Saturday, May 03, 2008


PLA Corner bracket failure mode

Looks like we have a failure mode for PLA corner brackets. This picture shows a relatively recent bracket torn apart by the forces of fitting an 8mm rod into it and over-tightening one of the grub screws.

As you can see, the break is hardly a neat delamination, and the line of separation is at the apex of a teardrop-shaped hole. The corners also have little plastic in them, even after being squared off so they parted easily too.

As a temporary solution I am putting bolts through empty holes to hold things together. We'll see how it goes, but I think we may need to remove some of the unwanted honeycombing from the design and beef it up a little. I'm also installing the next one upside-down, so the apex of the teardrop is at the bottom of the bracket.

Vik :v)

Labels: , , ,

I read that PLA is more brittle than PET? That may not give it ANY expansion room for forcing. Just a random look on Matweb shows one PET variant (Eastmen Easar EN058) has only 3.8% elongation. Not much. Probably less with PLA.

Sounds like we need to print some samples and do an Izod and pull strength tests? I was looking for data online from ANY PLA manufacturer for the impact performance and strength data, but nothing popped up yet.
If the teardrop apex is a source of weakness you could use the truncated teardrop somebody suggested a while ago. The point is flattened level with the top of the circle. It relies on the filament being able to span a gap without slumping, but it my experience it can.
To make a truncated teardrop, besides filling in the tip in a modeling program, you can also try the Truncated Teardrop Shaper script at:

It makes a truncated curve. To try it out, just drop it into the ArtOfIllusion/Scripts/Tools/ folder. The default truncation is 0. A truncation of 0 means pointy teardrop and 1.0 means the truncation barely touches the hole. No one has used it before, a truncation of 0.5 would be a reasonable first test.

There may be an issue that if the fill is going sideways on the first layer of the truncation, it'll have little to stick to. Just in case of that problem, I'll add a fill direction feature to the script.
Enrique has already sent the script, but I've not had cause to use it yet. For the existing corner bracket, I might just flatten the apex by hand. Also, I could probably make them 2mm thicker and adjust the Z motor bracket accordingly.

Once you get the bracing around one and some bolts in it, it seems much more stable.

For the top layer, I'm thinking of replacing the stubs for the Y axis bearing support with bits of 8mm threaded rod and bolting it all securely together. These will provide more compression, and might even simplify the design of the non-running end of the Y axis.

Vik :v)
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?

Subscribe to
Posts [Atom]