4

I have an idea for a 3D printed project, but I'm a total noob in this area and need someone to reality check it for me.

Basically what I want to do is a tricopter frame made of a tetrahedral honeycomb.

The whole model would be within 30 * 30 * 10 cm, the honeycomb edges would be approximately 1mm thick and about 15mm long and it would be printed out of nylon with SLS.

I have found Shapeways' design guidelines and my idea seems to fit it, but still it feels slightly more extreme use than what they had in mind.

Is this doable? Does anyone else use a similar method? Is there some software that I can use to generate honeycombs like this, or do I have to write it myself?

Edit:

This is how two layers of the honeycomb look like: honeycomb

In the actual model there would be several layer of this on top of each other and the shape would be kind of carved out of the honeycomb (+ some finishing to avoid spiky surface)

cube
  • 185
  • 5
  • 1
    consider designing and printing only corner connectors and then assembling the structure with match sticks or drinking straws. that will reduce the price significantly – fukanchik Apr 07 '16 at 15:29

3 Answers3

3

For designing your part, especially considering the repetitive mathematics involved, I would consider to learn to use OpenSCAD. I've learned the program and it fits your modeling requirement quite well. I feel it's easy to learn and is somewhat easier for folks who have a programming background. I don't have one, but it's still a logical progression to learn this program.

Regarding the SLS aspect, that also jumped out at me as a suitable answer. Solid shapes require to have "drain holes" to reduce the amount of powder consumed by the process.

I am assembling a Sintratec SLS printer and your model is the sort of thing I would enjoy to create with the printer. I've not yet listed my Sintratec printer on 3dhubs to solicit business, simply because it's not yet a fully assembled printer!

I did a quick Google search for "openscad tetrahedral honeycomb" and found this link:

http://forum.openscad.org/Beginner-Honeycomb-advice-needed-td4556.html

The result is more a polygonal honeycomb, not a true 3d tetrahedron, but it's a start. The file that created it is fewer than a few dozen lines of code.

The post is old enough that the internal links no longer work but the OP posted his module code and that does work:

enter image description here

I'm not sure how personal contact works in stackexchange, but I'd be willing to work with you regarding creating your code and if the printer ever gets assembled, printing out your part.

enter image description here

Adding a picture again, to show the latest revision, based on the updated information:

enter image description here

fred_dot_u
  • 10,532
  • 1
  • 10
  • 24
  • Thanks, that would be awsome. I just (very roughly) calculated how much it would cost to print at shapeways and came up with a value of about 1k USD. That's way out of my budget :-). For modelling I thought of doing the shape in SolidWorks, exporting to stl, and then replacing the volume with tetrahedrons using some script, but I'll have a look at openscad. You can find my mail address by following the github link in my profile :-) – cube Apr 06 '16 at 19:15
  • I've added to the original answer as I don't know how to generate an image here in the comments. How does this new image compare to your objective? It's sometimes challenging to determine what anothers' thought picture represents. The above model is 1mm high, 15 x 30 mm although OpenSCAD is measurement agnostic, mostly. – fred_dot_u Apr 06 '16 at 20:01
  • I added a picture of part of the honeycomb to the question. I couldn't do more because my computer started slowing down a lot. – cube Apr 06 '16 at 20:46
3

Is this doable?

It has been done, therefore it is doable.

I agree with fred_dot_u that OpenSCAD is a good system for programmatically generating highly repetitive 3d procedural content like this space truss 3d structure.

p.s.: A few links to people 3D printing various space trusses:

David Cary
  • 155
  • 3
2

In short, I don't think printing the full tetrahedral honeycomb design is a good approach considering the application of the part. Here are few things to note when attempting to 3D print the tetrahedral honeycomb:

I wouldn't recommend trying to 3D print this with a an FDM/FFF printer as you will most likely need supports and there would not be enough strength laterally. You may be able to print the design using SLA, but handling would be very difficult before post-processing as the part is very brittle post-print until a heat treat or curing process is done to chemically solidify. The post-process of the SLA could determine how strong the part is (ie. stainless steel powder, infused with bronze in a heat treat process would be good for such a part).

While SLS may be the best method for 3D printing this type of design, for that size part (30x30x10cm) you're looking at an expensive print regardless of whether or not you print it yourself.

Instead, I would highly recommend finding (or designing your own) a joint connector that would allow you to join wood/plastic dowels in the tetrahedral honeycomb shape. Not only will this be cheaper for you in the long run (easier to replace a few broken segments than an entire 3D printed model), but it could provide more structural strength for something that could potentially get banged around, like a tricopter.

For example, this model on Thingiverse (not my model) shows an example of how you can utilize 3D printing complex or custom joints that allow you to connect dowels in the shape you're looking for. It'd kind of be like building with K'Nex.

As far as designing said joint, you could model a single "inner" joint that has 18 connectors (8 on XY plane, 6 on YZ plane, and 4 on XZ plane). Below is a crude example of what I mean drawn in Google SketchUp: enter image description here

tbm0115
  • 6,234
  • 2
  • 18
  • 40
  • The post specifically asks about SLS, so what's the point of mentioning FDM and SLA? – Tom van der Zanden Apr 06 '16 at 20:55
  • 1
    I'll update the answer to include SLS. But to summarize my answer, I don't think that printing the full design makes sense either economically or structurally. SLS has the best chance, structurally, alongside high definition DLP, but that's a major expense still. – tbm0115 Apr 07 '16 at 01:29
  • Thanks for the answer. I roughly calculated the cost and it would definitely be too expensive. The underlying idea of what I was trying to accomplish is to make somehting fairly strong and very light using minimal amount of manual work. Vertex connectors and (probably wooden) pins would work, but it would kind of defeat the purpose :-) – cube Apr 07 '16 at 18:57
  • Are you completely sold on the frame being made out of a tetrahedral honeycomb? Even a basic truss should suffice if you're making a hobbyist tricopter, unless you're going for cool factor. If you go with a basic truss, your overall weight could go way down and the cost too if you print several panels of trusses and join them with lightweight dowels. – tbm0115 Apr 07 '16 at 20:29
  • If one creates a connector custom to this project, especially if done with an SLS printer, one could use carbon fiber rods for strength and light weight. Even a large quantity of connectors would be a reduction of cost to print via SLS, as well as fitting better in the build volume. – fred_dot_u Apr 07 '16 at 22:13