Bed adhesion, surface by surface

There is no build surface that is best. There is a surface that suits the material you are printing, and a different one that suits the material you print next week. Organising this by plate rather than by filament makes that obvious, and it explains why two people can give flatly contradictory advice and both be right.

One principle underlies all of it: adhesion is not glue, it is contact. The plastic must be soft enough to flow into the surface's texture, hot enough to stay there while it does, and clean enough to touch it at all. Every fix below is one of those three things.

Smooth PEI

The default on most modern machines and the best surface on this site for PLA, which sticks to bare smooth PEI at 55–60 °C without any adhesive. It gives a glossy, dead-flat bottom face that no other surface matches.

Its weakness is PETG, which bonds to it far too well. The stored warning is blunt: use a textured plate or a glue-stick release layer, or the part takes the PEI with it. Smooth PEI is also the surface most sensitive to skin oils — one fingerprint is a visible circle of non-adhesion.

Clean it with soap and warm water, not with isopropanol alone. Alcohol moves grease around and evaporates; detergent removes it.

Textured PEI

The forgiving option, and the one to reach for if you print mixed materials. The texture gives mechanical keying rather than pure surface contact, which means less dependence on perfect Z-offset and far easier release once cool.

It is the correct answer for PETG specifically: enough grip to hold the part, not enough to take a chunk out of the sheet. The trade is that the pattern is transferred to every bottom surface, so cosmetic parts that sit face-down want smooth.

Glass

Cheap, extremely flat, and the source of the best-looking bottom faces available if you are prepared to work for them. Bare glass does not hold much on its own, so glass is an adhesive-first surface — PVA glue stick or hairspray, applied thinly and evenly.

Two real hazards. Glass conducts heat well but unevenly if it is thick, so the edges of a large plate run cooler than the middle. And a part that grips glass hard enough can take a conchoidal flake out of it on removal, which ends the plate.

Garolite and PA-specific sheets

Garolite — glass-reinforced epoxy laminate, sold as G10 or FR4 — exists on a printer for one reason: nylon. PA6 does not stick reliably to plain PEI at any temperature, and it does stick to garolite.

The surface wants a light abrade with fine paper before first use and an occasional clean with alcohol. It is not a general-purpose plate; PLA sticks to it too well and leaves residue.

Polypropylene tape

Polypropylene sticks to essentially nothing except itself, which makes PP the awkward material of this whole subject. The standard workaround is exactly that: line the plate with PP packing tape and let the part bond to a sacrificial layer of its own polymer.

It works well, it is cheap, and it is fiddly — the tape must be laid without bubbles and replaced periodically. There is no elegant alternative, and anyone printing polypropylene regularly accepts this as part of the process.

Adhesives, and what each one is genuinely for

• PVA glue stick. Two opposite jobs. On glass and on ABS it adds grip. On smooth PEI under PETG it is a release layer, deliberately weakening a bond that would otherwise be too strong. Same product, opposite intent.
• Hairspray. A cheap sprayable equivalent for glass. Apply it off the machine; overspray on rods and belts is a genuine nuisance.
• ABS slurry. ABS dissolved in acetone, painted thin on a hot plate. The strongest option for large ABS footprints and the messiest.
• Purpose-made adhesion sticks and sheets. Worth it for polycarbonate and nylon, where the general-purpose options genuinely do not work.

If an adhesive is not helping at all, the problem is usually underneath it — a cold plate, a contaminated surface or a nozzle too far away. Adhesive not improving adhesion covers that case specifically.

The temperature you set is not what the plastic feels

A bed thermistor reads the heater, not the top of the sheet. A thick glass plate, a magnetic base and a spring-steel sheet all sit between the two, and the difference across that stack is not small.

Consequences worth knowing: a machine with a removable sheet generally wants a few degrees more than a machine with a fixed plate for the same material; a plate that is warped or poorly seated has cold regions that read as adhesion problems in one corner only; and raising bed temperature is a legitimate first move when a part lifts, up to the point where the base starts to bulge outward into elephant's foot.

Too much grip is a failure, not a success

A part that will not come off is not a well-adhered part; it is a plate about to be damaged. The rules that prevent it:

• Let the plate cool fully. Most materials release themselves as the plate contracts. Patience removes the need for tools entirely.
• Never lever with a metal blade on PEI. Flexing a spring-steel sheet is what it is for.
• Use a release layer deliberately where the material-and-surface pair is known to over-bond, PETG on smooth PEI above all.

If parts are already fused to the plate, that has its own diagnosis — and it is worth treating as urgently as a part that will not stick, because the repair bill is higher.

Matching material to surface