Various applications of 3D drawing

[3Dships]

Just joined the forum as I feel that after practising virtual ship building for decades things appear to be maturing. I did start with real modelling as computers where not available when I was young though.

 

I guess I was quite an early adopter with drawing ships (1:1) in 3D back in the late eighties. Initially starting with AutoCAD (when Bézier Curve algorithms where introduced) then switching to 3D modeling applications (for their Catmull-Clark subdivision surfaces) & then back to CAD (when NURBS and better meshing algorithms where introduced). In recent years, NURBS based CAD applications have also started to implement Subdivision modelling algorithms so you can do both the accurate & artistic stuff in one application.

 

Since I was early I got the opportunity to explore many of the various application options that drawing in 3D offers over the past decades. Re-engineering, design visualisation, simulation, construction, 3D printing, etc, etc. However, just like in the real world, the construction of watercraft much depends on building materials / toolset restrictions & the same applies to digitally drawing watercraft in 3D. These days I am getting the feeling that the toolsets most CAD application offer have become extensive enough to pretty much derive the many different possible end products from one accurate 3D drawing dataset.

 

Since I have an operational background in commercial shipping (Merchant Mariner) & the pressure on the shipping industry to reduce it's carbon footprint is growing, I recently have been researching the period where the shipping industry changed from wind to fossil fuel energy. Obviously, during this period wind powered design lost it's monopoly and was forced to improve it's efficiency while fossil fuel powered design was still suffering from not being very efficient leading to interesting design solutions.

 

I attached some visualisations of a reconstruction in 3D based on blueprints by a shipyard in the Basque Country dating back to 1919. It's a 24.5 meter long fishing vessel fitted with a 50 ihp steam engine that takes up around a third of the space below deck while still carrying some sail. The drawings are from a large collection gathered over 3 generations by a Basque ship building family and going through this collection it once again becomes quite apparent that copying was rife in the ship building industry in previous centuries. The design is clearly based on British Steam Trawlers from that period and it turned out that the elegant looking very long stern overhang didn't work out that well on the Spanish Atlantic Coast. Entering port often required getting over a river bar with sometimes large breaking surf, leading to a few of these elegant stern sections completely breaking off.

 

Visualised is a solid version of the 3D dataset that for example allows for low & high polygon export to 3D modeling applications like Blender for texture mapping, visualisation, animation & STL export for 3D printing applications. Obviously, perfect 2D plans are also easily derived from this 3D dataset. Still busy trying to find more photographic material for the details because there's little detail in the blueprints & since photography was still limited a hundred years ago some creative effort might be required.

 

 

 

 

Edited April 19, 2022 by 3Dships

[Waldemar]

 

Very attractive renderings.

 

Just curious, what software did you actually use for both 3D modelling and for visualisations?

[3Dships]

There's a difference between CAD & 3D Modelling applications in my view. I don't think there's any 3D modelling applications that offer NURBS surface toolsets but recently NURBS based CAD applications started to implement the Subdivision modelling toolsets that you find in most 3D modelling applications.

 

I use MOI 3D for the NURBS modelling & exporting meshes.  The developer of MOI 3D was initially involved with Rhino at its very beginning but wasn't happy where Rhino was going, left and started developing a more intuitive & user-friendly NURBS based CAD application. MOI 3D is much cheaper & is more focused on exporting good quality meshes for 3D modelling & printing applications than the full blown CAD applications. It doesn't have all the bells and whistles that full blown CAD applications like Rhino, Solidworks, etc offer but it's toolset is quiet sufficient to produce very accurate curvy designs and you can work fast. It uses the same main file format as Rhino. I produced the visualised fishing vessel in it's current state in a few days for example.

 

All of the test renders except the last one are from Lightwave 3D with materials applied to different surfaces. The last render was done in Blender after UV mapping with one texture bitmap image applied to the model for the colors. Recently I've started exploring Blender with an eye on building a virtual ocean with animated waves & boat wake as Lightwave 3D has not been going anywhere for some years now. Must admit that Blender has quite a steep learning curve but that it does appear to develop fast.   

 

   

[Waldemar]

 

Thank you for the information. I asked about the program you use to create 3D model/objects, as I am just using Rhino myself. Unfortunately, this program (which I like nevertheless) has a tendency to produce much rubbish when generating 2D sections and projections from 3D objects, and I will have to deal with this particular problem soon.

 

Maybe you have some suggestions on this particular issue. In short – how to create (in Rhino) in both exact and hopefully easy way 2D sections and projections from 3D objects, while retaining its vector character (as opposed to bitmap graphic). And these are ultimately needed for "classical" modeller's draught.

 

[3Dships]

I just use "intersection" for deriving 2D sections from a 3D object like for example frames, buttocks & waterlines. You draw lines where you want your 2D sections. Extrude them to surfaces that "intersect" your main 3D object and then get the curves from these intersecting surfaces & objects. You can "project" any 2 or 3D curve like for example the curves you got from "intersection" but also the outlines of a 3D object flat on any of the Cplanes and with some simple us of layers & rotating you can then derive a conventional 2D plan from your 3D drawing in a few steps.

 

I usually already set up 2D line grids at the x, y & z planes at the very start of a new project for this purpose since the 2D intersections are a good tool to check the fairness and accuracy of a 3D surface.

[allanyed]

Welcome to MSW 3Dships,

 

Now that you have come aboard, it would great if you would post a little intro about yourself in the new member forum.

Your drawings are terrific!   I've struggled just trying to do cannon barrel drawings for various time spans in the 17th and 18th centuries.

 

Sailed the seven seas in the MM?   Me too, lo those many years ago.   US or other flag?

 

Thanks

 

Allan

KP '69

 

[3Dships]

Sorry Allan, I wasn't even aware there was a new member introduction facility on here and during the past decades sailing the 7 seas & learning to draw ships in 3D drawing kept me so busy that it prevented me from joining any forums like these /-: However, 3D drawing a gun only takes a minute though. Makes a perfect solid object for 3D printing too (-: I am Dutch and staying in the Basque Country at the moment researching their boat building history after having done the same in England and Norway previously. 

Edited April 20, 2022 by 3Dships

[mtaylor]

Err, that wasn't me, that was Allan.   But it is a good idea to post an intro where he suggested.

[rth]

Delftship is a perfect solution for modelling ships with a low learning curve.

You can download a forever free version @ www.delftship.net

 

 

[allanyed]

3D

You make it look easy, but after your many years of doing this, that seems logical.  Not so much for us neophytes. 

 

 There are a number of articles here at MSW by a number of authors in subjects from basic lines drawings to planking to making various items such as stoves.  It would be SUPER if you did one with STLs of various cannon that can be downloaded and then they can be 3D printed in resin.  

 

Let me know if this would be of interest to you and I can email you 2D drawings of cannon designs and sizes from about 1675 through about 1815.

 

Cheers

 

Allan

[Kevin-the-lubber]

That's a very nice looking model and, as others are saying, you are probably light years ahead of the rest of us who have only recently got into 3D modelling and your knowledge would be very welcome here. Despite meaning to switch to Rhino a few months back I'm still just using Fusion 360 and, one day, will attempt a hull. Everything else in CAD/3D modelling apps is, as you say, fairly straightforward once you've learned enough. But I know a hull is going to suck up weeks of my spare time so I'm still ducking it for now.

 

What do you do with your designs? Print them, or are they purely for the pleasure of the exercise? I'm sure there is going to be a market for 'download and print' models

[3Dships]

Of course, it's true I have invested a lot of time in learning this 3D stuff. No prob to have a go at your cannons if you supply good reference material. Just PM me for the emails address.

 

Regards, Hans

[3Dships]

Hi Kevin, thanks for the kind comments. I guess having followed the whole trajectory of CAD & 3D modelling from the early beginnings did help a lot (-: 3D software has become a bit overloaded with functionality and the learning curve has become steeper & steeper over the decades. I guess it is also quite tempting to think that for example lofting in a CAD program works the same as in the real world but it really doesn't. 

 

I do love the exercise of doing a perfect hull surface with as little geometry as possible the most & generally, vessel depending of course, it takes me a day or two with the time consuming devil being in adding the details. The details also often require a lot of research because there's no drawings. At the moment I am trying to find photographs of the deck gear of the model I am working on. I have blueprints of 3 different versions of this "Vapor de Pesca". One has a lines plan only and then there's one with lines, a GA and some construction details and one with only a GA and construction details. The GAs only really offer some details in one dimension so either some photographs or creativity will be required to complete the model as far as that is possible.

 

I have been selling 3D ship models aimed at visualisation and simulation online for many years and there's clearly a good market but instead of just capitalising on it I have been investing my time in further refining a one in all workflow with the aim to produce 3D drawings I can derive all the different possible end products from. That goes from producing perfectly accurate 2D plans to low polygon simulation/gaming model, high polygon visualisation model to very high polygon 3D printable model. 

 

Regards, Hans

[Kevin-the-lubber]

Once you’ve done Allan’s cannons, I need a hull that’s an exact replica of the 1:96 Heller victory and, if you’ve got time, the same for the 1:96 revell Cutty sark 😆.

 

I’d add that, while I definitely couldn’t do the cannons fit to print as quickly as you (probably a couple of hours for me, including the carriages and fine details) they’re fairly simple to model and quite a good thing to learn how to do in a 3D package if you’re a beginner. Along the way you’d learn about simple shapes and extrusion, planes, revolving, cutting and combining and, if you really want to go the whole hog, surface projections. After which you’ll be able to make a very wide range of things for ships. For anyone interested, there’s an excellent beginners tutorial for f360 in an online magazine called diyode, Google it.

 

[Kevin-the-lubber]

Hans, you’re dead right about the devil being in the detail. I know I could get a reasonable basic hull done quite quickly but getting exactly what I want, or what it needs to be, in terms of detail and dimensions, is another matter.

 

That broader project sounds fascinating, albeit hard to get my head around, given the seemingly unique nature of the detail on most old ships.

[3Dships]

Did find some nice old photos of "Vapores de Pesca" but not any with deck details so decided to "get on with it" based on what I've got now. Ship building is not exactly "rocket science" in the end. Made a 3D printable manual winch in the last few hours.

[Azzoun]

That's lovely, are you going to post this anywhere for others to use?  Would these be suitable for 3d Printing?

 

 

Edited April 21, 2022 by Azzoun

[3Dships]

Yup, all solid models but things like the winch will require a resin printer I reckon. It's not finished so no plans posting it anywhere yet. 

[Dr PR]

Hans,

 

Nice work. I'm like you, I started in CAD long ago in 2D days and adapted to 3D when it became available. So I have had 30+ years to learn the tricks.

 

For prospective 3D CAD beginners there is one major hurdle, especially if they are familiar with 2D drawing/drafting. You have to forget everything you have learned and start over! In 2D work you create a flat drawing representing a 3D object, and there are many fairly standard techniques for this to show projections, dimensions,etc.

 

But in 3D you work in a virtual universe to create the actual 3D object. There is almost nothing in common with 2D drawing. You have to think in 3D, work in your virtual universe, and build the full dimension object. It is the same difference between drawing a house plan and then building the house.

 

So a beginning 3D designer has to learn the quirks and bugs in the software while trying to learn to work in a virtual reality. Like everything else it seems to be difficult at first. But if you are persistent it will eventually "click" and once over this hump it is a lot easier. It really isn't any different from learning to build a good scratch built model in the real world - but it is a lot easier to fix mistakes!

 

 

Edited April 22, 2022 by Dr PR

[3Dships]

Hi Phil, Thanks for your kind comments (-: Your "Okie" is an amazing piece of work! You even go down to the "nuts & bolts" here & there I see.

 

Yup, I agree with the getting around the "quirks & bugs" in the software & I've tried pretty much anything out there. It is simply building a boat but than different because you can't just bend a plank around frames, shave or sand off a bit to make something fit or roll a plate to get the desired curvature. A lot of trial and error and especially accuracy is required before you get to level where you can really become productive. Because there is hardly any limits digitally It's easy to get carried away without keeping an eye on what purpose(s) a 3D ship model is intended for.

 

I am really trying to focus on producing really clean solid objects these days so you can cover all the different purposes a 3D model can be used for. I also very much agree with people like Ab Hoving who does very realistic models of ships for which not much reference data is available in cardboard. It's interesting to note that so many plans turn out to be not accurate once you start to re-engineer them in 3D, creativity is then required and the same happens in real ship yards of course.

 

With regards to the easy fixing of mistakes I must admit that due to doing all kinds of very differently constructed ships I often start over again instead of trying to fix things. Even though the "Vapor de Pesca" is a very much a simple model I notice that I already created close to a 94 layers with 1966 parts. I guess some sort of structure in the way you work is very much important too.

 

Regards, Hans

[Kevin-the-lubber]

36 minutes ago, 3Dships said:

I often start over again

That’s quite a reassuring thing to hear from someone like yourself. I often find that what I’m really doing is a trial and error process, especially as it’s so easy to print objects. The first go usually highlights what I got wrong, and I then finesse. Recently I’ve started to do something I probably should have done from the beginning - treat 3D as just another tool and use other methods where appropriate, I.e. at 1:100 I’d probably print the body of that lamp but use wire for the guard.

[Kevin-the-lubber]

4 hours ago, 3Dships said:

Your "Okie" is an amazing piece of work

Ditto, both models are fantastic pieces of modelling. I haven't read the Okie log yet but I'm sure I'm going to find all kind of good tips in there.

[3Dships]

You mean trial and error with regards to the 3D printing of your objects? I try to keep the limitations of 3D printing in mind but with these resin printers you can already do pretty "thin" stuff. I'll wait until there is a 3D printer that can do a very detailed model that will fit in a bottle (-:  

[Kevin-the-lubber]

4 hours ago, 3Dships said:

You mean trial and error with regards to the 3D printing of your objects?

To be honest, not entirely. The trouble (benefit) of 3D design is that, as you’ve said yourself, it’s often quite quick and easy to correct mistakes, tweak the design and so on. Couple that with it being quick and often very cheap to print objects and you have a recipe for winging it. I know that if I was having to make the designs by hand, or outsource the printing, I would take a great deal more care to try to get the design right first time, think through the whole workflow for dependencies and so on. For the printing itself I always try to get it right first time because its such a pain when the supports fail or you find a plate of spaghetti in the morning. But even after a couple of years of printing, there is still quite a bit of trial and error to get the optimal result.

[Dr PR]

You are correct about getting side tracked and doing things that aren't really necessary for the model. Since I had the (almost) full blueprint set I started out to create the entire internal structure for the OK City - I was impressed of a CAD model of the Yamato that had all the frames, longitudinals and bulkheads. After messing with that for about a month I realized that it really wasn't necessary.

 

And I am an expert at getting sidetracked. I started my USS Oklahoma City web page (www.okieboat.com) to post modeling pictures. I got carried away and now there are something like 88 pages, most of it the history of the ship and the Cleveland class, plus the history of the development of the Talos missile (including GMM manuals), detailed descriptions of the weapons systems, engineering plant, ship's crew, sea stories ...

 

It took 14 years to complete the web page and the CAD model. I just like learning about things!

[3Dships]

The "devil is in the detail" I mentioned earlier and indeed seemingly simple things like a wooden block, shackle or railing can be quite challenging to do accurately with 3D CAD. I did it all before in 3D modelling applications like Lightwave 3D using subdivision but with nurbs it's very different ballgame. After some trial and error I am pleased with the possibilities and results though. Perfect tiny solid objects that "union" fine with the much larger solids.

 

One of the great advantages of MOI 3D over Rhino is that you can have one solid object with parts on different layers. If you take a wooden block for example that means you can separate the timber and metal parts while still keeping the object solid. Obviously, not an issue for export to 3D print applications but if you want to render the object in for example Blender this is actually a lifesaver as compared to Rhino that only allows for "join" and "union" a solid object on one layer and thus for example also only one material. I have no idea why such an apparently "complete" 3D CAD application like Rhino is lacking this important option. It is really no fun to having to completely deconstruct a complicated 3D Solid you have just put together.

 

  

[Dr PR]

You hit on a type of problem that is common in CAD software because the programmers have never used the program they are creating.

 

I cannot imagine a CAD program without layers, and the ability to create assemblies and sub assemblies on multiple layers and treat them as single objects. The CAD program I use has Groups and Solids. You can create multiple pieces on different layers, select them all and create a Solid or Group with them.

 

Solids behave like real solids, and you can perform boolean operations (add, subtract, etc.) on them to make holes, etc. You can add bits to existing solids, but if you explode the solid it falls apart into the individual basic components (planes, surfaces, etc.).

 

Groups are collections of different objects, like Solids. But groups can be nested, with groups of groups of groups, and so on. And each level in the nest can be exploded to release the groups and objects in that nest. The sub groups remain intact.

 

Think of an automobile engine. The engine block is a solid. It can be grouped with other solids like pistons, piston rods, crankshafts, etc. to form an assembly. The alternator can be a different assembly of solids like the shell, rotor, etc. Ditto with the water pump , power steering pump and air conditioning compressor. These can all then be grouped into a basic motor assembly. This motor assembly group behaves like a single object, and it can be grouped into a chassis assembly, and so on.. But if you explode this group each of the other groups remains intact.

 

It would be a great system, BUT!

 

For some reason the programmers placed an invisible Group "handle" on whatever layer was active when you selected all of the parts to be in a nested group, even if none of those parts are on the active layer (the handle should be on a layer containing a part of the group). Thereafter this invisible handle follows the nested group and you cannot select it. If the handle is on a locked layer you cannot select the group, even if all parts of it are on unlocked layers. And there is no way for you to discover what layer the hidden handle is on! The programmers had a great idea for nesting groups of groups, but totally screwed it up with the invisible, unselectable handle! It makes nested groups essentially worthless! If the bozos had ever used a CAD program to design complex assemblies they would not have created such a stupid design flaw!

 

Unfortunately, such design flaws can be found in almost all software, primarily because the programmers don't use the programs they create.

[3Dships]

I guess there's2 options why Rhino is not very export friendly with regards to solid objects. They took a wrong turn early in the development with no way back or they want to discourage users from exporting solid models to 3D modelling packages maybe?! MOI 3D is developed by just one guy and he is clearing taking the KISS approach with a focus on stability and avoiding UX clutter. There's no sub-layers and maybe that has something to do with being able to do all kinds of operations like "Join & "Booleans" on surfaces and solids kept on different layers. In the latest Beta that came out earlier this month "Groups" where introduced but I haven't looked into its use as yet.

 

Had to go back again for additions & thus also alteration work on the hull as I noticed in some photographs that these vessels had some planking on the inside of the bulwark that was initially not apparent in the drawings. Tricky stuff trying to match another solid that is curved in 3 dimensions over almost the full length with a bunch of other surfaces curved in 2 or 3 dimensions but it worked out alright. I can scale the hull solid object (1311 edges) from 1:1 to 1:1000 without getting any naked or non-manifold edges (-:

 

Now back to adding more details...

[Kevin-the-lubber]

Some of this is going over my head as a relative novice with only a year or so of 3D modelling experience. But across just about everything covered in the last 2 or 3 posts I’m thinking ‘but I can do that quite easily in F360’. Of course  ‘easily’ is a relative term here. Or am I missing the point? That’s not at all to say F360 is a panacea for 3D solid ship modelling, it has almost had me in tears of frustration at times, especially around patterning and lofting, but multipart assemblies are by and large quite straightforward. 
 

I had a look at rhino a few months ago as I was particularly fed up with f360 at the time, but as it looked like another long and difficult learning curve I parked that for another time. MOI 3D sounds interesting but a two week trial is too short compared to rhino’s 3 months (or f360’s forever).

[bricklayer]

Kevin

You`re sure about two week trial for MOI? Their download-page says "90-day trial".

 

Michael