Modeling boats in Freeship-The techniques of Using the sofware

[mick_allen]

Why are you using a separate panel to create your top instead of extruding the edge and shaping from there?

To try and clearly show the intention. The issue is that change has an effect.

The statement to be made again is that the same effect happens whether by connection (method shown), by extrusion, or by placing the 2 points on the ridgeline and making the cabin surface between.

As well, so called 'soft points' don't exist. They all (even 'locked points') have an effect like a magnet upon the surface or surface edge. But unlike a magnet, they have greater and greater effects the farther they are from the surface or surface edge.

Just to be clear to Frank and others:
Most of what I have written is NOT a criticism of the program. One has to recognize that a 'corner' is a singularity, but the program behaves essentially as one would hope: but you just have to record where you were, add a few more points and connections and massage the model back to whatever tolerance works for you.

The utility of joined faces outweighs separation by far. It just takes some adjustment.

[frank smith]

Cool , I'll work with that a bit

[lewisboats]

by soft points I meant points that aren't cornered. I'll have to play with it a bit more...maybe I'm missing something.

[mick_allen]

Here’s another look at the same example:

This is the situation that is wished to be retained. Simple hull, lots of curvature, few points. No hull change is desired.
Fig 201


The display only will now be changed to isolate the steps about to happen. The gunwale and chine only have been exported and reimported as ‘markers’ for reference. Stations on the isolated panel are shown so that surface changes may be seen. But nothing in the model has been changed:
Fig 202

Say for some reason, I place a control point in the chine curve by using ‘create new control points by splitting a control curve in two’, one would not expect much change as the arrow points to where the point will be placed and this point is really quite close to the chine curve. (no movement will be given to the point, the program will just be asked to place it between the two points c1 and c2)


But this is what really happens:
Fig 203

The chine curve has been displaced over a huge portion of the whole curve with the deviation being some inches away from its original position and the maximum deviation is not even where the new point is added! As well, from the station lines, one can see that the panel is dished somewhat towards the middle. One simple, seemingly inconsequential point has consequences.

***
So let’s step back to figure 202 again and instead of placing a new point the chine curve, place it into the existing control edge between point 2 on the gunwale and point c2 on the chine:
Fig 204

This placement (no movement has been done) does not affect the chine or gunwale curves (edges) but does bulge out the middle of the panel.

***

So then let’s go back the original position in fig 2 and add a gunwale point:
Fig205


This simple addition has caused the same sort of situation as that of the chine point addition: The gunwale curve has been significantly altered over more than half its length, the major alteration is away from the point addition and nearer the region of high curvature, and the panel has been dished.

The above has just been with normal 'uncornered' control points.

***

If you add points of any kind near areas of high curvature there will be significant effects. If the original shape is important, take steps to prerecord it so that the base position is not lost.

[Martijn_vE]

While trying to understand this phenomenon one has to be aware of the following:

• Points on the interior of the surface behave differently than points
  on the boundary or on crease edges. The crease or boundary edges
  will be closer to the controlpoints than interior surface points. Or To
  use your terms: The magnets of crease points are stronger than interior points.

• Further more, the algorithms for the subdivision surface adopted
  by the program are optimized for triangular and quadrilateral faces.
  By adding a 5th point on your quadrilateral face twist is introduced in
  the surface patch. This is clearly visible in the surface of your side panel.
  To avoid this you should also insert points on the chine and contour
  line, and connect these new points with edges all the way to the
  center line

[mick_allen]

Points on the interior of the surface behave differently than points
on the boundary or on crease edges. The crease or boundary edges
will be closer to the controlpoints than interior surface points. Or To
use your terms: The magnets of crease points are stronger than interior points

I regret using the term magnets: I think a better allusion is that of rubber bands from the control point to the surface or edge - the farther away, the more effect and vice versa. The demonstration is to show that even single point additions do influence surface/edge behaviours.

. . . . To avoid this you should also insert points on the chine and contour
line, and connect these new points with edges all the way to the
center line

Note that I have not shown solutions but just demonstrated the initial issue with a corner, or on a crease edge, or on a boundary edge, or on an interior edge.

Solutions involve adding more points in the manner you state, and then adjusting every one of them in order to approximate the previous location of the surface and of its edges.

[gonzo]

Does this program accept a table of offsets?

[yipster]

fef file
surface
carlson
chines
carene
vrml
polycad
offsets

quik download at: http://www.delftship.net/

[frank smith]

I have been adding some bulk heads to a design , and now my nicely fared hull
is all kinked up. I know that makers can be imported , but how do I set markers to a new hull design to keep tract of where I have been .

Frank

[Martijn_vE]

There should be a menu option in the intersection window that will convert any available stations, waterlines or buttocks to markers.

That way you can visualize the difference between any modifications you made to your hull

[frank smith]

I see that freeships can produce dimensioned bulk heads as well as
interior furniture . Dose any one have any experience with the accuracy of
the expanded panels? Could I design an egg crate like structure for a multi chine ply boat an have it work?

And is there an easy way to reduce the bulk head by the hull thickness?

[Rick Willoughby]

Originally Posted by frank smith

Does anyone have any experience with the accuracy of the expanded panels? ......

The accuracy of developed panels is limited by the accuracy of marking and cutting them. Although panel thickness has to be catered for separately if you are using thick panels.

If you have thin panels that are machine cut they will fit together neatly.

Rick W