Categories
Consumer Tech

3d printing some parts for the house

Intro
The end pieces of the pergola above my deck kept blowing off. I finally lost track of them and decided to 3D print a replacement to keep my skills fresh. I will use this as an opportunity to review my process for 3D printing.

The details
For people like me who don’t really know what they are doing, the kinds of objects you can 3D print is pathetically small. Yes, there are sites like thingiverse.com. But there are so many physical constraints if striking out on your own. If the extruder is pulling the filament horizontally it can only be unsupported for, I forget, 1 cm or so. That may be called a bridge. If you’re going from smaller to wider, the angle of increase cannot be greater than 45 degrees though I think 30 degrees is a safer bet. If you can want to violate either of those contraints you’ll have to include supports which you later saw away (yuck).

So those two constraints, combined, kill just about anything you’d ever want to print.

But, amazingly, my endplates could meet the criteria, if I made a slight compromise, so I went for it.

Openscad
I use openscad.org’s software to make a mathematical model of the object, built up from primitive geometric shapes (cubes and cylinders, mainly). When I learned you could write a simple program to generate your figure, and see the results rendered, I was like, sign me up.

Here is my openscad code, with some extra unused stuff left over from a previous design.

// DrJ 4/2020. Parameters in mm
  roundness=7;
  xsize=151;ysize=50;
  innerdelta=20; taperdelta=1.5;
  legthick=3; leglength=25;legheight=12;legoffsety=3;
  xteeth = 106; yteeth = ysize - 2*legoffsety - legthick;
  xhole=126;yhole=66;
  fullthickness=6; taperthickness=3;
  xreinforce=64;yreinforce=38;
  screwrad=1.6;
  divotrad = 5.1/2;
  Hhole = 15;
  screwholes=32.1;
  cablewidth = 16; cableheight = 10; cablex=16;
  cabley=-13;
  tinypropindent=9;
  // min facet angle and facet size (mm)
  $fa = 1.1; $fs = 1.1; 
 
  hull() {
  nice_rectangle(roundness,xsize-taperdelta,ysize-taperdelta);
  translate([0,0,taperthickness]) nice_rectangle(roundness,xsize,ysize);
  }
 
 // legs
 for(x=[-1,0,+1])
     for(y=[-1,+1])
        translate([x*xteeth/2,y*yteeth/2,legheight/2])
          cube([leglength,legthick,legheight],center=true);
 for(x=[-1,+1])
        translate([x*(xsize-2*legoffsety - legthick)/2,0,legheight/2])
          cube([legthick,leglength-12,legheight],center=true);
// thickeners
for(x=[-1,0,+1])
   translate([x*xteeth/2,0,taperthickness])
    cube([leglength+8,yteeth-1,taperthickness],center=true);
  //
  module nice_rectangle(roundness,xsize,ysize)
  {
   xcyl=xsize/2 - roundness;
   ycyl=ysize/2 - roundness;
   height=.02; // basically 2D - expect to 
  // be used within a convex hull
 
   union()
   {
    for(x=[-1,+1])
      for(y=[-1,+1])
        translate([x*xcyl,y*ycyl,0])
          cylinder(r=roundness,h=height, center=true);
    cube([xsize, ysize-2*roundness, height], center=true);
    cube([xsize-2*roundness, ysize, height], center=true);
   }
  }
Pergola endplate design

STL
You always make mistakes. The question is if you have the trouble-shooting skills to catch them before committing to printing… Anyway, when satisfied, render your object (F6), only then can you export to STL (stereolithography), which I guess is a universal description language.

Cura
Then you need a slicer! I think that translates your shape into movements that the 3D printer needs to make to lay down the filament, layer by layer. So import your STL file into Cura.

I think I set up Cura last year. I don’t want to mess with it. And don’t start playing with your model. It’ll do funny things to it. Just save as GCode. That’s it. Cura kindly estimates how long your print will take, and how much filament it will use up.

Anet a8 printer

It can be a royal pain to adjust the Z axis. Usually you have to adjust all four corners so that pulling a paper through gives just a little friction. The screws and assembly is so bad that you are constantly shifting between over-correcting and under-correcting. In this case I divided the problem in two because the plate was so narrow I just adjusted for the two sides, not all four corners, which is a nightmare. And I’m a little ahead of myself, but I have to say that this approach worked brilliantly.

Transfer the file to the micro SD card. Print, and enjoy! More fun is printing something with a lot of holes. The printer makes the coolest electric sounds as it quickly shuttles back and forth and to and fro. This endplate is pretty boring by comparison and mostly consists of long stretches of material.

End result

side by side

With its long flat section, the thing really adheres to the plate, unfortunately. I slightly nicked a corner prying it off with a screwdriver. But I think that beats the alternative of it not adhering to your plate. I’ve seen that once – and the printed thing just gets dragged around, which is fatal.

My piece came out a couple millimeters short in both dimensions. Still functional because of what it is, but still something I need to understand better.

Industrial design
I take it for granted that the PLA I’m printing is not as strong as the original plastic, so I should compensate by making edges thicker than needed. But I have to say the thing came out quite rigid and strong. This cheap 3D printer – if you treat it right – produces some high-quality output!

Bolt hole cover
In June I decided to print a bolt hole cover, also for the pergola above my deck. Modelling was pretty easy. I used a sphere for the first time to create the curved shape of the cover. Then I had to “get rid” of the rest of the sphere so I swallowed the unwanted portion of it up with a giant cylinder.

Here’s the openscad code.

// DrJ 6/2020. Parameters in mm
 
  innerR=8;outerR=11;plugHeight=2;
  epsilon=0.4;upperHeight=2.5;
  sphereR=26; big=50;
  tinyR=1;
 // min facet angle and facet size (mm)
  $fa = 1.1; $fs = 1.1; 
 
 
  difference() {
  translate([0,0,-(sphereR - upperHeight)]) color ([0.2, 0, 0]) sphere(r=sphereR);
  translate([0,0,-big/2]) cylinder(r=big,h=big,center=true);
 
  }
  translate([0,0,-plugHeight/2]) cylinder(r=innerR,h=plugHeight, center=true);
Bolt hole cover

I installed Cura v. 3.2.1 to make sure I had a version with support for supports, ha, ha. Most views don’t even show the support. I made sure I had it generate supports. Then I found that the layers view, when you drag it through the layers, shows the supports it will make. Some support options definitely would not have worked, by the way, namely, lines.

But, basically, supports are basically impossible to remove. I pried at it and rubbed with sandpaper but my vague hope that the supports might pop off were badly misplaced. The thing looks like a button. But fortunately the plug part was shallow so I just super-glued the thing onto my pergola and it looks like the real deal at a glance. In fact it looks better.

Mailbox plate

If I thought I was a whiz at this sort of thing, this mini-project proved me wrong. I was going to add a semi-circle at the bottom of a rectangle, and work through Pythagorean’s theorem to find the unknowns. I was just spending too much time and really, eye-balling it was more effective. And a more desirable shape was an ellipse. So the good ones make these parameterized models, but I find it’s not so easy and when you want to bang out some custom part quickly, it may not make sense.

There was a gaping hole in my mailbox post. This plate covers it up. Here’s the code.

// DrJ 7/2020. Parameters in mm
xsize=87;ysize=55; ylowest=65;
taperthickness=2.3;
// min facet angle and facet size (mm)
$fa = 1.1; $fs = 1.1; 
xreduce=8;
union(){
linear_extrude(height = taperthickness, center = true, convexity = 20, twist =0, slices = 20, scale = .98, $fn = 30){
square(size=[xsize,ysize],center=true);
}
linear_extrude(height = taperthickness, center = true, convexity = 20, twist =0, slices = 20, scale = 0.98, $fn = 60){
translate([0,ysize/2,0]) resize([xsize-xreduce,2*(ylowest-ysize)]) circle(d=30);

}
}
// bump out. Color blue.
translate([-xsize/2+35,-ysize/2+20,taperthickness]) color([0,0,1]) cube([10,10,2.9],center=true);
Mailbox plate
What techniques I learned from this simple mailbox plate project

I used the linear extrude method for the first time. Very useful. So, shapes like squares and cirlces (or ellipses) extruded along the Z-asix. My first effort did not include the bump-out, which I needed as a spacer. So I learned that you don’t always have to re-level after every print and this was the first time the dang filament didn’t break on me between print jobs – because I acted within 24 hours. I even re-used the blue tape. All real time-savers. So I used the 3D printer the way you imagine it to be – just there to print stuff, not to worry about fixing all the time.

I also had a reject – a first attempt without the bumpout. I kept it to test the breaking strength. So the thickness is 2.3 mm, right (I have to get calipers!)? It’s pretty darn strong. I was probably applying 10 pounds of force and it was hardly budging. Each layer prints orthogonally and a piece this thin is solid material.

Bird feeder C ring

I have a backyard bird feeder which hangs too low – I fear a squirrel could jump to it. So I wanted a way to shorten the hang by winding the wire around something that will stay. So I designed a C-shaped ring thing. works great. This is a first time for me for using a rotate extrude. I had to chop off the bottom of the ring so that I would not need supports. This is my shortest code yet, and least paraemerized. It’s just easier when you want to knock something out quickly…

// DrJ 7/2020. Parameters in mm
// min facet angle and facet size (mm)
$fa = 1.1; $fs = 1.1; $fn=100;
difference(){ 
rotate_extrude(angle=348,convexity = 10)
translate([20, 0, 0])
circle(r = 6);
translate([0,0,-9]) cube([60,60,10],center=true);
}
Bird feeder C ring

This is my first piece with interior fill. The bird feeder is pretty heavy – 10 pounds. but this little C ring is plenty strong for entwining the hanging string.

Time-saving tip for leveling the print bed

My anet a8 prints great once I get it set up. But that setup – boy is it a pain. Leveling the print bed being the absolute worst. Because, if you follow their advice and adjust each corner, well, they don’t seem to stay and you end up reversing your direction on the screws of the print bed. Or you’ll get one corner and then the corner you had just done will be off.

But think about it. For these small jobs, it only has to be level at the center. So, the heck with it, just make sure when you drag a piece of paper with the nozzle at the center that it has a bit of resistance, and don’t stress about what happens at the four distant corners. You will maintain your sanity this way.

Conclusion
I found the missing piece after printing this piece out, as luck would have it. BUT, then I found another place, right against the side of the house, that was never covered, but could have been. So I slapped my printed part there – it worked great!

A subsequent bolt hole cover project revealed that supports are nasty and basically impossible to remove. Yet the design was such that the thing worked anyways with a little superglue.

A plate to cover a hole in the mailbox post came next, and then a ring with a little slit for my bird feeder. Those two use an extrusion method.

You have to render your model or export to STL does absolutely nothing, and also does not show an error, except in the console, which you may not be showing.

References and related
Latest in 3D printing

This was the anet a8 printer. I don’t think it’s sold any longer, or not the cheapo $139 version which I was lucky enough to get. In fact someone bought one on my recommendation, parts were missing, and he never could get the missing parts. https://www.amazon.com/gp/product/B01N5D2ZIB/ref=ppx_yo_dt_b_asin_title_o01_s00?ie=UTF8&th=1

This is a great tutorial for openscad and constructive solid geometry: http://www.tridimake.com/2014/09/how-to-use-openscad-tricks-and-tips-to.html

Lots of free 3D designs are available at Thingiverse: https://www.thingiverse.com/

But there are a lot of other great 3D designs out there beyond just Thingiverse. This site introduces lots of these resources with example pictures from printed models.

Categories
Consumer Tech

What’s Next In 3D Printing?

In a lot of ways, the last few years have seen 3D printing “going mainstream,” so to speak. People are no longer astonished by the sheer capabilities of the technology, and we’re getting more accustomed to seeing stories about one interesting product or another having been 3D printed. People are also gaining more access to 3D printing; in some cases the costs of the devices have fallen, though we have posted about building one’s own 3D printer as well, and this is another option some savvier consumers are embracing.

The process of getting to this point, at which 3D printing is at least somewhat familiar to a lot of people, has been complex, fascinating, and enjoyable. But now that 3D printing is mainstream, what’s next for the technology? It’s hard to say with certainty, but we can look to a few current trends and developments and predict some of the ways that the technology could become more popular still.

Further Understanding Of Capabilities

Even if people are no longer astonished by what 3D printing can do in general, further understanding of the full range of the technology’s capabilities is likely to come with time. In a sense, the initial learning curve concerned the basic concept – that a 3D object could be printed into empty space by a machine according to a digital design. Next, people will inevitably come to learn about all of the various twists that can exist within that broad concept. For instance, people will get more used to the idea that 3D printers can create metal objects, and not only work with plastics; people will come to be familiar with larger 3D printers on a scale that can create car parts, or potentially even small homes and urban planning features. Ideas like these don’t necessarily fall into the most basic understanding of 3D printing, but they’ll inevitably be more widely understood in short time.

Combination With Other Advanced Manufacturing

This is perhaps more of a concern for large companies and significant manufacturing operations. Nevertheless, it’s a good bet that the next few years will see more people come to understand 3D printing as one of a few advanced manufacturing methods that can contribute to the creation of modern products. The technology’s most noteworthy counterpart may be injection molding, which is not as new, but which has grown vastly more sophisticated and capable. This is basically a process by which a given material is heated and molded to fit a shape, such that it hardens in that shape and becomes a final product. Fictiv describes injection molding processes as being ideal for design validation or for higher volume production, which largely conveys the idea that they are similarly useful to 3D printing. The truth though is that people may come to realize that the combination of the two technologies can benefit entire industries. For instance, there are some cases in which prototypes and early models are created via 3D printing (a better process for trial-and-error), and then used as the basis for large-scale molding efforts to design product lines.

3D-Printed Food

The very notion of 3D-printed food can sound absurd on its face. It reads very much like something from a science fiction novel or a fantasy film in which food can simply be created, and there are no more resource shortages. Amazingly enough though, while it’s not all quite so glossy and miraculous as the sci-fi version, 3D-printed food has become a reality. Hackaday explains how 3D-printed food can work in a piece written just last year, and the concept is actually surprisingly straightforward. Basically, pastes made of organic materials and broken down ingredients (such as, in one example, peas and seaweed) can be printed into mock versions of regular food items (in the same example, a steak). The same article acknowledges that taste isn’t necessarily ideal just yet, but the very idea of creating food products from a mishmash of edible materials is exciting. It’s a starting point for what could be a massive step toward addressing food shortages, and it’s likely to become one of the main topics driving further interest in 3D printing.

Command-Based Projects

This is something we don’t hear all that much about yet, but which could make for a significant step particularly with regard to consumer 3D printing. Singularity Hub took a look at 3D printing innovations we might see in the next five years or so and introduced this idea in an interesting way, stating that “hey” will become the “most frequently used command in design engineering.” Specifically the suggestion is that we’ll be able to say, “Hey 3D printer, design me a new pair of shoes” in much the same way we ask Alexa or Siri to make shopping lists or tell us the weather. If this proves to be true, it’s likely we’ll see a significant spike in consumer 3D printer purchases. For that matter, it could significantly change much of the retail industry.

Ultimately one of the cool things about 3D printing is that we’re all still figuring it out. It likely has countless applications we have’t yet considered or imagined. The suggestions here are grounded in current trends and speculation though, and should play a significant role in keeping this technology fresh and interesting.