HELP - surface trouble

[ld.power]

I’m having some trouble creating a surface, it looks simple enough going off the lines but is proving rather tricky. I first tried rail revolve - no good - as you kind of need two rails. I then tried sweep 2 but comes out very messy. I played about with loft and tried creating a top surface that would follow the dip but wasn’t really getting anywhere. I then went into gumball and control points pulling up the sides from a flat top but again wasn’t coming out very nice.

Any help would be greatly appareciated!! File posted

BIG SE DISH.3dm (58.0 KB)
Render is of a prior iteration that has a flat top. The current design uses raised sides as seen in the lines. (image for referenece)

[pascal (Pascal Golay)]

Hello - if I undertand what you want - this would be the process:

Two curves > extrude these > BooleanSplit the solid with the wavy thing and delete the bit you do not want, or BooleanDifference (need to mind the surface direction of the wavy one - see Help on Boolean operations, or Trim and Join.

-Pascal

[mdesign (Marcin)]

BIG SE DISH2.3dm (295.1 KB)

This was bad:

Top surface made with Xnurbs plugin.
The bottom is the flat plane trimmed to the curve
Side is Sweep2 command

Fillet or blend should be done as of last.

Edit: Oh sorry I haven`t seen that you wanted to have done it a bit differently. So Pascal method was maybe a way better for you (in my way middle of the top is pushed down).

[ld.power]

Hi Pascal,

Thanks for checking out this issue! I don’t think my drawings may have been clear, there is actually a dip in the middle going lengthways (in front window). I have drawn out some rough guides (green lines) to give reference to the form. Please see the new attached file. Also please could you export in rhino 5 - much appreciated

Many thanks!
BIG SE DISH.3dm (43.7 KB)

[pascal (Pascal Golay)]

Hello - is this it?
BIG SE DISH_Maybe.3dm (148.8 KB)
@ld.power V5 now.

-Pascal

[ld.power]

Hi,

Sorry can’t view it, could you please export it in rhino 5 version - thanks!

[ld.power]

Hi,

No, thats correct. The middle is supposed to be pushed down. The form is supposed to dip in the middle. it’s essentially a shallow tray, the sides scoop up a big higher than the ends.

I’ve drawn some rough guides (green lines) to give reference to the form.

BIG SE DISH.3dm (43.7 KB)

[ld.power]

Hi Pascal,

I don’t suppose you have that earlier rhino file, rhino 5 export, for me? Would greatly appreciate!

Many thanks

[pascal (Pascal Golay)]

@ld.power the file I posted above should be V5 now …?

-Pascal

[ld.power]

Hi Pascal

Yes! This is the form I’m after. Thanks very much for your help with this. Can I please ask how you did this - what were the steps?

Cheers

[pascal (Pascal Golay)]

Hello- instead of working from these curves directly:

I would make the big surfaces first - the top and sides - so they meet at a hard corner, then add the fillet as a separate operation.’

First, fix those curves - the inner parts are folded back on themselves - explode and make that bit a line.

Bypass the fillets at the outer ends:

Extend Type=Natural), followed by Connect, works well here. Connect by itself will as well but makes a more complex curve from the top one.

This no longer lines up -

I just moved a bunch of control points at that end of the perimeter curve by hand with Point and End Osnaps:

Set this up on all four quadrants - as I recall the curves you have may not be symmetrical, quite, so I mirrored one set and adjusted the perimeter curve, if I remember right.

Loft the inner lines:

Sweep2

Loft for the sides:

FilletSrf at R=2:

Edit: I’d probably prefer a PlanarSrf for the middle, starting with an InterpCrv through the end points of the lines, maybe adjust the shape to make it a bit more rectangular and less rugby ball, to be more like the perimeter by InsertKnot > Automatic then point edit and Scale1D to get it back on the end points:

-Pascal

[ld.power]

Hi Pascal,

Thanks very much for going through this - greatly appreciated as always.

Cheers