Back to the roots

[Ray Dunakin]

Nice work. The printed parts look sharp!

[Peter_T1958]

Thank you all for your kind comments!

The bad news is you forgot to order some parts. The good news is it will take months to clean up the beautiful prints that just arrived; you won't lose any time. -- Russ

Russ, you are spot on with your forecast: Indeed, I needed about half a month to clean up the major parts for just one side! But at least it's starting to look like a tank. There is not much left of the original kit parts.



All remaining fenders are just to support the weight of all additional parts (i.e. toolboxes, exhaust system, tool rack). Later they will be covered with PE parts to catch the look of thin sheet metal. For this I usually first draw templates from paper as may be seen here.

[Peter_T1958]

With regard to Bill's telegraph apparatus I must confess that I would not be (never been ) able to assemble such tiny parts. That's one more reason I try to print as much as possible in resin printing.
And «my» printer in Romania has improved once again with this one single piece on the left, the fairly thick rear armour with the attached delicate frames and, on the right, with this extremely fragile shutters (they are used to deflect hot exhaust air).



And it fits like a glove!



It's a pity, of course, but this will barely be visible at the finished model!

[Peter_T1958]

Back to my Centurion tank on the working desk. There are a lot of inaccurate parts waiting to be improved. This week I focused on difficult 3D parts – although they are rather small :

The towing hooks are a real nightmare to design:




And these weld seams consist of countless rounds and elliptical solids of different sizes and arrangements. My CAD program doesn't allow more complicated activities.
Number 1 is the initial construction, number 2 a refined one:

[finescalerr]

Whoever buys the kit you are designing should be extremely happy. -- Russ

[Bernhard]

It's amazing what you get out of your CAD. I always struggle when it comes to free-form surfaces and volumes like this. Hats off to you!

Bernhard

[Lawrence@NZFinescale]

I suspect I know what you went through to produce those Peter!

When scratchbuilding, it's common to try to get everything as square and true as possible.  With CAD modelling that's a given, and in searching for realism it can be useful to distort things a bit (The doors in the attached have been dropped and rotated to suggest worn hinges).  In addition things like weld beads, wood grain and more 'organically' shaped objects are generally not dealt with easily with traditional CAD.

After decades in AutoCAD I recently switched to Rhino.  That was fairly (but not entirely) painless as Rhino can be used in a very similar way.  But it also offers more integrated ways to do 'organic' things.  Not that I've explored that much, as I've got used to using Blender for soft modelling.  Rhino is much more accommodating of moving things in and out of the program though.  In any case a workflow involving programs (or feature sets) suited to 'hard' and 'soft' modelling is very useful. 

In the attached render, most of the model is classic CAD.  Started in AutoCAD and finished in Rhino due to my software switch. The hooks were done in AutoCad and like your tow hooks were painful, though ultimately successful. I strongly suspect that they would be a lot easier in Blender (or with surface modelling in Rhino).

The little chains at the doors were done in Blender.  They are a procedural thing, take just a couple on minutes, and remain easily editable for drape, numbers of links etc. They are a bit of a fudge.  The links will distort on sharp curves, the links pass through the body in places and so on.  In this case I'm happy with that as the chains are small and the compromise invisible in the final print.  To get them to print, making them more of a surface feature provides support.  I have used other methods in Blender to produce draped chains that behave more realistically, but they offer no advantages for this model.

Once I'm happy with the result they can be brought back into Rhino for a render, stl export etc. Doing them in CAD would be painful.  Drawing the links would be easy, but manipulating the chain as a unit far less so. I'd rate myself as a beginner when it comes to Blender, but there are plenty of videos for things like chains and weld beads that are easily followed.

In short, I've found that while most things can be done in CAD (or Blender for that matter) it is much less painful, and far quicker, to use the software that is best for a given task. One of my favourite features of Rhino is that it makes import/export pretty simple. This approach certainly made the attached wagon quicker and easier to do.

[Peter_T1958]

Thank you all for your feedback and Lawrence for his explanations! «Rhino» would have been one of my favorities, and I see the advantages are obvious especially regarding the limits of my CorelCad. I also had done certain things with «blender» as folds in leather seat cushions or folded hoods. The main problem was (and this seems to be a huge advantage of «Rhino», as you wrote), that I could not re-import the edited files in «CorelCad».
But, I don't want to fool myself, I see no way to obtain « Rhino » in my current period of life. My PC is more than overloaded with my actual attemps, and since I took an early retirement three years ago, the pension allows a good way of living, but not more.

BUT... That's absolutely no problem for me: I am living almost in paradise as you may see in this picture of this morning...  What else could you want? 

[Lawrence@NZFinescale]

The main problem was (and this seems to be a huge advantage of «Rhino», as you wrote), that I could not re-import the edited files in «CorelCad».
But, I don't want to fool myself, I see no way to obtain « Rhino » in my current period of life.

That's not the end of the world.  I couldn't reimport with AutoCAD either (which has been all but the last 6 months of my CAD experience).  You just need a more linear workflow where you do all your CAD work, export to Blender for organic stuff and then to your print file.  It doesn't need to be completely linear as you can always reimport new versions of the CAD data into Blender at any time - but obviously cannot go the other way. Reimporting to Rhino is nice, but I still follow this basic workflow.  Reimporting just allows me an easy way of rendering and visualising - I haven't used it to help with the actual modelling yet.

[Peter_T1958]

Oh, I forgot to mention: Very clean construction of an interesting wagon...

[Peter_T1958]

Side note: I was asked, whether it would not be simpler to use putty for these seams. The problem is, that modern tanks have more sophisticated/complicated weld seams than WWII vehicles. I had no clue how to represent such flaky ones like the ones on the engine deck. And I am very pleased with the outcome...

[finescalerr]

Superb. -- Russ