervigilant, if something can go wrong it eventually will.
