3D Print Time, Size and Accuracy

With the announcement of MakerBot's giant Replicator 3D printer, we're now entering a stage where most personal 3D printers have a significantly large build envelope, typically around 200mm wide by 200mm deep by 150+mm tall. This means we can print big things! But should we? 

 

Sure you can print something "twice as large", but what does that mean? Usually it means printing the a 120mm cube instead of a 60mm cube. In other words, when we say "twice as big", it usually means one of the axes has double length. 

 

This leads to a big issue that owners of large-sized personal 3D printers must face: print time. When you "double the size" of something you must consider that the actual object volume goes up twice x twice x twice = eight times the volume. This often means a print time of up to eight times as long! So your 47 minute print suddenly becomes six hours and sixteen minutes! 

 

But wait, it gets a lot worse. Everyone prefers higher resolution prints because they're visually more appealing with their smoother finish. The same equation results when you double the resolution: if you're printing at 0.5mm resolution and change it to 0.25mm resolution you've just created (more or less) eight times as much work to do! So even at the same physical size, a higher resolution print can take much longer. 

 

The situation is worst when you double the size of an object and also double the resolution: a theoretical 64 times longer to print. Of course these numbers are simply theoretical as the actual print time depends on a variety of factors, not the least of which is the number of shells, the complexity of the geometry and the fill density and pattern. Maybe it means we'll all be printing things overnight instead of while you wait. 

 

Nevertheless, the moral of the story is to carefully consider your print time when you get your brand new giant 3D printer. 

3D Printing Reality Checklist

An article posted on ExtremeTech showed up recently and introduced 3D printing to the general public in the usual overly-optimistic way: 

 

• a replicator and teleporter in every home
• You can plug it into your computer via USB, load up some freely-available 3D modeling software, and print stuff; it really is that simple. The only real barrier to mass adoption is the initial purchase price
• wireless versions that can sit on the kitchen worktop won’t be far behind
• it is becoming increasingly easy to simply download a 3D design, right click it, and press “create.”
• Imagine pressing the “bowl” or “cup” button on the 3D printer in the kitchen, followed by the “fork” or “spoon” button. It would even work for larger objects like cutting boards and colanders and laundry baskets
• With our current grasp of additive manufacturing, we could do these things now

 

Ok, that's enough quotes. There are a lot more in there along the same line. 

 

We believe in a 3D printing future, where yes, you can print things on demand. But that's just not true today. Articles like this tend to make the general public overly optimistic about 3D printing. The truth is that we will slowly develop more capabilities and features over the course of years, and gradually the public will enter the space and begin to print simple things, followed by more complex things as hardware and software evolve. 

 

All too often we speak to people who are truly enthusiastic about 3D printing, having seen or read similar reports in the mass media. But then we have to explain the actual state of affairs so they can understand where things really are at the moment. Here's some statements you can use when when you encounter someone whose head is exploding after first hearing about personal 3D printing:

 

• Big time commercial 3D printers can produce some really amazing objects, but home/hobby 3D printers are far less capable at the moment
• Print resolutions are pretty rough, analogous to the dot-matrix 2D printers of the 1970's. But they're slowly improving
• Not every shape can be printed. There are limits to the size of objects and the geometry of some shapes can only be printed on certain types of printers
• Simple objects are generally quite printable - but objects with internal or moving parts can be difficult or in some cases impossible to print
• Just because someone did a successful experiment printing in titanium/pasta/concrete/avocados doesn't mean you'll be able to. Current "stock" personal 3D printers are quite limited in the materials they can print
• You can't just press a button and get your object immediately; printing takes a long time. Often hours or even days in some cases, particularly for large or high-resolution objects 
• Personally affordable 3D scanning of objects produces only rudimentary models/sculptures 
• Designing 3D models for printing takes serious skill and sometimes expensive software

 

That should cool them off and keep heads from exploding. What did we miss?

 

We do think higher quality 3D printers will eventually emerge because there are tons of very smart, well-funded engineers working on ways to improve 3D printing today, but it will take a few years before we begin to approach the personal 3D printing vision that many seem to seek. 

 

Via ExtremeTech

Defeating Warping

If there's one thing that personal 3D printer operators dislike, it must be the evils of warping. Warping, the phenomenon that occurs when a coincidence of material, temperatures, extruder speed and stretching results in printed models that have curled up corners. Typically warping is much more prevalent with ABS plastic and this may have led some people to switch to PLA as their primary print material. 

 

But all is not lost. MakerBot has posted a very comprehensive list of all known techniques for minimizing warping if you insist on printing in ABS. Some of the techniques are obvious (slowing down your extruder speed to reduce the "stretch" factor) and some not so clear (adding anti-warp pods to the corners or reducing fill density). These may not eliminate warping, but they'll provide you with at least a chance of doing so. 

 

One more thing: We lied. PLA does in fact warp. It just doesn't warp as much, but it does come in some very cool colors. 

 

Via MakerBot

Essential Calibration

You've assembled your 3D printer after great effort and are anxiously awaiting your first printout. It will be amazing. It will shock everyone watching. Printing an actual object! 

 

But it turns out to be a sullen lump of misshapen plastic. 

 

This is the dilemma of many who assemble their own 3D printer. The effort is not so much in the construction (which can be a lot of work), but instead is the seemingly endless tuning required to "dial in" the right parameters that produce perfect 3D prints. 

 

There are hints here and there around the net, including a pretty good tutorial by MakerBot and excellent instructions by Dave Durant, as well as innumerable hints hidden in blog posts, forums and wikis. 

  

We found a very straightforward explanation in Thingiverse by Coasterman, who includes a series of "essential calibrations" that step-by-step lead you through the basics of tuning your 3D printer. The approach builds on the basic elements first, then follows with more challenging settings. Also included are all the objects you'll need to test with. 

 

If you're having trouble getting objects to hold together, we'd recommend you try these out. 

 

Via Thingiverse, MakerBot and Dave Durant

The Tribulations of Printing 3D Models

You've read yesterday's post about where you might find amazing 3D models for printing, but you're wondering what to do next. Is it as simple as downloading the model and printing it? Not exactly. Here's some considerations:

 

• Are you legally permitted to print this model? What license was it distributed under? Best to check before you get into trouble. 
• Some "3D" models are not at all suitable for printing. We searched for a 3D cat, and found several. But they were simply 2D planes with cat textures pasted on them. Not Impressive When Printed. 
• What format is the 3D model stored it? If it's not .STL, you're gonna have to convert it. This may involve using one or more intermediate programs and formats. Meshlab is your friend. 
• The model might include unprintable overhangs (that is, if your 3D printer doesn't include a means of printing support structures). Check your printer's overhangability - how much of an overhang angle can it handle? Does the 3D model exceed that angle? A careful visual spin around your model may tell you.
• Does the 3D model's utility depend on textures? What might it look like if printed in monochrome material? You might be terribly disappointed when you hold a bland print in your hand. Check this out beforehand.
• What size is the model? Will it fit in your printer's build chamber? Or perhaps it's microscopic? Regardless, always check the size and adjust accordingly.
• What level of detail appears in the model? Is it intricately detailed far beyond the 3D resolution of your printer? If so, you might end up printing a blobby mess. Select a model whose detail matches your printer's resolution.
• Is the model one single piece? Or is it a series of separate pieces. You might want to split them up or join them together first to simplify printing, depending on the nature of its design. 
• Those using extrusion-based printers might watch out for 3D models having a number of tower-like structures. Why? Because you'll end up with a significant amount of post-print cleanup work picking off stringy bits formed as the extruder moved between the high points. 

 

Yes, using a design not originally intended for 3D printing can be a lot of work, but if you get through it you might find yourself with a very unique 3D printout. Happy hunting!

Fixing Model Problems

One of the issues facing 3D modellers is the ability to create models perfectly suitable for 3D printing. A correct visual appearance does not mean it will print correctly, because the design might be ambiguous, have non-printable shapes or be "leaky" (non-watertight). Shapeways writes:

I've found that it's often difficult to retrofit your existing models for 3D printing, so it's best to start from scratch and with each step keep the goal (3d printing) in mind. This is especially true for models that you find online in model libraries - these are often made for visualization, not for 3d printing. Fixing these for 3D printing will make you want to pull your hair out!

Blog Casainho EMCRepStrap "About my RepStrap 3D printer based on RepRap and using EMC2 on GNU/Linux" had these non-manifold problems and found several useful resources to assist the repairs:

• Shapeways on Fixing Non-Manifold Errors
• Shapeways Video on Fixing Non-Manifold Meshes
• Shapeways Tutorial on Fixing File Errors
• The Meshlab open source system for processing and editing of unstructured 3D triangular meshes
• The Lucid Ubuntu package for processing and editing triangular meshes

Via Casainho EMCRepStrap

Cutting Modeling Costs

Sweet Onion Creations (whom we've covered before) has posted several very useful tips for reducing the costs of printing 3D models. The article focuses on the Architectural crowd, but the tips are useful for anyone printing 3D objects. Some tips we noticed:

• "... think about doing a hybrid model - This entails a complete model with the site and the building done at a small scale and smaller sectional models are built individually at a larger scale to show specific features of interest."
• "... saved tremendously on their modeling costs by hollowing out the model ..."

Via SweetOnionCreations

3D Printing Tip

An interesting forum post at 3DSociety.org talks about how best to reduce the "little sand-like jaggies on the surface" of your printed objects. Tim-Gibson writes:

Sometimes you can place your 3D model on the bed in a certain way so as to reduce the stepping on the flattest surface of you model. For instance if you were RPing a basic cube, a cube placed flat and parallel with the printing bed would be produced smoother than one you positioned at an angle with only one corner 'touching' the tray. make sense? kinda like you're going with the grain of the wood. Check with your printer.

Too bad we often need to print more than basic cubes, but a great tip nevertheless.

Via CGSociety.org