patmat41959

[patmat41959]

So here’s the second most important thing for DS users to know:
Lines imported from a lines drawing DO NOT create a hull true to those lines!

First most important? It can be fixed.

The developed hull shape does NOT go through the control points, which are input from the original lines drawing… they just INFLUENCE the shape. Fine except it’s not, because the developed form no longer matches the original. There must be a work flow that gets the generated shape closer to the desired form.

Indeed the DS developed hull cuts corners, it’s always well to the inside (concave side) of input curves, except at the “corners”.

Imagine you have a magnetic rubbery sheet pegged by its 4 corners to the floor, exactly where you want the corners.

Then you hang a bunch of magnets from the ceiling, and each magnet is hanging exactly where you want the sheet to stretch and rise to, defining some sort of hill (I’m picturing Richard Dreyfuss building Devil’s Tower in his living room).

But while the sheet is stretchy, it’s not THAT stretchy, and it hovers over the floor but doesn’t quite reach the magnets. Oops. That’s what we have in DS, with the developed hull not reaching the input points.

 

How to fix it? Just keep raising the magnets until the sheet rises to the proper place.

In DS, I did this, though there may be more efficient ways:

1. Import the control points once to define the control mesh and its resultant malformed hull.

2. Import the control points a second time as “marker curves”. Markers are static curves, immutable, and they serve as references of where the original control points and desired hull sections are.

3. In an end view, compare the developed body sections (which are wrong) to the desired shape shown by the marker curves. Now start moving control points sideways or vertically to slowly pull the body sections towards their markers. Yes, this modifies the original control points and mesh!

4. Note that this is iterative… moving one point affects the adjacent surface faces, so you’ll need to sweep back and forth over all the points to get everything close.

5. Also note that the four edges of your rubber sheet between the corner pegs were also pulled off the floor when the magnets started lifting the center. Same thing in DS- the edges at deck edge and keel also have moved out of place. So the adjusting exercise needs to address those too.

 

I did this for my tugboat hull, and while it still needs some adjustment, it’s much closer to the desired shape now (below).

 

More: While you can get close by doing this, you can also turn on curvature displays to better smooth (fair) your new shapes. I haven’t done this yet, may not need to for my present purposes.

More #2: The Pro version of DS has an “automatic fairing” tool. I’m not sure how this works, but it could be of help.

 

Here’s an end view showing some of these features:

White: Original control points and the control mesh with its straight Tinkertoy connectors. The control points lay on the proper body sections from the lines drawing.

Green: Body sections cut through the DS developed hull surface… they miss by a lot. Even at the deck edges.

Pink: Nice curvy marker curves that DS draws for the second import of the original control points. These markers are the target, I need to move the green lines to match them.

To move any one point: Click on it, and you can enter new numbers or use the little arrow to move the point by set increments, very handy. Move each and every one this way to drag the developed sections around.

 

• This reply was modified 4 months, 3 weeks ago by patmat41959.
• This reply was modified 4 months, 3 weeks ago by patmat41959.
• This reply was modified 4 months, 3 weeks ago by patmat41959.

Attachments:

You must be logged in to access attached files.

[patmat41959]

Yes, something is off, but I cannot make it go away!

As I want the elliptical shape, I will just do it the hard way… or in another software…

Thanks.

[patmat41959]

Here is the cambered deck generated at Faces = 4. The control point mesh extends past the centerline, but not as far as where the starboard deck edge should be. The mesh seems to be curved on the starboard side only? And at the fore/aft ends, the centerline is high, as if the absolute value of camber height at midship (at 2% there) was applied all along the length.

Attachments:

You must be logged in to access attached files.

[Author]

Posts