PETG: Bond Strength vs. Support Tradeoff?

The place to discuss filament...
Posts: 257
Joined: Tue Jul 11, 2017 3:38 pm

Re: PETG: Bond Strength vs. Support Tradeoff?

Post by Gwhite » Sun Oct 29, 2017 12:17 pm

The calibration I'm talking about is for the filament. Minor variations in diameter and melting characteristics means you won't get the extrusion width you are after. The simplest calibration is to print a one pass thin walled cube, and then measure the wall thicknesses. There are cube models all over the place (Thinigiverse, etc.). Set one for zero infill, 1 pass perimeter, and set the top & bottom layers to zero. That should get you a hollow cube.

That will get your extrusion multiplier in the ballpark. You may still need to tweak it up or down a little to get multiple pass perimeters to stick to one another well.

Posts: 34
Joined: Sat Jul 29, 2017 4:18 am

Re: PETG: Bond Strength vs. Support Tradeoff?

Post by GBuchwitz » Mon Oct 30, 2017 7:28 pm


I found/read/performed the extrusion multiplier calibration procedure... which yielded a factor of 0.83 for my M3 when using eSUN's black PETG.

Entering this value in S3D resulted in even weaker inter-shell layer bonds. Even straps printed directly on the build plate - which hadn't failed on me before - exhibited shell layer delamination when flexed even slightly.

Reverting back to 0.88 resulted in much stronger inter-shell layer bonds. S3D Tech Support recommended going with a higher hot end temperature, but I personally found that extruding at 241C yielded stronger bonds than 247C. This finding seems counterintuitive to me, but it's consistent with Jules' recommendations.

Aside from the notion that pushing a little more filament out for a set extrusion width might be better filling any voids between layers, I guess I really don't quite understand the relationship between the Extrusion Multiplier setting and the resulting bond strength - or why PETG shell layer bond strength seems to degrade when support material is used.

But at least gluing a direct-to-build-plate-printed strap to a block with a receiving slot gives me a functional (though less elegant) solution to my problem.

V/r, - guy

Post Reply