I'm designing a mount for a cylindrical speaker to attach to my bicycle. It will mount on the bottle cages. I've printed a few iterations with various infill settings (using PLA) and the weak point is always the bolts holding the entire mount to the bike. They can't handle the compression needed to secure it properly. I had thought about using an imbedded metal part to distribute the load, but it's not easily replicable and I want to make the design public and fairly accessible to others with the same speaker. I currently have PETG and ABS, would one of those perform better or should I order a specialty high strength polymer?

  • 2
    $\begingroup$ it might not be a material but a design failure. Can you show how you orient the part in the slicer, and how you set the wall thickness and infill? $\endgroup$
    – Trish
    Commented Dec 8, 2019 at 0:21
  • 1
    $\begingroup$ I'm at work at the moment, I can upload some pictures when I get back home. It's a strut about 100mm long, on either end is a 74mm loop perpendicular to the strut. It's printed with the strut vertical, I'm just realizing that that's probably a major weakness. My priority was to keep the layer lines running along the loops so they wouldn't break when clamping down, but that means I'm bolting the strut against it's weakest axis. I feel like a goof, that's really obvious. $\endgroup$ Commented Dec 8, 2019 at 0:44
  • $\begingroup$ You can have "Different infill in the same part" to locally reinforce the structure to prevent compression of the structure. A picture of the part would help a lot! $\endgroup$
    – 0scar
    Commented Dec 8, 2019 at 7:46
  • 2
    $\begingroup$ Also, TPU would be perfect for something like this. $\endgroup$
    – towe
    Commented Dec 12, 2019 at 7:38
  • $\begingroup$ What @towe said. Try TPU with high infill and wall count so it's mostly rigid. It will be indestructible. $\endgroup$ Commented Feb 2, 2020 at 21:05

4 Answers 4

  1. Infill has minimal effect on the strength of printed parts, so I would expect the part to break in the same spot regardless of what infill percentage you used.

  2. PLA is especially poor in this exact application, and it undergoes significant creep/cold flow under mechanical compression over time, so even if achieved the necessary strength by changing settings (which you can), it would require that you periodically tighten the bolts more and more, as the PLA would slowly deform under the mounting pressure.

Perimeter width and number of perimeters are what primarily influence the strength of a printed object, infill has very little impact on strength in comparison, and unless you're using an exotic pattern like gyroid, what impact it does have is not even close to isotropic (will add strength in some directions while doing nothing in others). But even then, infill only really has an effect when we are talking about forces that are spread evenly over the entire object, not concentrated strength of a specific spot of the part. And that effect is always much weaker than what perimeter count or width will have.

Just bump up your perimeters to 4 or even more and that should make a huge difference.

And also, don't use PLA. I think PETG is a much better choice in this situation. It is more ductile only slightly less rigid than PLA, making it much more durable overall than PLA, and less prone to cracking under compressive forces. PLA theoretically has higher tensile strength, but that often doesn't mean much.

I would not recommend ABS, it tends to have similar issues with brittleness and is one of the weaker materials one can 3D print.

It terms of ordering a special high strength polymer.... unless printer and hotend is rated for in excess of 400°C, no such 'high strength polymer' exists, at least not that you can print. PLA and PETG are close to the best you can get, with Polycarbonate inching out ahead but not by a huge amount (~20%). Despite what filament companies would like you to believe, carbon fiber reduces the strength of PLA, ABS, PETG, PC, and probably nylon, and instead simply makes those polymers more rigid and increases dimensional stability. The only filaments that would actually be made stronger with added fibers short enough for filament manufacturing processes are ones with glass fiber. But you don't want to print those filaments, trust me. They will dull the teeth on your hobbed gear(s), even if made from steel, and will just ruin all but ruby nozzles very very quickly. And they still wear out ruby nozzles even then.

There are exotic polymers, but none of them print at less than ~350°C, and are generally exceedingly expensive. All the polymers that can be used at normal printing temperatures are all fairly similar to each other in terms of tensile strength at least.


I realize this is an old thread but I’d still like to share some thoughts for others who may find this post.

If I understood correctly, you were printing this part in to see shape You were printing this part in a C shape. Loops horizontal, strut vertical. Yes, that will produce a very weak strut that is likely to split open when tightened.

Next option is to print it in a U shape. Now you have a rock solid strut that will sustain compressive strength from bikts much better but depending on thickness and forces, your loops at the end may be prone to fail.

You may want to consider designing this in three parts, strut plus 2 loops, that you glue together. I know glueing doesn’t sound very sexy but 1) each part is printed with optimal strength, and 2) glues like Weld-On produce incredibly strong bonds, as strong or even stronger than inter-layer printed strength.

And lastly, it may also help to use overdimensioned washers on BOTH sides the strut to spread the compressive force of bolt and bike frame and prevent splitting.

  • $\begingroup$ I actually ended up doing something similar to that. Originally tried printing all in one part, two parallel rings joined by a rectangular strut. Decided to separate the strut and print them all flat, so the layer lines are aren't inline with the dress when mounted. Joined both rings to the strut with M3 bolt's. Can happily report it's been working for three years no problems. :) $\endgroup$ Commented May 11, 2022 at 3:36

For your particular application, you should be able to get something acceptably strong with PLA as long as you design it right - both the model itself, and the way you slice it (orientation, number of walls, etc.). metacollin's answer covers this pretty thoroughly. Using PETG could be an improvement, if you're willing to spend time learning how to print it right (if done wrong it's super-brittle).

However, as towe suggested in a comment, I'd give TPU a try. When printed with many walls or with high infill rate, TPU can be fairly rigid, and there's no way you're going to damage it with a bolt unless you just keep going after it's already tight. It's a relatively easy material to work with, and pretty much any printer that can print PLA can print TPU (although you may have to slow it down quite a bit) so it meets your requirement of being accessible to others with the same speaker.

(Note to self and others: it'd be a neat experiment to take a torque wrench to a bolt with metal washers through a solid brick of TPU and see how much it can take before it fails, and how it fails when it does.)

  • $\begingroup$ What do you mean by "print it right"? What would that involve? $\endgroup$ Commented May 3 at 20:54
  • $\begingroup$ @Technophile: Sufficient temperature for layers to weld, fully dried filament so it doesn't bubble, etc. If you want more details on printing PETG that's the subject for a question of its own, not a comment on this old answer to an old question. $\endgroup$ Commented May 6 at 16:20

Try heating op your PLA while mounting it. The PLA wil temporarily weaken and will become a bit moldable. PETG is a bit more brittle than PLA is my experience so I'd say stick to PLA. Otherwise you might try Arnitel ECO. I can try to find an amazon purchase link for you if you want. Anyway, it is more rubber like and stays flexible, it will never crack but depending on your design it might become a bit more wobbly. If you share a picture of your design or an STL i might be able to help you a bit better.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .