Tip: Bed adhesion

[SandervG 1,521]

Good bed adhesion is one of the most important facets of 3D printing. When your print properly adhered to your bed you can sit back and relax while your print finishes. At the same time, when it does not adhere properly you are guaranteed of a fail. So, how can you ensure good bed adhesion?

A few ingredients are important. 

 

1) Bed leveling. 

Make sure your bed is level, and the space between your nozzle is not too much and not too tight. 

An Ultimaker 3 has active leveling which can help you achieve proper leveling. It is advised to manually level your bed before active leveling, so a potential offset is reduced to a minimum already. The less it has to compensate for, the better. For an Ultimaker 2, 2+ and Ultimaker Original you can use a calibration card supplied with the Ultimaker. The calibration card has a thickness of 0.15mm, which is exactly how much space you want between your nozzle and bed. 

 

2) The first layer

Ultimaker Cura has default settings for the first layer. In Cura, they're called 'initial layer' if you want to find them. This concerns among others the speed and thickness. We recommend printing your first layer at 20mm/s. A slower first layer gives the filament time to be squeezed onto the bed through the pressure it receives from the nozzle. Even when printing a high detailed 3D print, we recommend to print a thicker first layer, for the same reason as mentioned above. You want the first layer to squeezed on the glass plate. 

 

If the distance between your nozzle and build plate are correct and your first layer is printed at the right speed and thickness; your first print should look like #1. The layers are spread out and squeezed on the build plate evenly. 

In example #2 you can see some transparency in the first layer, this is usually an indication that the build plate was too tight to the nozzle. You can see near the edges how the excessive material is being pushed to the outside.

In #3 there was too much space in between the nozzle and build plate because of the roundish layers you see at the top. They are not squeezed on your build plate. 

 

3) Heat.

Heat can be a great source of adhesion. Different materials require different temperatures to ensure good adhesion. What temperature exactly may differ between filament manufacturers. For Ultimaker filament we recommend using the values proposed to you in Cura. 

As a rule of thumb, you could use these values as a starting point:

PLA: 60ºC - ABS: 80ºC - CPE: 70ºC - CPE+: 107ºC - Nylon: 60ºC - PC: 107ºC - PP: 85ºC - PVA: 60ºC - TPU:60ºC. 

 

4) Adhesives

Some materials are more prone to warping than others. Warping is when a material shrinks when it cools down. This shrinking can apply forces on the base of your 3D printed model which can result in the corners pulling up. To prevent this, you want to prevent a material from cooling down too fast (like by adding a door) and you can apply an additional adhesive. This can be a thin layer of glue applied with a glue stick, an adhesion sheet or other appliances like spray or partially dissolved material substances like ABS-slurry. 

Make sure you inform yourself if your material requires an additional adhesive. Your supplier should know. 

 

These tips should help you nail that first layer. Do you have any other tips or questions about your bed adhesion? Post them below!

 

 

 

[gr5 2,234]

Parts coming loose can be a disaster worse than just having a failed print so I tend to squish more than sander indicates.  All the above steps are important but squishing like photo #2 above results in parts sticking even more. So for large prints or if you are still new at this then aim for #2 until you are more experienced.

[geert_2 558]

In the beginning, 2 years ago, I did several tests on the effect of bed temp on bonding of PLA.

 

At 60°C bonding was optimal for me.

 

At lower temps, for average objects usually it still worked well until 50°C. However, at 40°C models were likely to pop-off suddenly, with a loud snap, in mid-print. Below 30°C there was no bonding at all, for my models.

 

At higher temps at 70°C the PLA became too soft. And then corners would lift and be peeled off the bed, due to the shrinking forces from the layers above, and the whole model would warp and come off.

 

Thus a balance has to be found between hot enough for a good bonding, and not too hot so it does not stay too soft. The default value of 60°C for PLA seems to work best on both of my UM2. These tests were done on light blue Ultimaker PLA, and on colorFabb white and orange PLA/PHA, but they seem to be valid for other Ultimaker and colorFabb PLA too.

 

As bonding agent for PLA I only used my "salt method", because this worked better than the glue stick at that time. But back then I didn't know yet how thick the glue layer had to be (probably I made it too thick?), and I did not wipe the glue with a wet tissue afterwards, which seems to equalise the glue into a very thin layer and improves bonding.

 

For PET (or is it PETG) from ICE, a bed temp of 90°C seems to work best for me.

 

Before printing real objects, I recommend you design an inverted prism, with a very small ground base, and wide top. This excerts huge warping forces on a very small base, due to the overhangs. And the edges curl up, causing the nozzle to bang into them. So it is a very hard bonding test. Print this model with various bonding methods, and bed temps, stay with the printer (!!!!), and carefully watch what happens, and try which bonding method works best for you. If you can print such extremes well, it should work for average models too. I could not print this test with the glue stick, nor on a bare glass plate: the models came off and produced spaghetti. But it worked with the salt method, although edges did lift, so it was on the edge.

 

Picture of a sort of inverted prism being printed. If I remember well, the base was only 2mm wide, but the top was going to be 10mm wide, and somewhere around 5 or 6mm high. And as you can see, at that time I also didn't know about atomic pulls and about worn-out white teflon couplers, so there is a little bit of underextrusion.

 

[Brulti 177]

On 20/02/2018 at 1:06 PM, SandervG said:

The calibration card has a thickness of 0.15mm, which is exactly how much space you want between your nozzle and bed. 

 

Question: is this also true for the UM3/E?

[gr5 2,234]

yes.  Well here is what I think is a more precise explanation for leveling on both UM2 and UM3.

 

The calibration card has a thickness.  Say it's 0.15mm (not 100% sure).  when you level with this the firmware is assuming the nozzle is this distance from the glass (0.15mm or whatever the thickness is).  At this point the printer realizes that the nozzle and the glass are 0.15mm apart so it stores this calibration information such that forever in the future if you tell the printer to go to Z=0.15mm it will go to this location.  If you tell the printer to go to Z=0 it should hopefully touch the glass.

 

Typically the bottom layer thickness is 0.27mm meaning the printer sets Z=0.27mm while printing the bottom layer and it also extrudes just the right amount of material to fit that gap.

 

I prefer to level the printer without the calibration card (although then I tweak the 3 screws again when it starts printing) such that when you tell the printer to go to z=0.27mm it actually prints about 0.12mm between glass and nozzle.  This squishes the bottom layer extra hard.

 

Alternatively changing bottom layer height in cura to 0.1mm seems to work very well also.

[gr5 2,234]

glass preperation

 

After "squish" the next most important thing is glass preparation.  Keeping it super clean is important.  Wash in soap and water and then dry AND THEN also wash with glass plus or similar window cleaner.  This will remove the last bit of soap and oil.

 

That's good enough for most prints but if you want it to stick even better I recommend a very thin layer of PVA.  The most common ways are:

1) 3DLAC - you spray this on the glass.  Easy.  But expensive compared to:

2) Hairspray.  I like aquanet unscented (blue can).  Make sure to remove the glass from the printer so you don't get PVA all over your printer and on the z screw

3) glue stick - for CPE you want a very thick layer so you can get the parts off but for PLA, ABS, Nylon and most other materials use a wet tissue to remove most of the glue and spread it very thin.

4) Wood glue.  I use Elmer's wood glue.  Mix it 10 to 20X parts water to 1 part wood glue.  I like to use a glass jar with wide lid.  Shake it up well.  Then open and use a small paintbrush to spread the milky liquid on the glass.  Heat to 60C and usually right about when glass reaches 60C it dries.  it should be thin enough that it's hard to see once dry.  Rinse the paintbrush afterward to keep it clean. 

 

The strangest thing about PVA is the thinner you make it the better parts seem to stick.  But obviously, at some point, you have no glue left.

#4 is my favorite method but I've tried 2 through 4.  They all work quite well.

 

5) Instead of PVA You can try salt water.  google it.

6) For ABS you can try ABS juice.  Google it.  Make it yourself.

 

I don't use any PVA glue for CPE as it can chip the glass (remove chunks of the glass) because it sticks too well.  I just clean the bed extra well.  Or I'm very careful not to squish the bottom layer much.  Or some people put extra thick glue stick down for CPE to keep the part from bonding to the glass too much.

[Brulti 177]

Side note on this, about the window cleaners: having made some test, most standard window cleaners you find in shops or malls are not to be used because, beyond cleaning, they deposit a thin layer of some anti-adhesive. Which is nice on your windows, and very bad for your printer. In my experience, avoid the cleaners that are blue, and most of the transparent ones as well. So far, the only one I've found that works well is the yellow one from Karcher: https://www.leroymerlin.fr/v3/p/produits/nettoyant-karcher-e1400355260

 

 

[Alex L 29]

I swear by Dimafix four our PLA prints to date. Works great and self releases once the bed cools down. Requires nothing more than a quick rinse under the tap to remove all traces of the adhesive.

[LePaul 331]

Good read, I often encounter what Sander showed as number 3...a tad too high.

 

@geert_2 good to know about PETG.  I just received some and everyone seems to think this stuff is better than PLA.  The datasheets don't entirely agree with that.  But good to have some bed temp info.  (Side note...from what I've read, it's very similar to ColorFabb nGen in  tensile strength)

 

I've wound up using hairspray as my goto surface for the past year.  Tried PEI sheet and it was a real struggle to remove prints.  I often had to put them in the freezer to pop free.  Then you wind up re-leveling the bed all over again.  

 

I got many months out of the diluted wood glue method but extra hold hairspray is just so much more convenient for me.

 

I use rubbing alcohol to clean the glass bed.

Edited July 12, 2018 by LePaul

[LePaul 331]

I meant to inquire...what's the salt method?

[tinkergnome 927]

3 hours ago, LePaul said:

what's the salt method?

 

wiping the glass plate of a 3D-printer with salt water

[geert_2 558]

12 hours ago, tinkergnome said:

 

wiping the glass plate of a 3D-printer with salt water

Yes indeed.  :)

 

Or see for the full PDF-manual and some other stuff here: https://www.uantwerpen.be/nl/personeel/geert-keteleer/manuals/

 

[Brulti 177]

After spending a couple days 'playing' with ABS and various sticky stuff, I will vote hands down for the UHU glue stick like what is shipped with the UM3E. It may stick a tiny bit less than all the spray stuff, but at least it doesn't hold on the part for the dear life when out of the printer. The bond between the glue stick and the ABS part breaks when cooling down, and I know when I hear a series of 'crack' sounds that the part is coming off on it's own.

 

All the samples of spray glue 'special 3d printer' that I got with various purchases of filaments and such proved to stick way too much, and were hard to clean even with warm water and soap. Glue stick dissolves even in cold water with a tiny bit of soap.

[TMicke 23]

Interesting read!

I've been using some kind of PET tape the last couple of years, it's the best thing ever, at least for PLA. It sticks super well and comes right off when cooled down, with smaller parts you can break away immediately after printing is done. I used the same tape for well over a hundred prints and I honestly didn't have one warping once the bed was level and settings in slicer optimized, as long as the first layer looked like Example #1 in top post you where good.

*edit* Example #2 also assured the print to stick, but with an annoying "edge" at the bottom of the print ? 

 

So I have a question; whats the downside to adhesives like the tape that I use? And today I read about all kinds of sheets, like gecko tek's EZ-Stik for example.

My former printer could not go over 210C so I was 'stuck' with PLA, so how it works with other materials I cant say.

 

My question obviously assumes that there is a downside, maybe I'm all wrong ? 

 

*edit #2* There is a very good reason for me not being an writer.

-Why I ask is because I get a feeling that most people use glue, hairspray or similar. Maybe I'm lazy but I think it's to much work with cleaning the build plate after every print.

My intention so far is to buy one or two extra glass beds so I can have different adhesive sheets for different materials.

 

Cheers!

Edited July 15, 2018 by TMicke

[Brulti 177]

I never needed glue for PLA. I use UM PLA, it stick very well on it's own on the glass bed of my UM3E. Did you had trouble with PLA that were solved with using PET tape? Do you use UM PLA or another brand?

[gr5 2,234]

For small parts a clean bed works well.  But for something that goes the length of the bed and heavy on the infill (or just skinny) and with sharp corners, parts are likely to pull off the bed.  For example if you printed a ninja throwing star with no brim that just barely fit in the build area (partly because of the sharp points).  Or if you printed a cuboid (a box) that is long and skinny like a pencil.

 

Also some parts barely touch the bed and get quite larger as you move up off the bed.  Something like this that runs the width of the bed will need brim and glue and heat and squish.  Imagine a typical pencil --  which is hex shaped in cross section as most pencils are. Imagine one of the 6 sides is resting on the bed but obviously it gets larger as you go up in Z.  With very strong forces on those upper layers pulling inward.  Especially if the pencil is as long as the print bed (say 220mm long).

 

[Brulti 177]

I see. I haven't printed many things that fit what you described, but this usually had some sort of support or the mickey ears for the sharp corners.

[geert_2 558]

Why no tape: for me the seam lines where tape strokes meet, would hurt. Most of my models need to slide along each other, sort of custom calipers for medical use. Also, putting tape down without bubbles, is not easy. And we all have had bad experiences with tape or labels that we couldn't get off cleanly (usually not related to 3D-printing, but in general). Maybe that is why a lot of people are a bit reluctant to use tape? Although in itself there may be nothing wrong with the method.

 

If your tape does not have any of these problems, feel free to post the brand and specifications and a few photos. We can always learn from it. Different materials may require different bonding methods. So it is good to have lots of bonding methods in our inventory.

 

Once people have developed a reliable method that they like, they tend to stick to it, because it is predictable and they master it well. This may outweight the possible benefits of trying other methods.

 

I mainly use my "salt method": this works well for my brands of PLA (Ultimaker and colorFabb), but not at all for ABS, and poorly for PET. So, for PLA, I just wipe the glass with a tissue moistened with salt water, and I gently keep wiping while it is drying into an almost invisible mist of salt. That is all. It takes maybe 10...15 seconds. Bonding of PLA is very good as long as the glass plate is hot (60°C). When cooled down, models come off by themself, without any effort at all. No cleaning required, no other messing, no need to take the glass out of the printer. Then I just wipe again with a tissue with salt water, and I can start the next print.

 

If I would ever need to print PE or PP, probably I would go to tape too, since these are very difficult to bond to glass, using glue.

 

[TMicke 23]

15 hours ago, Brulti said:

I never needed glue for PLA. I use UM PLA, it stick very well on it's own on the glass bed of my UM3E. Did you had trouble with PLA that were solved with using PET tape? Do you use UM PLA or another brand?

At the time I used Velleman PLA. When I started using the tape and it worked so well I never tried anything else. 

I had an vertex 8400 and it came without a heated bed using ‘buildtak’ adhesive sheets, and man did it stick to that, often it stuck too much and I ended up cutting pieces of the buildtak trying to remove my print.

Later some kind guys posted guides how to add an heated bed and made a new FW for it. By that time i was absolutley positively sure that you HAD to have some adhesive layer of some kind. 

Then I read about this guy telling about this PET-tape and at the time and in my ears it sounded just to good to be true, but I ordered a few rolls, 10 to be precise ? and never looked back. This was in late 2015 or early 2016. 

 

All the PLA I have tried has worked like a charm; Velleman, Eu makers, eSun, Primafila, colorFabb and so on. PLA with copper and wood worked equally good as did Ninja flex.

[TMicke 23]

@geert_2

Thank you for a very good explanation!

This salt method really sounds like a good thing.

 

You are absolutely right about the seam lines using a tape, even though I dont use my prints professionally it bothers my eye quite a bit actually. 

And then we have the bubbles ? Spot on again.

If you take the time to use soap water and a flat plastic thing to press the tape with afterwards to press out the soap water between glass plate and tape (cant find the words to explain this properly, but I’m sure you understand) it is possible to make it without any bubbles, and if you get some small ones they usually disappear once the bed is heated for a little while. 

 

I have ordered a ‘gecko tek EZ-Stik’ that I’m looking forward to try. 

But today, or rather tomorrow, I’m going to try the salty thing ?

 

*edit

Would you mind sharing how to make this salty water, I would appreciate it a lot hearing from a person whom I now have become to trust

Edited July 17, 2018 by TMicke
Added an request

[cloakfiend 996]

I'd also like to try the salt method, but I just use pritt stick power, (stronger than regular pritt stick and far stronger than the supplied glues). I change it maybe 3 or 4 times a year.

[kmanstudios 1,120]

I make a slurry out of the PVA waste that is from the prints I make. I have  also recently started to add alcohol to it to cut down on surface tension and it really lays out a much nicer PVA layer over the glass.

[Alex L 29]

I know I have said it already, but DIMAFIX is really, really good. Have had no problem with bed detachment, the coating is super even so doesn't have any adverse effects on the print quality, giving a good even sheen across the base, models generally self release after 5 mins or so in the fridge. Well worth the £10 spend.

 

The only downside is that you need a well ventilated space to apply the spray as it is pretty noxious, (although it smells amazing!) I just open the fire door and spray it outside to overcome this issue.

[DF-Werkzeugservice 9]

I print large ABS objects for industrial purposes. So besides a good bed adhesion, downtime on the printer is also a big factor to me. That led me away from any glue or something similar on the printbed, as a wanted to save the time from continuous coating and cleaning of the buildplate. 

I use BuildTak with two Flexplates that I switch after every print, which is working great for me. I know they're pretty expensive, but if you take care of them they last pretty long. 

Taking care means whenever putting a new sheet on, making sure to get the correct distance. If the distance is to big, it won't adhere. If it's to close it adheres to good and the sheet will get damaged when you remove the print.

A good way is to level the bed an put a Z- Offset in. Then make test prints, slowly lowering your Z- Offset until you get the desired result.

One downside is that the active levelling of the UM3 doesn't really work well, as the surface is kind of soft and gives inconsistent results, so I just stay with manual levelling.

 

You don't really need it for PLA, although this also sticks like hell to it. Before removing a PLA print, you should heat the buildplate to around 40°C, otherwise you'll pretty sure damage the sheet.

 

P.S.: I don't hold any shares on BuildTak, just like that stuff so much that I thought it should be mentioned here ?

[Hardlec 0]

A solution of vinegar and water will clean glass very well with no dirt-releasing additives that impede adhesion.

 

If you are worried about a tape seam, make a deliberate overlap of about 1 cm.  With an exacto knife or similar tool and a metal straight-edge, make a cut all the way through both layers.  Peel the tape back and remove the 5mm wide strips.  The wider pieces will fit back together very precisely.  

 

Stencil adhesive can be found in most craft stores. A  thin layer on the bed will encourage the plastic to adhere, but stencil adhesive will not form a permanent bond.  Stencil adhesive is designed to hold a stencil in place, keep paint from flowing under the stencil, and allow the stencil, often a delicate thing, to be removed and re-used.