inter-layer bond strength purely a function of contact area
110 views
Skip to first unread message
Whosawhatsis
Aug 17, 2021, 2:42:32 AM8/17/21
to 3DP Ideas
Just saw Sanjay from E3D talking about this in an interview: http://fullcontrolgcode.com/wp-content/uploads/2020/07/Author-version-Interlayer-bonding-has-bulk-material-strength-in-extrusion-additive-manufacturing-New-understanding-of-anisotropy.pdf
I've only skimmed the paper so far, but his summary was that inter-layer bonding strength seems to ONLY be affected by printing speed and temperature insofar as they affect the final contact area as one line of extruded plastic flows over another. Here's the interview: https://www.youtube.com/watch?v=PFQK9fPdwUs
This, of course, supports my argument for printing with extrusion width equal to nozzle diameter plus layer height (at a minimum, with layer height plus the width of the flat on the end of the nozzle as an obvious maximum). I've also been doing a lot of profile tweaking to eliminate the interstices between adjacent perimeters, largely for aesthetic reasons when printing with transparent materials, but also because it seemed obvious to me that more contact area should mean more strength. The few parts I've printed this way that have broken seem to have done so in ways that completely ignored layer lines.
I've only skimmed the paper so far, but his summary was that inter-layer bonding strength seems to ONLY be affected by printing speed and temperature insofar as they affect the final contact area as one line of extruded plastic flows over another. Here's the interview: https://www.youtube.com/watch?v=PFQK9fPdwUs
This, of course, supports my argument for printing with extrusion width equal to nozzle diameter plus layer height (at a minimum, with layer height plus the width of the flat on the end of the nozzle as an obvious maximum). I've also been doing a lot of profile tweaking to eliminate the interstices between adjacent perimeters, largely for aesthetic reasons when printing with transparent materials, but also because it seemed obvious to me that more contact area should mean more strength. The few parts I've printed this way that have broken seem to have done so in ways that completely ignored layer lines.
tray
Sep 30, 2021, 9:07:43 PM9/30/21
to 3DP Ideas
Interesting. Maybe you can answer this: Since the slicer asks for the extrusion width and layer height, why does the slicer care about the nozzle diameter?
Whosawhatsis
Oct 1, 2021, 1:57:49 AM10/1/21
to 3DP Ideas, tray
Depends on the slicer. In some, the nozzle diameter setting is used in leu of an actual extrusion width setting. In others, it is used to calculate default extrusion widths (which may or may not be equal to the nozzle diameter setting). It may also be used to warn the user if they seem to be setting the extrusion width/layer height setting too low/too high.
--
You received this message because you are subscribed to the Google Groups "3DP Ideas" group.
To unsubscribe from this group and stop receiving emails from it, send an email to 3dp-ideas+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/3dp-ideas/cdf113ef-54e4-4a87-b573-6e15ba43052en%40googlegroups.com.
Ryan Carlyle
Dec 9, 2021, 12:24:55 PM12/9/21
to 3DP Ideas
I've been thinking about this a bit, and I think you're likely to see some behavior differences between different polymers. Because FFF 3D printing deposition doesn't remelt the previous layer for a true fusion weld, to achieve full bulk property performance at the interface I think we have to rely on either crystal growth effects or diffusion bonding for adhesion between adjacent strands/layers. Simple van der waals type adhesion such as you see on a glass build plate isn't as strong as interlayer bonding, so I think there has to be something mechanically connecting the layers together.
- With semi-crystalline plastics, adjacent polymer chains have significant tendency to self-order into micro-crystal domains as they cool, and it stands to reason that the previous layer's exposed surface has crystalline domains which nucleate growth of those zones into the cooling fresh extrusion. If the crystalline zones grow across the interface, you should get very high mechanical performance at the interface equivalent to bulk properties. PLA and nylon both have this kind of semi-crystalline behavior. (As do most polyethylene and polypropylene variants).
- For amorphous plastics like PETG and ABS, there's no crystalline domains to speak of. Thus, diffusion bonding should dominate, where you get bonding because free ends of the molten polymer molecules are permeating into / tangling with the warm polymer molecules on the other side of the interface. Diffusion bonding is a "time at temp" process where the speed depends on temp (and the strength of the bond can be theoretically limited by rapid cooling).
Then, we also know that warping stress puts shear force on the interlayer bond and weakens it by "using up" some of the bond strength. So it makes a lot of sense that a low-warp, semi-crystalline plastic like PLA will have full bulk properties at the interlayer interface, subject only to interstitial void strength loss.
My question is, does the diffusion bonding of ABS and PETG happen fast enough to reliably reach full bulk strength? PETG tends to have high interlayer bond strength, while ABS is well-known for cracking at layer lines, but does ABS actually develop full bond strength and lose capacity due to interlayer warping shear stresses, or does ABS not develop full bond strength if it doesn't have enough hot contact time to get a good diffusion bond? The same environmental variables are going to affect both of these effects, e.g. hotter chambers or hotter nozzle temp have more diffusion time and also put more heat into lower layers so less warping stress. Can we separate these effects?
Bob Bilbrey
Dec 13, 2021, 8:50:10 PM12/13/21
to 3DP Ideas
I print lots of ABS mechanical parts - the low shrink stuff - with a fairly tight fit ooze shield. The shield in most cases maintains the part usually between 45-60 C. (I've inserted probes to confirm ). I experience minimal shrink issues and hardly ever any bed corner lift - no shrink splitting. But one still wants greater layer bonding.
Ryan, riffing on your assumption of an ABS/PETG diffusion bond mechanism I'm wondering if a medium temperature ( +/- 60C) post print heat soak of several hours would promote more layer grip. The temperature being low enough so as not to encourage dimensional distortion.
Whosawhatsis
Dec 13, 2021, 8:56:50 PM12/13/21
to 3DP Ideas, Bob Bilbrey
Reminds me of the salt re-melt method that some people have done with PETG. At a high enough temperature, it would definitely have an effect. The question is whether you can get the desired effect without softening the plastic enough to allow it to warp/collapse.
Another method that has been done is printing with a filament made of ABS with a polycarbonate core. After printing, it is heat soaked to soften the ABS and create stronger bonds, but it doesn’t get hot enough to soften the PC, which acts as a skeleton to keep the print from loosing its shape.
Another method that has been done is printing with a filament made of ABS with a polycarbonate core. After printing, it is heat soaked to soften the ABS and create stronger bonds, but it doesn’t get hot enough to soften the PC, which acts as a skeleton to keep the print from loosing its shape.
--
You received this message because you are subscribed to the Google Groups "3DP Ideas" group.
To unsubscribe from this group and stop receiving emails from it, send an email to 3dp-ideas+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/3dp-ideas/f69bcc35-8673-49a8-a57d-ceba74d97bd9n%40googlegroups.com.
Reply all
Reply to author
Forward
0 new messages