Need help designing an interlock between two 3D printed components

Hello, I’m designing an angled desktop picture frame which will contain a lightweight LED screen:

Attachment 13850
My question is, how to create a clean interlock between the base and the frame?

Attachment 13851

A few more details:

• The screen won’t be interacted with touch wise or anything, it’s just a free-standing display
• I would like the design to be as clean and minimal as possible without any need for supporting brackets in the back. The interlock should be able to easily support the cantilever force.
• The interlock doesn’t have to be designed for rapid assembly and disassembly
• I’m flexible on what program to design in (I’m using tinkercad at the moment)
• I don’t own a 3D printer, preferring to print at 3DHubs or other using a fairly inexpensive material such as Standard ABS (FDM)

I have read the guide here: How to design Snap-fit Joints for 3D Printing and have some ideas of my own, but this will be my first experience designing a snap-fit joint so I thought I’d reach out to see what ideas you have.

Thanks!

If you require that there be no bracing on the back of the angled support, it's critical that you have sufficient mass at the bottom where you plan the interlock. Additionally, consider the direction of the layer lines to ensure maximum strength in the design. As an example, consider the lowly cylinder, or in this case, an upright cylinder. Layer lines means the cylinder get printed as if it was a stack of coins. It's going to be just about as weak as a stack of coins, too, since it's easy to snap it into two pieces. If your display support is printed in the same orientation as the photograph/image, it will snap easily if not printed thick enough to handle the stress.

It's the kind of structure that would print better on edge, that is, rotate it 90° on the Y-axis. This would put the left edge on the print bed and the right edge at the very top, but the layer lines would be in the best position for handling the load.

And now to the joint.

I would create a slide-in type of joint for maximum strength and ease of printing and assembly. If you picture the base in the orientation I suggested, the layer lines would be set down in a straight line at the front of the part, left to right, with the back side of the part containing the joint.

From the top view, here's a coarse, simplistic drawing of how the bracket/base support could look. Obviously rounded corners, faired in edges, plenty of room for improvement.

Hi fred, thanks for that. Agree about the orientation of the layering, and I like your idea about a slide-in type of joint.For the slide-in joint, do you have any idea what tolerances I should use for a Standard ABS (FDM) joint? Or how I can find out? In other words, how much "play" should I allow in the parts for a snug fit?

I purchased a model print from Angus of Makers Muse to determine the print-in-place capability of my printer. I can print models with 0.4 gaps and expect the parts to move, but at 0.3 mm, they are stuck together. That's print-in-place, but separate print clearance is different due to "elephant's foot."

For your application, you don't have to have precise, tight fit and a small bit of play won't hurt the assembly. If you sand or otherwise cut away the elephant's foot, you could probably build a model with 0.5 mm clearance and have minimal problems, unless your printer is out-of-tune.

Consider also to create a block test of only the portions of the joint that are in contact. You need only to print it about 4-6 mm high for a good test piece. Short duration print, minimal filament use and you'll be able to check easily if it's safe to make the full print.

Sounds good. I don't own or have access to a printer, so i'll be ordering the print online, probably from 3D Hubs. Thanks for the info on gaps, that gives me a starting point. Maybe I'll split the difference and go for .4mm gap :)

Also, I just found this article on 3D Hubs which confirms your recommendation!

How to design Interlocking Joints for fastening 3D Printed parts

Attachment 13856

If you want to toss the test portions of the models at me, I can print them to check your clearance before you pay the big bucks to have them printed. If you're in the USA, I'd be willing to print your final model as well for a far lower price than the big operations would charge. What program do you use for your modeling?

Hi Fred, that's very generous of you!! I may take you up on that. Currently i'm using tinkercad online to design my models, but i'm open to other suggestions.

That's good news, as Tinkercad has nearly one hundred percent compatibility for 3D printing. I can't think of a single model that caused a problem for me or others in our makerspace. You can also send a private message to me with your shared link if you make your design public, once you've settled on a design or part to be tested.

We've been playing around with codeblocks internal to Tinkercad. It's a great feature if you want to be able to more easily adjust parameters for individual primitives, but as it is in Beta form, there's far too many features not working or nonexistent to make it useful enough for more complex models.

If you're not afraid of a little typing (okay, a lot of typing) and some low-level programming, OpenSCAD is a fun program to learn. Only one other of our makers likes that program, but he's a heavy-duty coder and could be expected to appreciate the package. The other members have this irrational fear of logic, math and related thinking processes.

Hi Fred, will do, i'm working up some designs right now. Glad to hear Tinkercad has a good reputation in your group. At first I thought it too simplistic, but now that I've gotten the hang of how it works, I enjoy using it very much. It's quite capable for a browser-based tool.

I am a coder type, so thanks for the tip about OpenSCAD, I'll keep that in my back pocket if I hit some snag in Tinker.

I'm surprised to hear there are maker types out there that would have fear of logic & math, but hey, to each their own! I'm sure they bring their own original insights a logical type like me would miss :)

why not just do it as a single print ? don't see why you need a snap fit at all.