POM woes

So I'm trying to expand from printing just PLA and ABS into a bit more exotic materials on my i3v Prusa. The first one I tried so far is POM, as I need some gears and bushings and I thought that this material's toughness and low friction coefficient would be what I need.

But the low friction coefficient is causing me trouble as so far I haven't been able to find an adequate combination that will allow it to stick to the printbed reasonably enough to complete a print without extreme warping.

So far I tried (J-head 0.5mm, nozzle 225 deg, bed 130 deg); test object is Greg's extruder small gear:

-plain glass: doesn't stick at all, not even the first layer; ends up all curled and sticking to the nozzle
-glass plus hairspray: sticks in the middle but the edges lift as soon as the first layer is done
-glass plus blue tape: no sticking at all.
-glass plus sheet of manilla: see plain glass, doesn't stick at all
-glass plus manilla plus hairspray: a little better than glass plus hairspray, but edges still curl up during the second layer print
-glass plus textured cardstock, plus/minus hairspray: see manilla

I haven't tested Garolite as I don't have any and I've read that it usually needs machining to be used as the sheets you can buy aren't exactly flat. I'm not prepared to do that.

If you print POM, what works for you?

Have you tried PEI? I've heard (no personal experience) that that stuff works with just about anything.

Either that or Mutley3D's Printbite.

---

greghoge.com

HUGE 3D PRINTER PARTS SALE!!!

I haven't tried PEI yet, and I see no mention anywhere of it being used with POM. I have asked in the printbite thread about using it with POM but haven't received any replies yet.

One problem with both is they come in 6x6, 8x8 or 12x12 inches, but my print bed is 10x10. Either I have to buy a sheet smaller than my bed and limit what I can print - which would not be much of a problem, I only intend to use POM for small parts- or spend more money on a sheet bigger that I can use then have to cut it myself, which I'd rather not do (I heard that PEI is really easy to scratch).

Edited 4 time(s). Last edit at 12/05/2015 04:22PM by stamasd.

Mutley will cut you a custom sized piece I'm fairly sure of that. But no experience with POM myself.

Actually I've just looked at the pdf on printbite's site, they mention acetal (=POM) as one of the materials that cannot be printed on it.

It however mentions printing POM on glass covered with acrylic lacquer... I'm off to the car parts store.

Edited 1 time(s). Last edit at 12/05/2015 04:44PM by stamasd.

stamasd - yea i will cut a piece to size for people upon request, but as the site says, it wont work with acetal/POIM/delrin

POM is Acetal is also the 3M brand material called Delrin - the problem with it is the shrinkage factor of nearly 3% compared to 0.2-0.8% on ABS and PLA etc. The POM shrinks itself away from its bed adhesion, ie pulls itself up before it can complete a bottom layer. I have had some success using the lacquer but it is limited and problematic. A print friendly POM is required, with less shrinkage, but the compound that gives its properties inherently shrinks, take that away and its not POM anymore (at a chemical chain level).
Hope the above info helps. keep us posted on how you get along.

Well I have one of my spare glass beds coated with a heat-resistant enamel (the ones not marked as "heat resistant" are only rated to 93C/200F). Dunno if it's acrylic or polyurethane, the clerk at the store couldn't tell me and it's not marked on the can. It's not clear either, they only had black and silver - I went for the silver. We'll see tomorrow or the day after, when it dries up. I plan on curing it in the oven for an hour or so before using it in the printer.

I actually picked up several different cans and will experiment with them. I'll post back here with details if I find something that works.

Edited 2 time(s). Last edit at 12/05/2015 09:00PM by stamasd.

Has anyone tried mechanical hold downs like one would use on CNC machines? For example, print 3 or 4 layer tabs and clamp those to the print bed.

Could we print POM on a sheet of POM?
I'd then print the part with a raft and hope to separate them easily later.
-Olaf

Quote
o_lampe
Could we print POM on a sheet of POM?
I'd then print the part with a raft and hope to separate them easily later.
-Olaf

I have read somewhere that it has been tried, and it works, but separating the printed object from the sheet (even with raft in place) is very likely to damage the sheet therefore you don't get much mileage out of it before it needs replacing.

Quote
etfrench
Has anyone tried mechanical hold downs like one would use on CNC machines? For example, print 3 or 4 layer tabs and clamp those to the print bed.

POM warps very much, vary fast. In my attempts so far I see it beginning to happen even before the first layer is finished. By the time the second layer is down, the edges have already lifted by half a millimeter or more.


I have in the meantime asked a couple of suppliers of PEI (one on Amazon, another one on ebay) about printing POM on it, and both have replied that either they haven't tried, or that they have no information on that.


My plate should be ready later today and I will see what it does. Alternatively I could try some kapton tape, but the problem is I only have it in 20mm width and it will be a chore to cleanly cover a 250mm plate with it...

Edited 5 time(s). Last edit at 12/06/2015 08:25AM by stamasd.

I did some test with this filament: [www.3dcpi.com]
After some tries, the only kind of pieces I can print correctly are bushings.
I use IKEA mirror and a good coat of Imedio-UHU glue stick:
[plus.google.com]
For gears, I get betters results with Taulman 645:
[plus.google.com]

Printed with Catnozzle V2.3 hotend without fan.
POM: 235 hotend - 115 bed
Nylon: 245 hotend - 100 bed

Be carefull, POM fumes are very dangerous.

How hot do you need an enclosure to be for POM?

Why not just use PC filament - its hard as nails and slippery

You could also use Igus filament - again hard as nails

Quote
Mutley3D
Why not just use PC filament - its hard as nails and slippery

You could also use Igus filament - again hard as nails

I haven't had much luck with all-metal hotends (I actually built my printer originally with a Hexagon hotend bought from Colin at Makerfarm, but it went bad very quickly - the metal itself became warped to the point where filament wouldn't pass through it anymore; then another bad experience with another metal hotend has soured me on all-metal). Currently I'm using a J-head as it has proven very reliable, but doesn't print PC.

And Igus is very expensive. The prices I see are around $300/kg.

As for my coated glass, I haven't tried it yet. I did try kapton tape, with the result that around the 3rd layer the whole piece detached and slid off.

Edited 2 time(s). Last edit at 12/06/2015 08:23PM by stamasd.

I've had limited success printing a POM object today.

To get this out of the way, the glass coated with enamel didn't work. At first it looked OK, but then around the 3rd layer I noticed that the object was actually rocking as the print head was going around it, as it had completely detached from the printbed. Failure.

What did work though was the following. It wasn't a complete success, the object did not detach and the print completed succesfully but upon inspection had a serious flaw - which would prevent it being used.

The test object was, again, Greg's extruder small wheel. It's small, moderately complex but not overly so. It's also representative of the type of objects that I plan on using POM for.

The settings were: J-heat 0.5mm, nozzle 225 degrees, bed 120 degrees, otherwise default settings for ABS for my 10-inch Makerfarm Prusa i3v.

The printbed was a sandwich comprised of, from bottom to top (likely not all are necessary but I had to start with what I had in a pinch, as I didn't want to mess with my last spare glass plate):

1. glass, 3mm
2. Elmer's glue
3. thin sheet of manilla cut from a folder
4. two thick layers of hairspray
5. kapton tape
6. a thick and rough layer of ABS glue. This was quite messy because my ABS glue had thickened up, I didn't have any acetone to thin it with and I was impatient.


The printed gear stuck fine on top of this sandwich, with no detachment or visible warping. There was some stringing but not too bad. At the end of the print, the place where the printhead left the object was left with a little cone of material, which would require sanding off. The main hole and the side hole printed very well, The small rectangular hole had some material over its opening, likely ABS glue which stuck to the bottom of the piece. In fact the adhesion to the ABS glue was so strong that it came off with the object from the bed.

The object is however not usable for its intended purpose. Most of it printed well, but there's a whole spoke missing. Not a nub, not a bump - it looks like the printer failed to print a whole spoke altogether. Instead of 9 spokes, the wheel has only 8. The one missing is the one located directly above the side hole. I'm not sure why it happened, I checked the .stl and it looks fine. I didn't have time to check the gcode.

Other than that, the gear seems to have printed well, the other spokes are well made and clean, it appears very solid and the layers have fused well. I had a very tough time trying to detach the ABS glue from the bottom. Acetone? I have to buy some.

I'll post some pictures soon.

Suggestions?

Edited 1 time(s). Last edit at 12/07/2015 05:16PM by stamasd.

Pictures as promised:

1: gear being printed. [dl.dropboxusercontent.com]
2. after printing finished still attached to bed [dl.dropboxusercontent.com]
3. same from different angle [dl.dropboxusercontent.com]
4. detached (sorry for the lighting, phone camera) [dl.dropboxusercontent.com]
5, 6 zoomed in [dl.dropboxusercontent.com] [dl.dropboxusercontent.com]
7. bed after object was detached. All of the ABS glue on top of the kapton came off. [dl.dropboxusercontent.com]

8. spare glass coated with heat-resistent silver enamel [dl.dropboxusercontent.com]
9. printing on the enameled bed; Seemed fine at first [dl.dropboxusercontent.com]
10 however it quickly detached and slid off. The failed object is turned upside-down to show bottom. [dl.dropboxusercontent.com]

Small update, I found a seller on aliexpress that has 1mm PEI sheet and will custom-cut it to my bed size; the price is good, so I ordered two. But it will be shipped from China, so I don't expect that I will have it before 2-3 weeks. I am otherwise very busy this week, so I don't think I can attempt another print until next week anyway. I am however encouraged by the strong adhesion I got to the ABS glue to continue experimenting with it.

So I did 2 more attempted prints.

1. Used the enamel-coated bed plus a layer of kapton, then a thin coat of ABS glue. The object detached at about 50% completed and started rocking under the print head. It looks like the point of failure was between the kapton tape and the enamel underneath, or between the enamel and the glass. At this point I scrapped the enamel as a usable printbed layer altogether.

2. Plain glass with 2 thick coats of ABS glue. This worked the best so far. No detachment of the object, though there still was some warping on the bottom. Print completed with the same defect as in the test from December 7th, a spoke missing on the wheel. The STL and gcode look fine, I don't know why this is happening. I welcome ideas. If none comes, I will give it another try with reduced print speed (70% perhaps) and the object rotated by 45 degrees on the print bed to see if it's a problem with the printer or with the software. If there's again a missing spoke in the same location relative to the side hole then it's a problem with the software. If there's a missing spoke in the same position relative to the printer or no missing spoke, then the problem is with the printer.

Quick update: I received the PEI sheets I had ordered. That was much quicker than I expected, coming from China. They are 250x250mm, 1mm thick. The surface is scratched but not badly, I don't think it will impede function. The scratches are very superficial, none seems to be more than 1/10mm deep. They are not perfectly flat either, having a slight concavity. They are quite flexible though, so they could be clamped flat onto another surface.

My question is: to make a bed, should I place one of these sheets on top of my existing glass plate (increasing thickness to 3+1=4mm), or remove the glass and place the PEI directly on the heated bed (for a total thickness of 1mm)?

The replies to the questions I had in this thread lately have been positively underwhelming. I will go ahead and do what I think. If/when I get results I may post updates.

The difficulty with acetal is whilst being pretty strong it also has a very high shrinkage rate. At about 3% its probably the highest shrinkage rate out of all the materials that Im currently aware of as being available in filament form. Most other materials are are around 0.5 to 1.5%.

I wouldnt be put off by the lack of replies, there are plenty viewing the thread and so the interest in finding a solution is there. A few people (including myself) have tried many surfaces with this material but with limited or no success.

Sometimes it can take many people many attempts to crack a nut until someone comes up with a novel or well tuned solution.

Anyway, here's the answer: POM does NOT stick well enough to PEI. My best out of 6 tries went to layer 10 then detached. Nozzle 230 degrees, bed 120 degrees (1mm PEI on top of 3mm glass), print speed 20mm/s.
The same bed works amazingly well for ABS and nylon though so I don't regret the purchase.

Edited 1 time(s). Last edit at 12/20/2015 11:57AM by stamasd.

To sum the past few days of experiments.

The ONLY way I found to keep the POM reliably attached to the printbed is with a thick layer of ABS glue. Using any other method has resulted in sooner-or-later the object detaching while being printed. The simplest system that worked is a clean glass plate coated with ABS glue.

Downsides: the adhesion to the ABS glue is very strong, and the glue will come off the plate with the object. It's easy to clean the glue up from the object, but it's still a hassle because to print the next object you have to let the bed cool down completely, reapply ABS glue and have it dry at room temperature (applying ABS glue to a hot bed results in bubbles and gross irregularities in the layer). This slows down printing rate considerably.

The bottom of the objects after cleaning shows some warping, but not as bad as without the ABS.

So far so good. But there is another problem I found with POM: inconsistent print resolution along the X axis. Basically the greater the X, the better the resolution. At small Xs, i.e. the right side of the object as it sits on the platter, the loss of detail is so bad as to lead to loss of function. Remember my post above where I mentioned the missing spoke on the extruder wheel? I printed another wheel, this time rotated on the platter by 45 degrees CCW. There is again a spoke missing. It's the one that's oriented towards the right side of the wheel. There is also considerable loss of definition to the 2 spokes adjacent to it. The other spokes are well printed.

I have watched closely what happens at that place to result in this effect. When the print head gets to that spoke, which is the last one printed on the layer, it pauses briefly and the POM forms a little blob there; the blob attaches temporarily to the side wall of the print head. As the print head moves to the left next, it draws the blob to the left and therefore the spoke doesn't get printed correctly. The viscosity of the POM at 230 degrees is fairly high, I wonder if that's what's causing this effect.

When I have time again I may attempt printing the same object again at a slightly higher temperature; perhaps a lower POM viscosity will help? I can't go too far up though, my hotend is a J-head and I've heard of bad things happening to J-heads above 240 degrees.

Any input appreciated.

Edited 2 time(s). Last edit at 12/20/2015 06:25PM by stamasd.

I'll have to get myself a roll to try. Delrin is great to work with on the lathe, if I could print with it as well that would be really useful.

One thing I'm not clear on, are you printing with an enclosure or out in the open? Trying to deal with high shrinkage in the open sounds like a pain.

Quote
JamesK
I'll have to get myself a roll to try. Delrin is great to work with on the lathe, if I could print with it as well that would be really useful.

One thing I'm not clear on, are you printing with an enclosure or out in the open? Trying to deal with high shrinkage in the open sounds like a pain.

Open.

Something else I've been playing with. Unfortunately I won't have much time to experiment for the rest of the week, but I thought I'd put it out there, perhaps it will be useful for someone else.

For printing difficult/slick materials I came up with something which I call "hotswap". Esentially the idea came to me observing what happens when you change the filament in the hotend. Usually a small amount of the previous material remains on the walls of the hotend bore, and will be pushed out by the new filament at the beginning of the next print. Sometimes you get a mix of the two materials that gradually converts to the new material over the first and even the second layer.

So my thinking is: perhaps we can use this to our advantage to print materials that don't stick well to a standard printbed. I can get POM to stick reliably to glass now, but it's messy and involves coating the glass before each print with ABS glue. What if I printed the first layer of an object in ABS, followed by a transition ABS/POM mix then POM (as an example - but you can imagine it being applied to other material combinations provided that they are compatible and will mix)? I can print ABS very reliably with no warping at all on PEI, and I have seen very strong adhesion of POM to ABS glue, so I assume these 2 materials would be compatible.

I plan on replicating the above observed phenomenon on purpose by inserting a small length (5-10mm) of ABS filament directly into the hotend before loading POM filament from spool into it, and attempt to print this way. Probably it would be even better to use different colored materials ( i.e blue ABS with white POM) to see exactly how they mix and what the composition of the layers will be. I hope that I will get strong adhesion to the bed and minimal warping this way.

A cross section through the hotend loaded and ready to print would look like this:



Even better, if a gradual transition from material 1 to material 2 is needed, the profile would look like this:



This may be a little difficult to obtain; ABS is easy enough to cut at an angle, but POM is very tough; harder than PLA and much less brittle. I usually cut it with a small hacksaw, and doing it at a precise angle would probably require that I make a custom jig . Plus there's the issue of trapped air between the 2 materials, which makes me think that they'd better be glued together first, perhaps with a rubbing of acetone. Then it'd need to dry for a while before being used.

As always, comments/suggestions welcome.

Edited 1 time(s). Last edit at 12/22/2015 10:10AM by stamasd.

these folks:

[www.mosaicmanufacturing.com]

might be able to tell you something about filament transitions.

That is a brilliant idea! Print a raft in ABS and switch to POM - genius

Coming to the party late here... I commend you on your effort to pioneer something but the physical properties of acetal just make it a poor choice for printing. If you a want acetal parts I suggest you buy a lathe, mill, or pay someone with these tools to make what you need.

Don't fight physics.

I tried printing POM. I used Delrin tape on an aluminium heatbed. Printed at 220-230 for layer adhesion - much lower and it would delaminate. Heat bed was at 100C, but only because mine doesn't go any higher.

Using the Delrin tape, I was able to print objects about 20mm cubed. I printed a ~15mm major diameter herringbone gear and a 20mm calibration object. The latter had pronounced warp (several percent of the height at the worst corner.) The Delrin tape is one shot - it becomes your "zeroth layer." You can cut it off of course if you don't want it, and the adhesive easily comes off with acetone or MEK.

The adhesive on the Delrin tape seems to weaken at typical heated bed temperatures. Objects larger than what I mentioned above (e.g. a 4cm diameter gear I tried to print) pull up the tape from the aluminium bed. I have yet to try a cooler heatbed (paradoxical, but it might make the tape adhesive adhere better - I didn't realize it was having trouble with the temperature until I attempted to remove the residue from the build platform when it was cold the next morning.) I also haven't yet tried any measures to enhance the stickiness of the Delrin tape at elevated temperature.

I'm also experimenting with some more radical concepts which could have applicability beyond POM. I do think POM is worthy of further investigation as a FDM material - it's got quite low friction with itself (a third of nylon or a tenth of ABS with themselves) and is extremely suitable for moving parts. ABS is not so great tribologically for gears and PLA has service temperature issues for extruder parts, so printing things like POM and nylon is certainly useful for the "self replicating" aspect of the RepRap.

Incidentally, there is a website floating around out there that claims Delrin can be solvent welded with methyl ethyl ketone. This is not true. I even put a failed print in MEK, hoping to make some "POM juice" for experiments, but it was completely unchanged after spending all night in a flask with MEK. At this time I have not experimented with other solvents for joining or polishing POM.

Edited 1 time(s). Last edit at 12/28/2015 03:38AM by Bryce.