Riveted Ship Hulls

I was wondering if anyone has a good solution for going from a molded hull form to a riveted hull form.

If I have the molded line (red), the exposed edges of the plates (green), thickness of each strake and know whether the the edge is on the inside or outside of the molded line, and amount of overlap, is there some tool to build a hull shell?

My understanding is “molded lines” for most ships are to the outside of the frames. The frames are built based on the lofted lines, and then the plating is applied. To the extent the plating fits tight against the frames the molded lines would also be the inside of the plating.

I have modeled lapstrake wood hulls but not riveted metal hulls. For the lapstrake hulls I start with the inside of the bottom strake. OffsetSrf (with loose option) to obtain the outside of the plank. Then work my way up the hull by modeling the inside surface, OffsetSrf for the outside surface, and then the next strake.

What will the Rhino model be used for?

That is the simplified description. The molded lines are are really the reference lines that define the shape form of the major structures of the ship.

Normally, the molded lines for deck go under plating. But, if the plating thickness changes, you’d want the deck to be level on top. In some areas, the plating goes under the modeled line.

In the case of a riveted hull, the plating can zig-zag across the molded line because of overlaps. Ship designers appear to have taken artistic pride in how many variations they could come up with. In some cases adjacent strakes may be flush and have a cover plate. In other cases a strakes can alternate such that one is along the molded line on the inside and the ones above and below are on the outside. In other cases, they just zig-zag.

This is why riveted ships were built from the shell in. The frames were templated to the hull shell as the shell was constructed. Consequently frames invariably have jagged edges. This is also why the actual measurements of large riveted ships deviates from their designed measurements by several inches.

The model will be used for illustration.

Still hoping on this.
In this illustration I have a cross section with the molded lines grayed out.

In this particular example, the P strake sit inside the molded line while the other strakes lap.

The thicknesses and laps change over the hull length.

I can map out all the sight edges, laps, and thicknesses but translating them into plating for study is impracticable without some kind of tool.

You’ve presumably got a set of moulded offsets and you also seem to have plating details.

Is there a way you can reduce this to a series of geometrical steps, starting with the first plate on centerline bottom and describing how to position the next one?

Is each plate a well described flat plate or rolled arc with flat ends to mate with the adjoining plates and you have an Ikea kit and need CAD help laying out the parts on a curve and riveting them? Or are does the curvature of some/all curved plates need to be reverse engineered from the station curves (offsets)?

You’re sitting at the interesection of two different areas of expertise here: naval architecture, specifically a historical corner of it, and CAD.

If you can get it translated into a specifically CAD problem (no NA terms of art or contextual knowledge required), you’ve got a neat application so someone might jump forward with a solution.

The problem is that you have to deal with plate bending, especially at the edges of a strake. It’s a complex problem. I always wonder if someone has tried to tackle it.

From this structural section, it looks like they might be just alternating plates, one side outside and one plate inside the moulded dimension:
Naval Gazing Main/Armor Part 4

Does that change over at some point along the hull to match your diagram of what looks like a forward section and the original post which was vertical?

Overall, are the Iowa’s simple, or an assemblers and modeler’s nightmare like the Titanics appear to have been for the bottom plating?
RMS Titanic Under Hull Plating Kit for 1:200 Trumpeter Model (woodysmodelworks.co.uk)

Since you’ve already been waiting since last year, If I were you, I’d probably start with the historical and modeling communities to figure out how it was built up and then here to turn that into a CAD process.

It’s much more than alternating plates. It’s defense department gold plating. The relationships of the strakes changes throughout the length. I’ve created giant diagrams showing each plate and the type of joints among them. Some places use ship lap. Other use alternating stakes (which is appears to me Titanic used). Some times the plates are butted. Sometimes the plates are scarfed. The direction of a lap joint can reverse. Only the military would do it that way.

The objective is to have a reference for inspection. the 2D diagrams were used for diver inspections. There is a desire for 3D to compare when the ship goes into drydock in a few months.

A weirdness is that many of the lap joints are welded, rather than riveted.

This sounds so random that it seems like coming up with some simple process is almost undoable…but what are we working from, or could you be working from?

I am working from original blueprints and microfilm. It does appear random. To me it looks like engineers were doing things just because they could do them.

However, much of the information is encoded so, in theory, a program could solve the problem.

Needs to be printed out at least 6 feet lone to be readable:

A weirdness is that many of the lap joints are welded, rather than riveted.

To me it looks like engineers were doing things just because they could do them.

There’s a method to the madness: vertically, they were worried about bombs and plunging fire up on and near the deck and below the water shells and torpedoes or mines down low. Up near the bow, they probably had it in their minds that the ship might need to ram an enemy which leads to different structure.

We could put together a set of rules to build up structure, but I doubt if there’s a better way to produce something close to what they actually did than pulling straight from each structural detail drawing.

You could try to identify a handful of patterns and then create scripts to use locally to make the input less tedious, but trying to guess exactly where to transition from one set of rules of thumb to another is a hard problem.

There are a few general paters

  1. Strake parallel to the molded line.
    a) The plate can be on the molded line.
    b) The plate can be inside the molded line.
    c) The plate can be outside the molded line (ie, has plates on either side of the same thickness)
  2. Strake can be nearly parallel to the molded line. (ie, the has plates on either side of different thicknesses.
  3. The plates can be ship lapped.

The first group (1.a-1.c) are reasonably easy to do. One can split a molded line surface then offsetsrf.

The second two groups are problematic.