Fading chine hull in Rhino

[Ben B]

I am designing a dinghy with a chine that fades to round hull as it goes forward. I'm working with Rhino and I would like to know if anyone has a good way of modeling this. One Surface, Two, Three, variable surface fillets,....

Any suggestions would help. Tools or technique

Thanks,

Ben

[DCockey]

A proceedue which works for me:

- Start with four surfaces per side of hull: upper forward, upper aft, lower forward, lower aft. The split between the forward and aft surfaces corresponds to where you want the chine to completely vanish. Adjoining surface edges need to match in position. Also the adjacent corners of the surfaces need to match in position.

- Use "MatchSrf" to match curvature of the adjoining edges of the upper forward and lower forward surfaces.

- Use "MatchSrf" to match curvature of the forward edge of the upper aft surface (select first) to the aft edge of the upper forward surface (select second). DO NOT use the "average surface" option.

- Use "MatchSrf" to match curvature of the forward edge of the lower aft surface (select first) to the aft edge of the lower forward surface (select second). DO NOT use the "average surface" option.

- Use "MatchSrf" to match POSITION (not cuvature) of the lower edge of the upper aft surface and the upper edge of the lower aft surface.

You may need to go through the "MatchSrf" steps a second time to ensure everything matches properly.

The forward upper and aft upper surfaces can be merged after matching if desired. Likewise for the lower surfaces.

[bhnautika]

Ben B another way of getting to a simple hull shape like your dinghy shape that is editable, is what I call control sections. I have supplied a file of what I mean. I start with a keel line to temporally anchor my bow shape, mid section(s) and the stern. For this I use “control point curve” for curved lines, with the minimum number of points to get the basic shape I want or a “polyline” for straight sections or a mix of both. The object is to get a continuous line from keel to shear (blue lines in file). If you select one of these line in the non surfaced model then turn on the “edit control points” you will see I only used between 2 or 4 points for any section. Once I’ve created this grid I go to the “record history” (right click on it first to make sure the “update children” is ticked) and left click to turn on, then go to the “loft” command. Select the bow then section(s)then stern, in the loft options pick “tight”. This now gives you a surface that is connected to the lines you started with, just select the one you want and turn on the “edit control points” you still have the same number of points as before that can be moved around and the surface goes with it (it a good idea to turn off the “osnap”so things don’t jump around unexpectedly)

[Ben B]

I've tried both techniques and just can't seem to get the hull to come out right, chalk it up to Rhino inexperience.

I've attached the hull file so you can see what I'm trying to do. I want the chine near waterline to disappear just aft of amidships and the shear's hard edge, "chine", to disappear near the bow. I've used 3 surfaces and I've done a lot of control point "massaging", but I still don't think the surface is very fair. Any Ideas?


Also I'm using Rhino OSX, and it still has a lot of "bugs" e.g. Record history doesn't work.... yet

[Joe Petrich]

Can you save the file as Rhino 4.0 I can't open the one you posted.

[DCockey]

Revised so deck has continuous curvature. Attached file replaced.

I developed a hull using 4 surfaces. See attachment.

- Imported Ben's model. Saw that the one set of the control sections didn't intersect the chine at ends of the control section curves.
- Copied his sections and stem curve to new layers
- Created new chine curve using InterpCrv
- Created new "keel" curve on centerplane using InterpCrv
- Copied Ben's sheer curve.
- Extended sheer curve copy forward past stem. Extruded sheer curve copy on both sides. Found intersections of extruded surface with stem curve and first control section aft. Split stem curve with extruded surface. Split first control section with extruded surface. Joined upper split part of first control section to first control section deck curve. Refit resulting curve to tolerance of 0.1 mm.
- Created new sheer cuve thorough upper ends of control section sides and stem.
- Noted that the forward side and bottom sets of control curves only have tangential continuity, not curvature continuity. Assumed that tangetial continuity was sufficient for forward portion of hull.

- Used NetworkSrf command to create four surfaces.
- MatchSrf for Tangential continuity between forward side and forward bottom.
- MatchSrf for Curvature continuity of aft side to forward side
- MatchSrf for Curvature continuity of aft bottom to forward bottom
- Match Srf for Position continuity of aft side to aft bottom (to ensure no gap due to previous two MatchSrf commands.
- Copied hull surfaces to new layer
- MergeSrf to merge sides surfaces
- MergeSrf to merge bottom surfaces

- InterpCrv for deck centerline curve
- InterpCrv for cockpit side curve
- NetworkSrf for forward portion of deck
- NetworkSrf for aft portion of deck
- MatchSrf for curvature continuity between forward and aft deck surfaces making sure "Preserve isocurve direction" is checked to preserve the kink at the front of the cockpit
There are some problems with the forward portion of the deck. An alternative arrangement of control curves may work better. I'll experiment.

[CaptBill]

Originally Posted by Ben B

I've tried both techniques and just can't seem to get the hull to come out right, chalk it up to Rhino inexperience.

I've attached the hull file so you can see what I'm trying to do. I want the chine near waterline to disappear just aft of amidships and the shear's hard edge, "chine", to disappear near the bow. I've used 3 surfaces and I've done a lot of control point "massaging", but I still don't think the surface is very fair. Any Ideas?


Also I'm using Rhino OSX, and it still has a lot of "bugs" e.g. Record history doesn't work.... yet

T-Splines is a component add on for Rhino. It implements fading chines very well. This is probably the most complicated thing to do in 3d right. (not sure they have OSX version?)

I believe they use the 'crease' function on this copter tutorial.

[DCockey]

I find it very beneficial to make sure the curves I'm using to construct a surface are fair and as intended before creating the surface. It's alot easier to fair curves than sufaces by moving control points.

If a surface has more than 7 or so points in one direction its very tedious fair if the corrections are not very local, and I find I'm likely to introduce other problems.

If a surface with many points needs fairing it may be helpful to rebuild it before trying to fair it by moving control points. Two ways to do so:
- Rebuild, input how many points in each direction, check deviation from original surface with the Preview option, then Calculate option.
- FitSrf, specify the maximum deviation from the original surface and it simplifies as much as possible staying within that deviation.

SoftEditCrv and SoftMove can be very useful when making local shape alterations.

CageEdit is very powerful for global "distortion" of a surface. The trick is to pick the right type of control object with the right number of points based on what you want to do.

[DCockey]

How does the attachment look?

Hull is the same as before. Deck created using Sweep2.
- Copied deck curves to DC Deck Lines Rev layer (to keep the old ones for comparison purposes)
- Joined foredeck centerline with curve which was top of stem- Point at forward apex of foredeck
- Deleted the two forward deck sections
- Created foredeck with Sweep2 using centerplane curve and sheer as rails, point as first section, deck section at front of cockpit as second section
- Created side decks using Sweep2
- MatchSrf with Preserve isocurve direction and curvature continuity to match fore and side deck surfaces
- MergeSrf to create one deck surface with Smooth=No to preserve kink at front of cockpit

The appearance can be improved with a better mesh:
- DocumentProperties > Rhino Render > Mesh
- Click on Smooth & slower
- Click on Custom
- Set Maximum angle to 5.0
- Set Maximum aspect ratio to 6.0
- Made sure Refine mesh was checked

[Ben B]

Hey Joe I've attached a Rhino 4 version.

[DCockey]

What improvements are you looking for?

Two problems I see are the edges of the deck and hull at the sheer do not coincide in the forward portion, and the deck comes to an acute point rather being rounded. How did you generate the deck surface?

In DocumentProperties under Units you might try changing the "relative tolerance" to 0.1 percent or even 0.01 percent, and the angle tolerance to 0.1 degrees or smaller. Those changes won't help the existing model but may improve precision for future work.

Also adjust the Mesh parameters as I described above. It will improve what you see.