the future?

[DKS]

My engineer friend said "It's not that complicated".

A basic law of technology (or just about anything in the universe): difficulty is relative. What's complicated to one person is virtually impossible to another. Even for trained engineers, some applications or processes are unfathomable. When anyone says something such as this is simple, I take it with a massive grain of salt, and question their veracity.

Geoff, just a basic example why simply rescaling the drawing won't work:  If you designed a locomotive shell in N scale and simply enlarge it to H0 or O scale, the walls will be overly thick. Same goes in reverse: if you designed the shell in HO scale and shrink it to N scale then the walls will be too thin. Other things also do not simply scale up or down either.  Sorry to burst your bubble, but this is where the technology is today.

Just to expand on this a bit: say the N scale loco shell has handrails rendered to the smallest cross-section possible by the printer; by the time it's enlarged to O scale, the handrails will be stovepipes. Or, going in reverse and starting with fine handrails in O scale, they will be far too small to render when reduced to N. In order to create "one-click re-sizing," highly intelligent software would be required to interpret how to adjust the relative dimensions of parts automatically. I am suspicious that, if such re-sizing is claimed to be relatively simple, the models involved are very basic, i.e. not very well detailed.

[wcfn100]

Sorry to burst your bubble, but this is where the technology is today.

Mostly. 

With the right drawing software there are a couple ways to make scaling just a couple clicks but it requires you really know what you're doing (in the sense that you need to think all the things that would need changing) and much more planning ahead of time to create your drawing in the right way.

I haven't looked yet, but I'm guessing some software will allow for logic statements such as 'if scale is 'ho' then 'min_wall' = .06''  or whatever situation you can think of.

And can we stop calling this 'new' or the 'future'?  I think I got my first Mark4 shells 7 years ago and had my first parts printed over 5 years ago. 

Jason

[DKS]

And can we stop calling this 'new' or the 'future'?

This.

[bbussey]

... I haven't looked yet, but I'm guessing some software will allow for logic statements such as 'if scale is 'ho' then 'min_wall' = .06''  or whatever situation you can think of...

You can, but it's not as absolute as it seems.  For example, that you "shell" out a solid detailed block at one thickness without incident doesn't guarantee that you can change to a different thickness without generating build errors.  Even with extremely careful planning, situations occur that require re-engineering.

David, you are correct in your assessment that 3d printing is difficult. However it all depends on the platform being used to create the drawing. On small parts it is more difficult and requires more work. My engineer friend said "It's not that complicated". But I apologize if I seemed cavalier about it...

One small point to make here.  I'm sure this is not the intent, but the text reads as advice based on theory/supposition being pushed on those who've actually designed solid models.  Degree of difficulty has little to do with the "platform," which I'm taking as referring to the workstation and/or design program.  But designing the parts isn't the issue, because they are virtual and vector-based.  The issue is rendering the design on a 3D printer.  Respectfully, it is difficult for me not to be dismissive of your position because you have no point of reference regarding the subject, whereas I and many other people here do.  Anyone who says that they can faithfully render a scaled-up or scaled-down physical model without modifying the virtual solid model does not have much practical experience in the matter.  Because no, it can't be done today on a consistent basis.

 

[DKS]

For example, that you "shell" out a solid detailed block at one thickness without incident doesn't guarantee that you can change to a different thickness without generating build errors.

I really have to wonder how this could be made to work. Let's say the shell must grow in thickness as the scale is reduced. In what direction does it grow? Inward? Outward? Both equally? What, then, happens to any features located on its surface--do they relocate as they, too, grow? I can envision the possibility of creating rules to compensate for all of the possible changes, but for a complex model--say, a steam locomotive--the rules to govern the behavior of the object's elements would seem to be overwhelming. Just saying...

[Mark W]

...I can envision the possibility of creating rules to compensate for all of the possible changes, but for a complex model--say, a steam locomotive--the rules to govern the behavior of the object's elements would seem to be overwhelming. Just saying...

This is exactly how it would be done.  And yes, it would take so much more time (factors of 100+) to script these expressions, that you might as well just start from scratch for a new scale.

[Iain]

 The issue is rendering the design on a 3D printer.  Respectfully, it is difficult for me not to be dismissive of your position because you have no point of reference regarding the subject, whereas I and many other people here do. 

My Lima is my first 3D printed project, and I tell you, there has been a lot to learn.  I am an experienced draftsman, working for several years doing CAD for a civil engineering firm.  Even with all that experience, taking on that project has been fairly difficult, and that's with Bryans advice on all sorts of things.

You say it takes one click to scale a model?  Well, my Lima shell is pretty darned simple (no surface details, for instance), and it flat out would not work if I scaled it up to HO.

[wcfn100]

I can envision the possibility of creating rules to compensate for all of the possible changes, but for a complex model--say, a steam locomotive--the rules to govern the behavior of the object's elements would seem to be overwhelming. Just saying...

Absolutely, so without defining what we're talking about drawing, we're just talking over each other.  Drawing style can effect this as well.  Some methods may work better for what we're talking about than others.

I already use variables in my drawing to try and account for different machines.  My standard two are 'minimum step' for stuff like plate metal, wood grooves or the minimum space between objects, and 'rivet diameter' for obvious reasons.  I also draw with a 'scale' variable since much of what I've done comes straight from 1:1 drawings.

In the end, most 3D design will be for personal use and relegated to a single scale which makes most of this moot anyways, but still fun to talk about.

Jason

[bbussey]

I really have to wonder how this could be made to work. Let's say the shell must grow in thickness as the scale is reduced. In what direction does it grow? Inward? Outward? Both equally? What, then, happens to any features located on its surface--do they relocate as they, too, grow? I can envision the possibility of creating rules to compensate for all of the possible changes, but for a complex model--say, a steam locomotive--the rules to govern the behavior of the object's elements would seem to be overwhelming. Just saying...

You can have it grow inward as you reduce the scale.  But if you have complex curves on the outer surface — the nose of an EMD E/F unit for example — then you can run into geometric problems on the inside as you "sharpen" the severity and reduce the proximity of those curved surfaces.  I've tried it, and it's not as straightforward as it would seem.