Like a mad scientist or any other inventor or creator, i am REALLY reluctant to just yet share out my Delftship-originated model which i now attach as a VCP file.

There are two primary layers/branches:

One is for the mesh
One is for the curves

The mesh layer is the 3D DXF output from Delftship. Notice how NICELY better than s..c..e..x those meshes conform to the stations in the other branch.

Why on this Earth would i want to decompose that to surfaces, which in VC and VCP produces hundreds of facets, only to restitch them again, and then have them convert to semi-ruled approximate placements and have new irregularities introduced into my model. If it looks so-so now for quick and dirty rendering at THIS stage of development, it is going to be horrendous if i recompose it by hand.

Out of curiousity, i endured a lot of frustration with AutoCAD 2008 last night to modify the polylines. Smoothing them out, AC 2008 shockingly and beautifully reduce hundreds of facets down to about 37 or 38 or so. It then was worth exporting back to VCP v856 and playing around. But, still (no, i didn't try to again use PolyCAD, as since that last 3 weeks ago attempt i just gave up again) ability to in ACAD to convert those faces to easy-to-use VCP-touchable faces. So, in VCP, i had turned about 20 of them into surfaces and then panned the drawing around. Not what i needed nor wanted.

Now, given the full size of my model and noting that i have had to by hand cut those waterlines at the bulkhead lines where the bulkhead thickness will be, it can start to become clear to some that cutting or segmenting the waterlines at their intersections with the bulkheads/stations produces some improperly-cut lines. This leads to a whole lot of manual movement of endpoints that i DO NOT WANT TO MOVE, even at the tiny inch/millimeter gaps being considered.

For what it is worth, i imported the model into AC2008, BOY was it a ROYAL PITA to select very specific stations and waterlines and expect to "just show me THESE selections" without the torture of Autodesk's implementation. I think it's easier in AC2009, or 2010 (on my work desktop), but i am not installing those on my personal laptop. Zooming and flipping the drawing with selected geometry in VCP shows the selected geometry even after the visual movements are completed. Not so in AC2008, which infuriates me beyond imagination, and which some people seem ok with. Not me.

But, cutting lines and panning to clean up the view in both apps led to some y-axis irregularities. Z and X were always fine, but ACAD gave varying decimal y-axis values whereas VCP was much closer to the values i used in Delftship. But, both apps seem to fail on cutting the waterlines at intersections of the station, and i concede this may greatly be due to my manual fairing of the stations and waterlines. So, they will obviously produce some non-smoothness. With the fully-automated version of Delftship, this would (i presume) be fully resolved before export to DXF even happens.

However, i do not want to use VCP for fairing/smoothing of the curves/lines/waterlines/stations because i need the ViaCAD model to be as faithful to the Delftship model as i can keep it, even with the fundamental manual fairing errors brought into VCP in the DXF file.

So, having said all this it is tear-jerking, heart-rending, mind-bendingly painful and saddening that i cannot natively in VCP convert those beautiful meshes of the hull into surfaces that contour-match the mesh.

I could convert the mesh to surfaces, and then at convenient locations join those surfaces, but it introduces a lot of nodes when i have to cut the surface for, say, overboard discharges. Walking away from that option, i tried various sweeping tools in VCP to arrive at manual surfaces that approximate my mesh shape. Given the curvature in the polylines (wait... do i need to do line conversions before attempting to make surfaces) and that there are gaps, i end up with these surfaces that do not match the mesh. Because the segment tool is not visually segmenting at the exact xyz coordinates (yes, x and z are apparently good), they y causes problem.

Please note:

-- I removed the rudders from the mesh model because the 480 objects caused an upload problem.

-- to visualize the model as one, just import one into the other

-- the mesh models colorized meshes do not divide where the curves model's stations do. That is something going on in Delftship, probably based on the curves and intersections i manually created.
Have fun...

I think i will run the model in Delftship again, but in millimeters instead of inches and fractions. Hopefully that will help the Delftship Free export to DXF 3D Polylines and the VC/VCP import. I'tll probably take a month by hand, and i'll undoubtedly remove some "dishing" in the resultant meshes/surfaces. I'll also have to even more closely make sure that there is positive curvature as much as possible.

Have fun with the hull model, but be advised you are not to claim it as your own. I'll eventually publish it in greater detail. Of course, due to US Copyright rules, there will be limitations i just imposed upon myself by publishing this part of the model. However, i expect and hope that *ANY* use of the model comes with disclosing the source (me) and that anyone using it uses it for personal, not monetary gain. That's about all i can be cared to bother with. Future models will be deferred for my own personal or monetary gain, so i'll be reluctant to publish in whole until i've filed the proper papers...

My home country adopted decimal currency in the late 1960s and metric units of measurement in the early 1970s, and people like me who were kids that grew up in that time of change are natively "unit-bilingual", in a manner of speaking. Imperial units are frustrating, arcane, inefficient and counter-intuitive. Congratulations on your epiphany.

I am happy to say that i am about 98% done with the hull surfaces and trimming the deck plating against the surfacing. The watertight bulkheads and the fuel and the ballast tanks are cut and i have my volumes. That was the hardest part, setting up the constraints of my model's working space, which i defined in DelftShip.

Nicely, the siding and deck plating file size is about 23.7 MB, whereas months ago whatever i was doing ballooned the file to past 85 MB. Deleting lines and having only solids and surface saves about 3 MB. This was about 3 weeks or so of working about 4 hours a day here and there. If i had tens of thousands of dollars, all this would have taken about 2 hours to generated each compartment's boundaries.

My next task will be to start sweeping sideshell stiffeners, deck girders, transverse beams, stanchions and more. I showed in some of my earlier posts what some of this will look like. Now, my current model (the one i hope i can copyright) should be sufficently different (at a technical level) enough since the linework is different (deck levels, hull beam in areas, better and cleaner fairing of the hull, all of which removed tons of steel and tens of thousands of pounds of resistance, and resulted in a slenderer hull... so, if any of you like the previous posts' images or files, have at them...) that i should be able to claim the ultimately finished drawing (what, 3 years from now?) as original, unpublished work.

I think moving forward after the hull girders material, things should start to become easier to model since all of the major compound curves items are almost behind me, well, unless i want to try to model the propeller...