While there are many in the 3D-printing community who loudly and proudly proclaim never to have stooped to printing a 3DBenchy, there are far more who have turned a new printer loose on the venerable test model, just to see what it can do. But Benchy is getting a little long in the tooth, and with 3D-printers getting better and better, perhaps a better benchmarking model is in order.

Knocking Benchy off its perch is the idea behind this print-in-place engine benchmark, at least according to [SunShine]. And we have to say that he’s come up with an impressive model. It’s a cutaway of a three-cylinder reciprocating engine, complete with crankshaft, connecting rods, pistons, and engine block. It’s designed to print all in one go, with only a little cleanup needed after printing before the model is ready to go. The print-in-place aspect seems to be the main test of a printer — if you can get this engine to actually spin, you’re probably set up pretty well. [SunShine] shares a few tips to get your printer dialed in, and shows a few examples of what can happen when things go wrong. In addition to the complexities of the print-in-place mechanism, the model has a few Easter eggs to really challenge your printer, like the tiny oil channel running the length of the crankshaft.

Whether this model supplants Benchy is up for debate, but even if it doesn’t, it’s still a cool design that would be fun to play with. Either way, as [SunShine] points out, you’ll need a really flat bed to print this one; luckily, he recently came up with a compliant mechanism dial indicator to help with that job.

Thanks to [Keith Olson] for the tip.

Do you find supports to be annoying, when you use a 3D printer? A lot of time breaking away surplus pieces of plastic and then cleaning up the resulting ragged edges on your prints is certainly an unwelcome chore. But printing in free space is beyond the capabilities of even the most expensive printer, so it seems we’re stuck with supports for the foreseeable future. [Adam Haile] may have a solution to some support woes though, in the form of a clever technique for printing inset holes without support. His designs have a significant quantity of screw holes with inset heads, too far for the printer to bridge over so his technique breaks down the bridge into manageable smaller distances.

In the video below the break he shows how its done, with successive single layers that contain polygons bridging chords across the circle, with each layer approximating further to the final hole and the last holding the hole itself. Over a few layers the hole is created, without any support but with the minor inconvenience of a not perfectly flat inset. It’s a very clever idea, and one that we’d be interested to see further expanded upon by others.

At the dawn of 3D printing, support structures were something to avoid. ABS is a hard substance to clear off, and the slicers did a comparatively poor job of making structures that were easy to remove. Today, supports are not a big deal and most of the slicers and materials allow for high-quality prints with supports. We were printing something with supports the other day and noticed that Cura has a support floor and roof function. Curious, we did a quick search and found this very comprehensive post about the current state of support.

With 25 topics in the table of contents, this isn’t a 3-minute read. Of course, you might wish to skip over some of the first parts if you get why you need support and understand the basic ideas. We became more interested when we reached the geometry section.

The post talks about using the same material for support as most of us do. But it also covers using PVA or HIPS for support if you have a dual extruder printer. The other often overlooked method is to make a model that doesn’t require support, for example the post shows a statue that has design elements that make that possible. It also talks about using chamfered holes to prevent needing to support vertically-oriented holes.

Some of the advice is pretty obvious. For example, try to rotate the part so it doesn’t need support. However, some of it is pretty tricky, such as printing a mannequin at a 45-degree angle to reduce the amount of support needed.

Towards the end, they do talk about Cura-specific settings, which is where we found the answer about support roofing. Overall, a well-done post with a good mix of elementary information and advanced techniques.

If you’ve mastered support, you have no more excuses for not printing the Smithsonian artifacts or some of those cool NASA models. Oh, and the model in the picture? A sponge holder by [ajwammock].

Have a look at the object to the right. Using a conventional fused deposition printer, how would you print the object? There’s no flat surface to lay on the bed without generating a lot of overhangs. That usually requires support.

In theory, you might be able to print the bottom of the sphere down, but it is difficult to get that little spot to adhere to the bed. If you have at least two extruders and you are set up to print support material, that might even be the best option. However, printing support out of the same material you are printing with makes it hard to get a good clean print. There is another possibility. It does require some post-processing, but then again, not as much as hacking away a bunch of support material.

A Simple Idea

The idea is simple and — at first — it will sound like a lot of trouble. The basic idea is to cut the model in half at some point where both halves would be easy to print and then glue them together.  Stick around (no pun intended), though, because I’ll show you a way to make the alignment of the parts almost painless no matter how complex the object might be.

The practical problem with gluing together half models is getting the pieces in the exact position, but that turns out to be easy if you just make a few simple changes to your model. Another lesser problem is clamping a piece while gluing. You can use a vise, but some oddly-shaped parts are not conducive to traditional vise jaws.

In Practice

Starting with an OpenSCAD object, it is easy to cut the model in half. Actually, you could cut it anywhere. Then it is easy to rotate half of it so the cut line is at the bottom of each part. That doesn’t solve the alignment problem nor does it help you clamp when you glue.

The trick is to build a flange around each part. The flanges mate with a few screws after printing so alignment is perfect and bolts through the flange holes can keep the parts together and immobilized while your glue of choice sets. The kicker is that I even have an automated process to make the design side of this trick very easy.

You might cry foul. After all, this is just another form of support right? Not really. At least, not in the sense of support generated by programs like Cura or Slic3r. Slicers automatically generate support that uses a special pattern made to make it easier to tear away but it contacts all the surfaces that are overhanging. Unless you use a second material and a solvent that can attack your support but not your main part, you are going to have scars all over the part. With the flange method, you have a small number of beams that connect the flange to the part. The beams are easy to remove and while they may leave a little scar, they are easy to remove since they are small and only in very specific places.

Does it Work?

I’ve actually used this technique on a few practical projects. Although the part I printed for this example is just a test object, it shows the results of the technique quite well.

There are a few marks where the flange beams joined the main part, but they were easy to file away. If you had printed this in any orientation with traditional support it would have taken a lot more time and effort to get to a similar appearance, if you could at all. With the flange method, I simply applied some glue, inserted two screws, waited a bit, and then cut the flanges off with flush cutters. The whole process took under five minutes although some glues can take longer, of course.

I didn’t do any filing or sanding, so with more effort this could look even better. I also dinged the sphere a little bit pulling it up from the print bed (BuildTak works almost too well sometimes). However, the part still came out fine and prying a part off the bed aggressively is always a problem. It doesn’t factor into this technique.

Here’s how the parts looked coming off the print bed:

How To

Of course, now that you know this trick, you could just cut your models manually and build the flanges and support structures. That’s how I did the first one a few years ago but ever since I have wanted to automate it. Ideally, it would be great to have an OpenSCAD function that just “did the work.” I didn’t quite get there, but I did build a framework that makes it pretty easy. I put the entire file on GitHub.

The framework assumes that you have a module called part that defines your object and should be cut on the XY plane. Of course, if you don’t have that form, it is easy to wrap your code in a module and rotate and translate it to the proper point. Here’s the module for the test object:

// This is the odd-shaped part in question
module part() {
 union () {

difference() {
 sphere(r=20);
 translate([-20,-20,0]) cube([45,45,20]);
 }

difference() {
 translate([-9,-9,0]) cube([18,18,30]);
 translate([-2.5,0,0]) cube([5,25,15]);
 }

translate([0,0,30]) cylinder(r=5,h=5);
 }
}

If you prefer, you could call that module something like part0 and then write:

module part() {
 rotate([0,0,0]) translate([0,0,0]) part0();
}

Of course, you’d change the [0,0,0] parts to suit where you wanted to cut.

The rest of the OpenSCAD file has the code to cut your part in pieces, flip them, and add the flanges. There are some variables you can set to control things:

// Flange parameters
od=60; // outside diameter
odr=od/2; 
id=48; // inside diameter
idr=id/2;
flangeh=2; // height of flange
flangeboltr=1.9; // size of bolt holes
flangebeamw=2; // width of flange beams
flangebeamh=2; // height of flange beams (usually same as flangeh)
flangerotate=34; // rotation of flange beams

// part offset
offsetx=50; // put the other part this far away
offsety=50;
bigcutbox=1000; // box used to cut away half the model; just has to be bigger than model

Of Interest: The hull() function

If you want to dig into the OpenSCAD code, most of it is pretty straightforward. There’s only one part that is a little tricky. You can assume that the part is flat where you split it, but you can’t assume that it is solid. Initially,  I just built the flange and beams and merged them with the part.

However, in the test object, this doesn’t work well. See the cutout on the face of the box? If you just merge the flange, the beams will exist inside that cutout! That is hard to remove and serves no purpose, so it had to go.

Subtracting the part doesn’t work because the cut out is empty and subtracting empty space doesn’t help you. The trick is to find the maximum points of the part using the hull() function. Technically this is a convex hull, but I like to think of it as an envelope. The figure on the right is the result of using the hull function on the test part.

Armed with that, it is easy to subtract the hull from the prototype flange and then merge back the original part:

 difference()
  {
  flange(); // add flange
  hull() part1(); // but cut away "outline" of part
  }
 part1(); // now add part

Post Processing

You can control where the beams intersect the model by changing the rotation. You could also comment out some of the beams for many models. Fewer and smaller beams are better because it reduces the mess when you cut them off. On the other hand, if the beams are too tiny, they will break off when you remove the part from the bed, so there is a trade-off.

I used two small wood screws to hold the pieces together for gluing. You could also use a nut and bolt if you prefer. I usually start the screws or bolts but leave a gap. Then I apply glue to the parts while there is still a gap between them. Tighten the screws and then wait. Obviously, you have to do any surface preparation appropriate for your glue of choice. I was using DAP RapidFuse with PLA, but you may prefer other materials or glues.

I usually unscrew the pieces to make sure the glue held before I remove the beams. If you are certain of your glue, though, you could just cut the flanges free. A pair of flush cutters will make short work of the beams and leave very little residue if you use them right. If you were really wanting things to look good a little sandpaper or an emery board would vanish those beam marks easily.

Other Ideas

You can hack this idea a few different ways. For example, if your printer doesn’t like to make nice circles, you might prefer a rectangular flange. You might want fewer beams or more alignment holes. If your parts are hard to figure out alignment, you might put a mark on the flanges to identify which part goes with what. In the test part’s case, the bottom can rotate freely and it doesn’t matter and in most other cases I’ve done this, the part only goes one way and that way is very obvious.

I don’t know enough about glue science to know if you could do something interesting to the surface during printing to make it hold glue better. For example, you could create little interlocking channels to create more surface area for bonding. You could leave pockets for some sort of glue catalyst. Maybe flat is best, I simply don’t know. Maybe [Dan Maloney] can help.

We all know what the ultimate goal of 3D printing is: to be able to print parts for everything, including our own bodies. To achieve that potential, we need better ways to print soft materials, and that means we need better ways to support prints while they’re in progress.

That’s the focus of an academic paper looking at printing silicone within oil-based microgels. Lead author [Christopher S. O’Bryan] and team from the Soft Matter Research Lab at the University of Florida Gainesville have developed a method using self-assembling polymers soaked in mineral oil as a matrix into which silicone elastomers can be printed. The technique takes advantage of granular microgels that are “jammed” into a solid despite being up to 95% solvent. Under stress, such as that exerted by the nozzle of a 3D printer, the solid unjams into a flowing liquid, allowing the printer to extrude silicone. The microgel instantly jams back into a solid again, supporting the silicone as it cures.

[O’Bryan] et al have used the technique to print a model trachea, a small manifold, and a pump with ball valves. There are Quicktime videos of the finished manifold and pump in action. While we’ve covered flexible printing options before, this technique is a step beyond and something we’re keen to see make it into the hobby printing market.

[LonC], thanks for the tip.

3D printing is getting better every year, a tale told by dozens of Makerbot Cupcakes nailed to the wall in hackerspaces the world over. What was once thought impossible – insane bridging, high levels of repeatability, and extremely well-tuned machines – are now the norm. We’re still printing with supports, and until powder printers make it to garages, we’ll be stuck with that. There’s more than one way to skin a cat, though. It is possible to print complex 3D objects without supports. How? With pre-printed supports, of course.

[Markus] wanted to print the latest comet we’ve landed on, 67P/Churyumov–Gerasimenko. This is a difficult model for any 3D printer: there are two oversized lobes connected by a thin strand of comet. There isn’t a flat space, either, and cutting the model in half and gluing the two printed sides together is certainly not cool enough.

To print this plastic comet without supports, [Markus] first created a mold – a cube with the model of the comet subtracted with a boolean operation. If there’s one problem [Markus] ran into its that no host software will allow you to print an object over the previous print. That would be a nice addition to Slic3r or Repetier Host, and shouldn’t be that hard to implement.