A buyer opens a carton of printed PETG trays. The rim skin lifts at one corner and the resin certificate said compliant. Nobody kept a drying log, and that is where the food claim fails.
A PETG part is food contact safe only when the print record proves it, not the resin datasheet
Print settings decide compliance on PETG, not the resin bag. Hydrolysis, cold layer bonds and porous surfaces show up after a certified pellet becomes a tray. No incoming certificate catches any of those three failures.
Two shops buy the same lot and still ship different products. One dries to a fixed dew point and records it every shift. The other trusts a sealed bag, an unchecked dryer and an unlogged hopper.
This covers the settings that matter, the failures they prevent and the checks a buyer can run without a lab. It is written for sourcing and quality staff who sign off on printed parts each week.
Moisture in the filament degrades the melt before the nozzle reaches the part
PETG absorbs water from ambient air and holds it inside the amorphous regions. At melt temperature that water flashes to steam in the barrel, and chains split at the ester linkage. The damage cannot be reversed.
A dry spool extrudes steadily and smells faintly sweet. A wet spool pops, foams at the nozzle and lays a matte, cloudy bead. Both cues appear on the first layer, before anyone prints a test bar.
Buyers comparing offers across the PETG market should expect resin cost to follow glycol and acid contracts. A price fixed for a full quarter is a feedstock bet rather than a service commitment. Ask which index it tracks.
A drying schedule copied from another polymer will not hold on PETG
Suppliers routinely hand over a dryer setting borrowed from PLA or nylon. PETG sits between them in how readily it takes up water. The defensible target is a measured outlet dew point, not a timer.
Residence time matters as much as temperature. A spool left in a hopper overnight reabsorbs moisture from the headspace whenever the dryer cycles off. Late-shift parts run wetter than the ones printed at the start.
Ask for the dew point log and the hopper residence time in writing. Compare both against the print date on the parts. A supplier without either figure is describing an uninstrumented dryer and hoping nobody checks.
Bed adhesion on PETG is a surface problem before it is a temperature problem
PETG bonds to bare glass so firmly that it can pull chips out of the plate on cooling. That is not a win. The same aggressive bond on a textured sheet leaves a hazy footprint on the part.
Most adhesion complaints trace back to a plate cleaned with the wrong solvent. Alcohol lifts grease but leaves release agent behind. A single fingerprint on the build area creates a local no-bond patch every time.
The working sequence is a hot wash with mild detergent, a full dry, then a slow first layer. Raise the nozzle gap in small steps and stop at the first thin, translucent skirt on the plate.
Layer bonding, not wall thickness, decides whether a food part survives washing
A food tray fails at the seam between layers long before the wall gives way. Wash cycles, hot water and detergent reach that bond line and work away at whatever gap the printer left behind.
Printing inside the recommended temperature band is not sufficient on its own. Part cooling and layer time decide whether the previous layer stays warm enough to fuse, and both shift whenever the geometry of the part changes.
Cut a printed sample across the wall and study the fracture surface. A bond that failed by tearing leaves a rough, fibrous edge. One that failed by release is smooth and glossy, and that is the reject.
Stringing and ooze are early warnings that the melt is already compromised
Stringing gets treated as cosmetics. On PETG it marks the visible end of a melt that has taken on water or spent far too long at temperature. The same condition weakens every layer below it.
A pull test on a purge tower printed beside the part gives a cheap read. If the tower bends before it snaps, the polymer has lost molecular weight. No amount of post-processing or annealing recovers that lost strength.
Retest the filament itself before blaming the profile. Dry a sample spool, print the same file and compare the two pull tests side by side. If the stringing disappears, the dryer is the item to fix.
Hotend and nozzle alloys can void a claim the resin supports
Brass nozzles wear and shed copper and zinc into the melt stream. A worn nozzle also widens, which changes flow and pushes the operator to raise temperature. Both effects undercut the food contact claim.
Hardened steel resists wear but is not automatically acceptable for food use. Ask for the alloy designation. Ask whether any plated or coated surface sits in the hot path and can flake at elevated temperature.
The claim follows the whole melt path, from the spool right through to the nozzle tip. A compliant resin printed through an unknown third-party nozzle assembly is paperwork with one unverified component sitting in the middle of it.
Print orientation leaves crevices that no cleaning procedure will reach
A part printed flat hides its layer steps inside the part geometry. A part printed upright puts them on the food-facing surface, where residue collects in narrow grooves that a brush cannot reach into properly.
Sharp internal corners are the other trap on food-facing surfaces. Radius them in the model rather than the slicer, because a fillet added as a variable-width wall leaves a seam that opens after a few washes.
Specify the orientation on the drawing and make it part of the acceptance record. If the shop is free to rotate the part for speed, the surface inspected is not the surface anyone actually qualified.
A food contact declaration covers a grade, never the part in the box
A resin supplier declares a grade. That document names the base polymer, the colorants and any processing aid, and it covers material in original packaging. It says nothing at all about the printing process itself.
The distinction bites when a buyer moves from rigid trays to lidding film. A food contact barrier packaging programme tests the finished article for migration, while a resin declaration describes only the pellet that went in.
Keep the two documents separate in the file. When an auditor asks why the printed part is compliant, the answer is a declaration plus a print record plus a migration result. None of those three substitutes for another.
An incoming check a buyer can run without a materials lab
Start with the spool. Weigh it, record the lot, and check the desiccant indicator inside the bag. A bag opened weeks ago and left on a shelf has taken up enough water to matter already.
Print a test bar from each lot on a freshly dried machine. Break it again and record the deflection. Compare the result against the retained bar from the last accepted lot, not against an absolute figure.
For the finished part, check wall thickness at three points, inspect the bond line at a cut section, and confirm the orientation matches the drawing. Photograph all three and attach them to the lot record.
Most PETG quotes answer a different question than the one you asked
A quote that names a resin but not a dryer, a nozzle or an orientation is answering a material question. Buyers asking a part question get a price they cannot compare with anyone else's quote.
The same trap appears again in grade selection, and it costs more. A copolymer tuned for sheet extrusion will still print, but its melt strength and crystallization differ from a filament grade, so thick sections warp.
Send the drawing with the orientation, the wall callouts and the cleaning requirement attached, then ask every supplier to quote against that same document. Only quotes prepared on one common basis can then be compared fairly.
FAQ: questions buyers ask
Does a food contact certificate cover the printed part?
No. A resin declaration covers the grade as supplied. The printed part adds a melt history, a nozzle alloy and a surface texture, and those are what an auditor asks about when the migration result is reviewed.
How dry does PETG need to be before printing?
Dry enough that the outlet dew point holds at a measured value through the entire run, not just at the start. Ask for the log. A dryer with no dew point reading is a guess with a timer attached.
Why does PETG stick differently to glass and textured sheets?
The two surfaces release differently. Bare glass forms a bond strong enough to chip on cooling, while a textured sheet needs a slightly higher first layer and a slower speed. Cleaning residue is often the culprit.
Can a food contact part be printed with a brass nozzle?
Brass wears and sheds material into the melt path, and a worn nozzle changes flow. If the article must pass a migration test, specify the alloy and replace nozzles on a recorded interval rather than visible damage.
What belongs on the incoming record for printed parts?
Lot, spool weight, dryer dew point, print orientation, wall thickness at three points, and a photograph of the cut bond line. Those items let a buyer defend the part without paying for a full laboratory test.
Comments (0)