Complicated Software

[753]

Well it's certainly interesting to hear members views on this subject. My conclusion is the possibility of a Widows style program for 3D printing aint going to appear any time soon. So if you want to learn to swim jump into the deep end, I had to with the CNC mill, and some programs are easier to learn than others.

As to programming I think Ralf's statement

" If you examine the .scad files that I have written in the WIKI they will look simple, short, and obvious. It has taken 40 plus years to get to this level!!! "

says it all, and as we all know there is not substitute for experience, know where's my hammer.

Mike

[John Candy]

I would add (for the benefit of the disbelievers) that it is NOT difficult to master the basics of OpenSCAD.

Yes, I did learn computer programming in the mid-1960's (my interest was sparked when I attended a series of lectures at the Science Museum, given by English Electric Computers, later to become ICL) but have not "used it in anger" since the early 1980s.

As Ralph has commented, it is similar in structure to BASIC but it is far, far simpler to construct a 3D print model than write a suite of programs to run a business.

Pretend you are a brainless machine waiting for instructions as to what to do.

Basically, you need to know what shape to draw, how big and where to place it.

The most basic instructions are cube,sphere,cylinder and with those you need only specify the dimensions and the x.y,z co-ordinates telling the machine where to place the object within the workspace.

Yes, there are very much more complex objects, some of which do need to be extruded from 2D designs, examples being a circle extruded to form a ring and a polygon for solid shapes which are not cubic and some do involve angles and more complex parameters.

I assure you that if you start with something simple, you will soon get the hang of it.

I put off trying for far too long and within a few days I was able to produce chimneys, domes and coach sides.

Regards,
John.

[IanT]

John C. - 'Slicers' basically take a .STL file and convert it to something the 3D Printer can understand - so they are all the same in that respect. However, once you need to start digging under that level of simplicity, things do get more complicated with respect to how much and what kind of customisation you can actually do.

I've only used Cura myself, so cannot make comparisons but I believe there are differences between products once you start customising things like supports, object scaling and placement, infill proportions & patterns etc. These differences might just be finding the right menu choices for your particular slicer - or that some just cannot do things that others can. I'd recommend following the Gauge 1 3D Circle at Group.IO for useful information in this area. They also have some G1 wagon designs that could probably be scaled up to G3 just using slicer scaling.

John B.  I would download Open SCAD and try typing in some simple objects. It is much easier to learn in steps than might be suggested by looking at complicated scripts. Generally (I think that) 3D Printing is easier and more affordable than any form of CNC.

Having said that, there are now fairly affordable 'router' style CNC machines that would probably be capable of cutting the work I suspect you have in mind.

I've thought about this myself recently and considered getting a 6040 type CNC machine (but fitted with an 800w water-cooled spindle) which would be powerful enough most of what I might want to do. This would cost about £1K (pre-Jan). You can buy much cheaper set-ups with smaller tables (3020) and small DC motors but YT is full of 'upgrade' videos to make them more usable. You can programme these machines directly in G-code (perfectly possible for simple cuts) but you will eventually need some kind of CAD/CAM software to do more complex work.

Your idea of the software taking a 2D 'plan' and turning it into '3D' is interesting but in many ways is the opposite of what actually happens. In 3D, you sketch a 2D plan, then extrude the third dimension. This continues in stages until you have built your model. You can export this in different file formats - so .STL for 3D print or perhaps .DXF for CAM/CNC.

One of the great benefits of modern 3D CAD is that it is parametric, so any changes are reflected back into the overall design. If you need traditional 2D drawings, then it's just a drag and drop operation. Change your design and everything is updated automatically, including all drawings and their dimensions etc. For the kind of work I know you do - I think spending time learning 3D CAD would be well worth while. I'd recommend Solid Edge - and it's quite capable of '2D' drawing as well as '3D' (something I didn't realise at first).       

I'm busy re-building my small mill at the moment - something I'd been putting off. But I managed to break the back-gear last year and finally decided to bite the bullet. I think I'm moving to the view that I might be better to fit an inexpensive DRO to my mill - rather than just dive into CNC - but everyone has different needs & interests. A small CNC router (for thin non-ferrous & ply work) might still be on the cards though.

Regards,

IanT

[Nick]

I've been reading this thread with interest. As I have mentioned here previously, I use Fusion 360 extensively and it does everything I need it to do (and a huge amount besides). Yes, the learning curve was steep and it took me three months of winter evenings until I became confident with it. At the same time I had a long and hard look at OpenSCAD (inspired by Ralph's Wiki to do so). Eventually I rejected it because I didn't want to go back to writing code (I stopped doing that 30 years ago), and because I like the facility to work with the 3D object directly on the screen. I'm not making a case for one or the other, I am just saying that anyone getting into CAD should seriously consider both the scripting and graphic options before making a decision.

Ian, please don't buy a 6040 machine or anything of that type, you will be wasting your money if you want to work with anything other than plastic or wood. The problem is not power, speed or accuracy, it is rigidity (lack of). The frame is simply not rigid enough, you will get tool chatter and a good surface finish is impossible on any harder material. Any machine made of flat plates, no matter if they are strengthened, will vibrate far more than we want. Cast iron is better, but best of all is epoxy concrete which is what many industrial CNC machines are made of because it damps vibration so well. I have some familiarity with them - even the small ones are massively constructed for a very good reason.

Nick

[John Candy]

This screenshot shows the OpenSCAD file (the full code) to print the ducket for the LNER Gresley PBV.

It is a fiddly shape to produce by fabrication but the program is very short and quickly produced.

No cut or burnt fingers....just make a coffee and let the machine do the work.

Regards,
John.

P.S. You will (eventually) be persuaded that learning SCAD it is worth the effort!

[John Candy]

A second file showing a "see through" version of the ducket (the version above was deliberately infilled.... the blue bit).
When printing the "see through" version, the software will automatically insert a fine mesh support structure which is easily "poked out" when printing has completed.

John.

[cabbage]

The people who use text based systems and those who use graphical based systems operate in two different modes... Here sat in Stoke station I can write .scad files on my phone without using OpenScad and by practice I can visulise what I am creating. The processing power required is very low.

If I run a program such as Bryce, I select the primative enter the position (x,y,z), colour, material and surface finish. The principle is the same as a .scad file but you use a mouse and enter numbers into the boxes.

The processing power is vast (!)

A 3D CAD system is as complex as you want it to be. Text based systems appeal to me because I only use what I need.

Regards

Ralph

[John Candy]

2 hours and 6 mins printing time at 50 microns and 50percent infill.

The completed ducket is loosely sitting on the pattern for the side panel section (it will not become part of the resin casting .... the side panel is in its enclosure, ready to pour the silicone rubber mould).

Another ducket is already being printed .... only the later LNER builds of Gresley coaches had duckets and then only on one side of the coach (left, looking forward.... the side most signals would have been sighted).

Regards,
John.

[AshleyW]

is the one side only an april fool john? as when they arrived at a terminal, they'd have to take the coach away to a turntable and turn it so the ducket was on the left for the return journey ?

[John Branch]

But if there were a brake end at the other end of the train, it would have its ducket on the right, that is left, side, when going home so as long as the guard travelled at the right end, that is the last vehicle, he would be on the left, that is, the right side to see the signals.

Are we all clear?


John

[John Candy]

No, not a joke!

John B has summed it up nicely ...... important trains were made up of sets and would have a brake carriage at each end. What is puzzling is that Gresley's GNR carriages did have duckets BOTH SIDES but his early LNER designs had none at all and the later ones had them on just one side. The logic escapes me but it may have been to save costs. Not sure how much use they really were in practice. The GWR didn't bother with them post-Dean and the SR switched to periscopes under Bulleid.

Regards,
John.

[John Candy]

The OpenSCAD for the 8ft 6ins width gangwayed end (as fitted to Gresley LNER bogie PBVs and van sections of many passenger brake coaches).

The gangway connector will be produced and separately fitted to completed models.
It took about 20 hours to produce the program and I will soon know how long it will take to print .... it will be used as a pattern for resin casting.

While looking at the finished item in OpenSCAD, I hit the "thrown together" view option.
The astonishing result is also shown ..... that design reminds me of the symbol for a certain organisation.

Regards,
John.   

[John Candy]

It took just over 20 hours to print at 80microns and 15percent infill and will be used as pattern to make a mould.

I am now "hooked" on 3D printing and have ordered a liquid resin printer, the Elegoo Mars 2 Pro.
Previously wary of the "mess" and smell associated with the process, I have also ordered the Elegoo Mercury wash and cure machine which receives the print still attached to the production plate, so you don't have to touch the resin before it has been washed and cured.

The Mars 2 Pro printer has a capacity of 129x80x160mm and is fitted with a carbon filter to remove the particles from the cooling system exhaust. It will be used to produce smaller parts than the Snapmaker can handle but should be much faster and give an even  better "finish".

Regards,
John.

[753]

John

No stopping you now, let us know when you are ready to receive orders for rakes of coaches. A set of GWR clerestory would suit me!!!!

Mike

[John Candy]

The Gresley coach project is now well advanced with the patterns produced for six different LNER diagrams.
The Snapmaker has now clocked-up getting on for 200 hours ... it is rattling away as I type.

OpenSCAD works well for me but for those less confident in their abilities, I wonder whether "Tinkercad" could be a solution?
It is web-based (you work through your web browser) and, from a brief look at the tutorial (link below) works by manipulating pre-defined shapes.

https://all3dp.com/2/tinkercad-tutorial-easy-beginners/?utm_source=push

https://www.tinkercad.com/

Regards,
John