Frameup

[middle_watch]

Following on from an earlier question re frames I came across a fascinating collection of 3D graphics models of sci-fi vehicles on the Internet. The creator had taken a bandsaw to his old Dinky toys and sliced them up. He then dipped each section in ink, made a pattern and scanned it.

By reassembling the frame sections on computer using a suitable CAD program he then had a perfect skeleton just waiting to be "skinned"

It occured to me that it should be possible to do the same with an old model kit hull to obtain the frame references to scratch build a hull.

(Casts a thoughtful eye on the POW kit Cabin Boy bought me Christmas)

[Francisco P. de Nanclares]

That is some really original way of thinking!!

[Mark Deakin]

Hi,

perhaps the Heller 1/400 Hood might be usefull for something after all

Mark D.

[Timmy C]

Of course, that's only good if the hull form of the kit is accurate, eh? What I don't get is how he transformed the scanned image into a file that CAD can use...

[Cadman]

These days you can take the hull and digitize it with a 3D scanner. This will give you a 3D model of any part. The problem with cutting a hull into frames is that the sections would tend to distort and give you an inaccurate section.

Skinning a model is a horse of a different color. Lofted surfaces are very hard to create with most software. While one can easily twist and shape a flat sheet to fit the frames of a scratchbuilt hull, doing it mathematically takes a lot of work, and usually results in gaps, and overlaps.

It is better to deal with a true solid model, rather than a surface model. Surface models are more suitable for games and simulations, while solids are ideal for just about anything that your machine can handle. Working with solids is like carving a solid chuck of material. You cut away material to create the final shape much like the way you machine a block on a mill.

[middle_watch]

Is it possible to take a model somewhere to be 3D scanned? It is not the sort of equipment most of us have in the living room!

Now a disclaimer, I am not a draugthsman, it just so happens I often have to work with them when they fry their computers, I like to have a basic knowledge of the various bits of software my Users play with, it helps when they start yelling it is the computers fault and not theirs!

I had a go at it last night using 3DS Max, but I cheated and used the frames already printed on a plan. I scanned the frame plan as a JPG and set it as the background image.

I then traced the contour of each frame line to generate a spline. Each spline was positioned at the correct distance and dotted with a dozen vertices (points on the line which can be used to modify the line or link to other vertices).

It was the work of moments then to link the vertices along the longitudal (fore and aft) axis and I had a skeleton hull which I transformed to a mesh (A solid surface) and applied a skin using the surface generator.

The hull shown hear is not accurate, in order to do it quickly I simply duplicated the central frames.



That was when things got funky, trying to fine shape the hull caused it to crumple just like real steel, you can see the effect near the bow, I am going to have to rethink the methodology! I will draw up the frames more accurately and have a bash. If anyone has already done this would appreciate any pointers!

Before people start to flame me for going off topic I would point out a 3D plan is got to be handy for the scratch builder, and if I can get this right it will also be great for figuring out the RC gear.

[middle_watch]

It is better to deal with a true solid model, rather than a surface model. Surface models are more suitable for games and simulations, while solids are ideal for just about anything that your machine can handle. Working with solids is like carving a solid chuck of material. You cut away material to create the final shape much like the way you machine a block on a mill.

Sounds cool, but have absolutely no idea how to. A tutorial I found showed how to chop chunks out using Boolean functions which I used to make the bearings for the shafts, pretty cool, but could not see how that would apply to shaping the hull.

Lofting - ERK! Our survey says Eh-Ahhhh! I just got a headache reading up on that. Well I have something to do at the weekend now, I presume you start with the first frame and loft to each in turn?

[Cadman]

Your hull looks cool, and yes there are services that will digitize for you. I will have to check my links at home. I have a good contact here for 3D printing, but not digitizing. I have always prefered to create my own geometry from the best plans I can find.

For solids, you should be able to do the hull as you have done this one. Then you can extrude the hull along the path of the splined frames. This will create a lofted surface. Better yet draw the frames as a closed spline with thickness and your hull shape will be a solid instead of a surface.

[middle_watch]

Your hull looks cool, and yes there are services that will digitize for you. I will have to check my links at home. I have a good contact here for 3D printing, but not digitizing. I have always prefered to create my own geometry from the best plans I can find.

For solids, you should be able to do the hull as you have done this one. Then you can extrude the hull along the path of the splined frames. This will create a lofted surface. Better yet draw the frames as a closed spline with thickness and your hull shape will be a solid instead of a surface.

Thanks! Will give it a whirl!

[middle_watch]

Ah I think I have it, when you said solid I was thinking the whole hull had to be a solid. But with thickness applied to the spline I get elements I can select around the face of the spline and extrude to form a hull skin. I like it, thanks. Have to do the work thing now but will play with it tonight.

[Cadman]

The main difference between a solid and a surface is thickness. A surface is like a cube made of paper, while the solid is like a block. The block can be shelled out and have very thin walls. A surface has zero thickness walls much like a vector line. These are more useful for visual effects than representing true geometry.

Solids are a little challenging to edit, but with a little planning you can create some elements that are extremely complex.

[middle_watch]

The main difference between a solid and a surface is thickness. A surface is like a cube made of paper, while the solid is like a block. The block can be shelled out and have very thin walls. A surface has zero thickness walls much like a vector line. These are more useful for visual effects than representing true geometry.

Solids are a little challenging to edit, but with a little planning you can create some elements that are extremely complex.

Erm, fink I got it wong again, Dad!

I sussed out turning the spline tracing of the bulkhead into a solid, on my system I added an extrude modifier (no option to extrude a spline without it) and then a shell modifier. Got a a really neat cylinder type effect. Traced them all, laid them out, then went to extend the extrudes to mate them.

Ah!

Each had a different number of faces and vertices, trying to weld them without getting a lot of diagonal lines, which I know translate badly when smoothing, was just impossible.

I am guessing now I should just extrude one then add vertices at the other bulkhead positions and position them to the right transverse points.

I talked to one of the CAD people I trip over and he told me boat and ship hulls are done on specialist software. Big help!

[middle_watch]

Okay, next try, I drew a spline for each bulkhead and made sure they each had the same number of vertex points. The vertex points linked fore to aft and a surface modifier added.

Getting there I think. the line on the aft midships deck is where a couple of vertices are not mated down the centre line.

[middle_watch]

Finally managed to do some work on this project after a long delay. Bound to say I am not finding it easy to control the smoothing processes, but after lots of trial and error I went back to basics and built the thing pretty much as I would a material hull. Finally getting to do some work on the superstructure, the hull is a long way from perfect, but good enough so I feel I can leave it for now.

[middle_watch]

A shot without the building frames.

[/url]

[Timmy C]

Looks very nice now ^^

[middle_watch]

And a quick peek at some detail, the troublesome vents that have been giving me a headache on the real model, the location of these changes with each varient and refit it seems. This is an Engine Room vent, though uptake or intake I don't know, an odd feature is that the main mast itself acts as a vent also, hence the open hatchway to be seen on photos, though the Kiwi ships in later life had these fitted with hatch doors.

Hope to have all done soon anyway, I believe that many of the hatches were kept shut by default and presumably only opened when needed in hotter climates, but they also had a part to play in NBCD when the ship was closed down and made gas tight; perhaps somone more qualified can shed some light.

[Cadman]

Those are tedious. I had a fun time with this Mk4 screen.


The screen is supposed to be oval openings but that would have created a very memory intensive part.

[scottrc]

Those are tedious. I had a fun time with this Mk4 screen.


The screen is supposed to be oval openings but that would have created a very memory intensive part.

Your MK4 looks a lot like a debris catcher I'm designing for a skid loader in Solidworks.

Great!

[Guest]

Just not enough time in the days to do all! But while working on an article on the Y-100 engine for the Leander Class I mocked up this quick Engine Layout. Hoping to able to get a plan of the steam and water feeds to finish it.
]