MakerBot Print is our newest print-prepration software, which supports native CAD files and STL assemblies,
allows you to interact with all your printers via the Cloud, and many other exciting new features.

Download Now

What to design?

Please Login to Comment

Hi, I know this might sound like a very dumb question, but:

I'm started Blender since several years ago, but never got above a very basic usage; I started designing stuff for second life (which had a very specific rules for meshes (or sculpties, back than), so its no much use for 3d-printing.

Since then I watched a bunch of tutorials but did'nt find anything that really suited my needs... I'm not very talented in creating any "organic" stuff, I'm more the practicanl oriented person with a preference for exact dimensions.

On the other hand, I can't get warm with any of the CAD-Softwares around so I'd really like to use blender as a tool, cuz I'm already kinda famliar with it: I created some basic parts for daily use already, or remixed things to fit my needs (most of it is too niche, or it doesn't have the quality i'd like, to post it here)

So I'm hoping you could give me any advice for anyday object I could design, as an exercise, which might be of acutal use - I know this is a very specific question which is hard to answer, not knowing my special circumstances - But I'm asking more for a brain storm than a real answer - Maybe some one can help me?

Thanks in advance!

I also learned Blender making Sculpties for Second Life and later Kitely that did not have the restrictions (like walking on sculpties) or the scale issues> If you purchased the means of getting from shape to sculptie you probably can get a free upgrade to the latest conversion app for blender and import sculpties to a mesh. They are likely to need a lot of work after that but it is a start.

Another big issue with things made for SL is that the scale is weird there, and much of the detail is done with texture rather than geometry, but again it is a start.( Istill have stores in the marketplace of SL and Kitely called Dragon's treasures, and still getting beer money though I have not been in, in over a year)

prosthetic duck bills

Blender is usable for technical parts and accurate dimensions. Maybe not ideal, but usable nonetheless. It does require a specific workflow and of course it does not have all the helpers and UI hints that a technical CAD program offers. However, I started out with Blender because I had some experience with it when I had bought my 3D printer, and I'm still using it for everything. I would love to make some tutorials showing how I usually tackle a design, but time is not on my side.

The way in which I make a lot of my designs is by using boolean operators that simulate actual machining operations. The booleans can be tricky and have a risk of producing broken models, but this risk can be minimized by pretending you're actually drilling or turning parts on a lathe. One rule to avoid problems with booleans is to never do two different operations with parts that share one or more edges.

If you want to have an idea what is possible with Blender, look at the designs by otvinta3d. They also have a lot of tutorials on their website. Even if a tutorial is above your level, you're still likely to learn at least a few useful tricks from watching it, or get a feel for how to get started with a design.

First: Thanks for all the answers.

I know that Blender isn't optimal for technical design, but I feel very comfortable using it. For now I could solve every task I had quite easily with it. I really like the workflow of roughly designing the parts and refining them with the grid for the dimensions I need. Pulling an edge, snap to grid, done. So the problem isn't the software, its that I don't have enough tasks to do. :)

Since I learn best by doing something, I don't get a lot from watching tutorials of stuff thats not straight at hand. By the time I could use the knowledge, its gone when I don't use it regulary.

For now I mostly designed stuff to solve own small problems: Holders for this and that, a handle here, a hook there. The design of static parts with defined dimensions isn't really a challenge anymore.

I guess I'll start with a bit more mechanical stuff, with moving parts, to get the hang of it. Already got a few ideas while thinking about it.

Thanks again. :)

Do not think what objects might be useful to other people. Think of something which is useful to you. It will then probably be interesting for others.

You lean Blender just like you learn any other complex software. Start with something simple and progressively add to your knowledge as you go.
1) Think of something that you need to create and that would be useful to you (not a tedious example) and make a start.
2) Build it out of basic mesh shapes (eg cube, cylinder). Join and cut them (with the Boolean tool) until you get the shape that you want.
3) Print your design and see what design features work and what do not.
4) Improve and adapt your design, learning the edit mode, manipulating vertices edges and faces. Use on-line videos and the help to learn perhaps just one or two keyboard shortcuts each time to build up your learning curve.

I started as an AutoCAD operator in an engineering context. Then 3dsmax and a smidge of maya in a game dev context, then also blender for SL. By the time I suddenly dropped out of SL dev I had pretty much mastered lsl and sculpties, but no doubt it's changed a lot in a decade and now I'm a noob again.
I use blender because it's free, linux native and adequate for 3d printing, but I'm not sure what to recommend or what tutorials to watch....but maybe stuff like: Learn the keyboard shortcuts, that's really blenders strength once you remember the basics.
Maybe also of interest: http://www.mechanicalblender.org/

As for model suggestions: Find components that you intend to integrate into 3d prints, and model them primarily with the intention of making a slot in a 3d design that the real object fits perfectly, and not so much to copying actual details. Dummy objects that are accurate enough to desighn around are really useful, often quick to do, and good practice.

I know your situation. I wanna Switch to Fusion 360, one day, but I'm too lazy for now.

I think you should design some mechanical parts, like threads / screws and gears. Also housings for your gears.
Start a project, like an R/C car or a robot.
Or - one thing I really appreciate - parts for 3D printers. There is a lot space for improvements.

I too have been fiddling with blender for years (also got a start creating things in Second Life). Organic modeling and sculpting are just no good for me (I do prefer to work with accurate dimensions).

If you are looking for things to model, I'd suggest doing things that you have an interest in. For me it's cars, vintage cars because there is more exposed detail in the suspension and such. I've also modeled Star Wars ships, aircraft and anything else that takes my fancy. My goal has always been to try and create as accurate a model as possible (photorealistic), and that is all in the smallest detail.

I haven't really purposefully designed anything for 3D printing, but if accurate modelling is your thing, Blender shouldn't stop you in that regard. I am currently designing a 3 axis CNC machine I plan on building. I have gone into obsessive detail to make sure nothing is overlooked that can come back to haunt me when I do get to build the real thing.

As someone who has also tried using blender, someone who is also not so good at organic models, I'm not sure blender is suited for practical, exact dimensions type models. Personally, I've used Sketchup for a decade, so that it what I've been using, but recently, I've been trying to learn FreeCAD. You might want to take a look at that. Sorry I couldn't offer you a suggestion for blender...

Blender's just fine for "practical, exact dimensions type models". Even more so now than it used to be, ever since they added boolean operations within edit mode that DON'T require tedious mucking about with modifiers. You can get sub-micron precision if you really want to, though I don't know of any existing 3D printers that actually present a need for it.

I model all my objects in Blender, and they're inorganic, hard-surface type things, often with components that require quite a bit of precision (like parts that screw together, or fit together with existing non-printed objects).

"boolean operations within edit mode that DON'T require tedious mucking about with modifiers"

Say what now? Can you elaborate? What version?

"I'm not sure blender is suited for practical, exact dimensions type models"

No. It is. Very much so. :)

In Edit mode, if you've got two or more separate hulls (e.g. if you add a primitive while in edit mode, or you've joined two separate objects), select one of them and use Mesh -> Faces -> Intersect (Boolean). No shortcut is defined by default, but you can define your own - or hit the spacebar and type "boo".

The first time you use it after starting Blender, it'll be the Intersection boolean type, as the menu name suggests, but you can switch it to Difference or Union in the tool properties (on the tool shelf). Subsequent invocations remember the last boolean type you used.

In Difference mode, the selected hull is used to cut away from all other hulls which overlap with it, but there's a "switch" checkbox in the tool properties that will make all other hulls cut away from the selected hull. If you have more than two hulls, you can hide any that you don't want to be affected.

Looks like the feature was introduced in 2.77.

Nice. Good explanation.
Seeing it again now I recall that I have used it before. I'll try to remember it exists so I can use it.

""I'm not sure blender is suited for practical, exact dimensions type models""

meh... I've been known to have been wrong before...

I agree that CAD software may be much better suited for mechanical engineering i.e. you can test the interaction of several moving parts right in the software - but for me (and I guess some others) it isn't that important yet.
I can see, though, why it will be very handy if the designs get more complex.

In my opinion Blender is still completly fine to get design parts with exact dimensions as long as you exactly know what you need and don't need much testing. I tried a few interacting gears and beside some minor tolerance problems, everything worked right on the first print.

Set a clock on a device to go off ever odd number of minutes throughout the day. Whatever you are looking at, write it down until you have at least 10 different items.

Pick one of the ten.

Make sure it is the most challenging one (except people or animals to start with).

Set a reasonable length of time to give yourself to model that design.

Don't design it to be for anything specific, be it printing or anything else. Just as a model for learning.

Do this once a week, and eventually, you'll learn how to do things on your own, and even search for new methods and shortcuts. (I still find new things after 7 years).

Best of luck!

(I also have some Youtube tutorials, but you're probably at or beyond what I've shown in a lot of them)

Many thanks for this advice.

I was stuck with the idea that each design has to be useful, cause thats what was most satisfactory in the end, but it restricted me to stuff I already knew (ideas that came to mind only used knowledge I already have)

The basic idea beeing: "I have no idea how to do this!" -> Thats where the learning curve comes from. Stuff doesn't work with my current skills or are hard/tedious to achieve, so I look up a way to get the result I want (either at all or more easily).

In the end, it sounds very obvious, but I really was overlooking it...

Conclusion: If I don't want to do it cause its too much trouble; that's exactly what I should do :)

Don't forget to save whatever models you do as well. That way, down the road, you can look back at your progress, or have a base model for something need that you can improve. Just like art and drawing, it doesn't matter what you draw, as long as you draw, you will get better with practice. :)

Exactly! Just doing it over and over to develope a routine for the workflow and to reinforce recently learned things.
Of course I'm keeping the models for future reference and to look back after some time to see how I improved :)

The whole video is relevant, but for a simulation in blender jump to 19:30