Pitfalls of 3d Design

There are several things to be aware of when you're trying to design a model for 3d printing. They are:

Manifold Edges
Surface Facing
Material Thickness
Material Optimization

Manifold edges are probably what people struggle with the most when they start out. It's a big, scary word that actually doesn't mean very much.

It's basically saying that all the edges of a model must be connected. Imagine you have a box. The box must be connected on all sides in order to hold liquid, right? Something like this would be NON-MANIFOLD:



The edges aren't connected, so the the computer (and therefore the printer) cannot build the shape. Now, this is an obvious example. Lets look at one that is less obvious:



Now, you may be asking what's wrong with this. After all it's an open box, right? Well, not exactly. See, a line and surface has 0 thickness. That is, it doesn't actually exist. The computer is simply telling you that you're working with a surface this size - but the surface has no thickness yet.

In the real world, everything has a thickness. Even paper and atoms can be measured, so we have to make the same true of our models. Lets give our box a thickness:



Alright, now our box has thickness. It's ready to print, right? No. Now we address the second problem. 3d printers need to know what surface is on the outside and which is on the inside. You'll notice the box has a blue-grey color and a white color. The blue-gray marks the interior of an object, and white the outside of it.

The rule of thumb is to make certain everything on the outside is white:



Alright, now this box is 100% 3d printable!

But, if you were to put it up on the market, it would probably cost around $10,000. Why? Material thickness. Shapeways charges by the amount of material required to print the model. For this reason, you'll need to make certain that your model is A) The size you want it to be and B)hollowed out (material optimization).

I'm not going to get into hollowing models because that is a full post on its own.

But let's discuss material thickness. Each material has a minimum thickness. If the material is too thin, it will break when it is cleaned or during shipping. Don't worry, Shapeways will detect the problem and reject your model so you don't waste money. But it won't tell you where the problem with thickness is.

There are a lot of different materials, and I won't go into them all right now. You can find this list (along with a lot of other helpful tips) here.

A plugin that is extremely helpful is the solid inspector. This plugin will scan your models for breaks in the surface and a few other problems and highlight them. This can save you a lot of time, especially if Shapeways is consistently rejecting your model.

Stay tuned for my next entry, where I'll explain how to make rounded surfaces!

First Post

Ah, so here we go, a blog. Where to begin? Maybe I should start with my intentions.

Keep in mind, I am not an artist, not in the least sense of the word. I make horrible stick figures and scraped by my high school drawing class with a D+. But, I was rather taken recently by the prospect of 3d modelling. Interestingly enough, 3d modelling is not a normal medium like oil or charcoal. No, it is its own, unique form of art.

Those of you interested in it might rejoice in the fact that it does not require great artistic talent to master. At least, not that I have seen so far.

Also let it be known that I have taken no classes or made any studies in the field of 3d art or animation. Everything in this blog will be from my own trial and error, and perhaps, a way for you to avoid the same pitfalls. I do not assume to be a master, expert, or even journeyman in such things. But maybe that is for the better. There are more students than teachers, and there is something to be said for group learning.

As a brief history: I started playing with Google Sketchup (v8) in mid-December. My objective - to design a 6mm scale tank (roughly the size of a quarter)for 3d printing by a company called Shapeways. After a few hours of trial and error, I had something that looked like a tank:



I promptly uploaded it to Shapeways, where their automated system promptly laughed me off the internet. See, it's not enough to simply make a 3d model. Printers are very expectant machines. They expect to know where to begin and where to stop. What is on the outside, and what is on the inside. Where one object begins and another ends.

Being only vaguely aware of such things, I started googling anything and everything I could about 3d printing. Finally, I figured out exactly how the model should be put together. But I didn't feel like continuing on my tank yet. I felt I needed to start smaller, with a basic structure that already exists.

So, I started playing with a bolter from Warhammer 40k (a 28mm scale strategy game). I decided I needed to do 2 things. The first, was make a base model to drive all others from. This would have the same dimensions as a normal bolter,so I could add or remove additional features without worrying if the bolter would fit the model it was meant to go with (Space Marines):



Bummer. With that in mind, I decided on
With my base model completed, I decided to make a larger, more "assault rifle" version. I increased overall size by around 40% and and removed the front sight aperture in favor of the holographic sight. I also decided to incorporate combi-weapons, the 40k equivalent of under mounted weapons:



The problem? Well, it was too big:



Bummer. So it was back to the drawing board. I decided that I would decrease the size so it was only about 20% larger, and add modular optics, combi-weapons, and magazines. This way there would be much more customization:



It looked better, but there are still size problems. I'll introduce those in my next post.

The second thing I realized I needed to do, was begin making a component library. In Sketchup, a component is an object or objects that are generally considered finished. You can copy components out of the library and right into the workspace. In this way, I could make a base scope mount. Every time I make a new scope, I just copy the mount in and start working, then save the new scope as a different component.

Because I can very rapidly build from a base model or interlocking component, I'm able to add much more variety to my models than would be possible through conventional modelling. At least, in the span of time it takes me.

The downside is that any changes to the base model will require changes to the components. For example, if the top of the bolter were to become wider, the scope mount would have to be changed to match.

Making modular systems requires a lot of pre-planning so you don't end up doing the same work several times, something I've been learning the hard way and will share in a future blog entry (you're going to notice a pattern here).

So, stay tuned. I have some interesting things planned for the future.