How to get extrusions to stick well?

I've been asking random acquaintances the following question:

"Can you think of anything that's about 1' square, 1/4" thick, porous, rigid and reasonably heat-resistant?"

I've gotten a lot of blank looks but I've also gotten "unglazed ceramic tile" a few times now. Might be worth a try. I wonder what would happen if you warmed it in the oven first.

... for ABS you could solve the adhesion-problem with painting a porous surface (e.g. wood or floor-covering laminate-pieces) with ABS-dust solved in acetone.

It's known as ABS-glue and solidify to a hard ABS-surface with nearly 70% stability of the original ABS-material.

Should work for every other plastic too when you find the right solvent ...

Viktor

Other suggestions I've gotten: pumice, foam rubber mat

I played around with some sugar tonight to test it as a possible support material and it seems to have some potential. What can I say I have a sweet tooth.

I set the oven for 370f and made an approximately 1/8 in. deep, 6 in. diameter disc of granulated sugar on top of some parchment paper in a pan. I didnt time it but if I had to guess I would say 25 minutes to completely melt. It caramelized a bit but did not burn. After letting it cool I had a pretty nice disc of hard smooth sugar. After it cooled it was starting to get sticky from the very mild humidity (coastal california).

Lacking an extruder I lite some HDPE milk carton pieces and dripped them as best as I could onto the sugar disc. They seemed to stick quite well. They were definitly above 365f (what I found when I googled sugar's melting point) as the surface of the disc melted and then resolidified. If I moved the plastic too soon it just pulled away some sticky melted sugar, if I waited a bit it was very difficult to separate the two. I also dripped a piece of an old mouse housing which I am guessing is ABS onto the disc. It also adhered quite well, in a strength test it started to break the disk before it separated.

The HDPE is the hard to see clear droplets and the easier to see grey burnt looking plastic in the bottom middle of the disc. The ABS is sitting on parchment paper above and to the left a bit of the disc, it was attached below and to the right a bit where you can see a circular fracture in the disc.

Thoughts? Anyone with an extruder, and oven, and a cup of sugar have the inclination to try this?

Quote

I played around with some sugar tonight to test it as a possible support material and it seems to have some potential

That would cerainly show that sugar has potential as a bed material.

Using caramelised sugar as a support material is slightly inconvenient, as it implies a 3-toolhead system... part material, sugar (presumably in the form of water icing, or similar paste) dispenser and a sintering head, presumably something tried and tested like candyfab's hot air heater. So long as the sintering occurs before plastic parts start melting, of course.

How easy is it to remove caramel from something? Without rushing to the kitchen right now and burning things, I recall that it is reasonably resilient, and so is going to be awkward to remove from fine features. Possibly leaving the part in a tub of water for a while would do, but you wouldn't want to do that with hygoscopic plastics like PLA, I assume?

It is worth noting that candyfab is intended to sinter sugar particles together rather than melt them outright, so it will have different mechanical properties than pure caramel. This will possibly make it easier to remove mecahnically or by dissolving it.

I believe a granule extruder for sugar is possible. bag of sugar goes in the hopper, molten sugar comes out. similar to the previous experiments with the PCL experiments with a granule extruder. Or a similar approach with premade sticks of melted sugar similar to the concept here: [www.instructables.com]

Nice work!

Removal would be a trick though... the soak-it-in-water approach is good because it doesn't require much human effort, but on the other hand I suspect it would take quite a while to dissolve away

Perhaps there is some sort of acid or catalyst that would cause the sugar to dissolve rapidly, without affecting the plastic?

For sugar : boiling water - heat + water solvent + ajitation = quick clean model

Alternative sugar removal ideas.

leave it on an ant hill.

Give it to some small children.

Seriously though:
I believe a dishwasher might work
or any other application of hot (boiling?) water or steam. (obviously not a solution for PCL but I doubt it would stick well anyway.)

I have seen vinegar listed as a potential solvent.

It is very brittle, a certain amount could be shattered to loosen or remove.

Other possible base materials could be particle board or asphalt shingles.

I got great results out of a thin piece of Plexiglass last night, right before my extruder broke down.

There was a bit of warping due to the thin sheet being unsupported, but I'm off to buy some thicker plexi and bolt it to the bed. Seems very promising!

Almost finished a new motor mount; I think enough of it was built to be useable thouh, a first for me.

I have just moved to a kitchen chopping board. I do not know the material but it is the common 'plastic' variety. Using HDPE at 240C, I put down a single layer of the Skeinforge raft under the product, setting the nozzle close to the surface to get good adhesion, and it works much better than any other base I have tried. The product can be removed using a chisel leaving the board reusable.

That is because the board is likely to be HDPE as well.

I put this on the Wiki, but I should probably mention it here as well - 1/4" acrylic sheet (aka Plexiglass) has been working really well for me with ABS. I've printed over 80 good parts on the same sheet now, although I had a heater failure yesterday that put a good dent in my acrylic bed; it now has a nozzle sized crater right in the middle. I'll have to blog the sordid details...

Wade

degroof Wrote:
-------------------------------------------------------

> "Can you think of anything that's about 1' square,
> 1/4" thick, porous, rigid and reasonably
> heat-resistant?"

Gypsum-board (AKA sheet-rock) might work, especially if the paper surface was roughed up or peeled off. Cement board, used as an underlayer below tile in shower stalls (at least in the US), might work even better -- no paper layer to remove, and a bit more porous. Both come in large sheets, and scrap from construction/remodelling jobs may be available. (I have gypsum board, if anybody in the Boston area with a working extruder wants to give it a try. I'll keep my eyes out for cement board as well.)


-- Larry

A headed base plate worked well for us. First we used some nicrome wire between two epoxy plates that are bare epoxy on the inner side (so it does not short circuit the nichrome wire) and a thin layer of copper on the other side (we used this for etching pcbs normally). We print onto the copper side. It does not stick very well but at temperatures of about 100C or so it sticks well enough for a raft. The big advantage is that the stress is removed and the parts don't bend upwards after printing - they are completely flat. Disadvantage ist that you will not be able to get the raft off so we only print 2 raft layers and don't care about the extra layers. More info and pictures can be found on our blog

[reprap.soup.io]

In the next version we just etched a long winded trace of copper as a heating element. Sadly the resulting resistance was much higher than the calculated value so it needs way more than 12V to be heated, so we'll have to redesign this one.

One word of caution: While the heating helps reduce the warping stress don't overdo it. If the base plate is too hot the material sticks very well but also shrinks and gets crunchy. About 100C seems to be the best value although I don't fully trust our readings.

All I wrote here applies to ABS we haven't tested it with other materials yet.

---

greetings
wizard23
[reprap.soup.io]

Now that is a cool idea! Please post the details on your new PCB - I'll be etching one myself very shortly!

Wade

Wade Wrote:
-------------------------------------------------------
> Now that is a cool idea! Please post the details
> on your new PCB - I'll be etching one myself very
> shortly!

If you are interested in the pcb: I made some python scripts that generate the serpentines as a kicad .brd file. The rest (adding some more traces and vias for mounting i did in kicad). the python scripts and the board I actually etched (generate_meander2.py and reprap_heaterv03.brd) are in our subversion user directory:
[reprap.svn.sourceforge.net]

I calculated the pcb with this PCB track with calculator to be 8 ohm at 100C:
[circuitcalculator.com]

but when I etched it it had more than 18Ohm so something must have gone wrong or the PCB trace calculator is wrong. A friend who is very much into analog electronics built a small cuircuit so we can now drive it with 30V from an arduino but because the standard reprap PWM board uses 12V I want to redo it with thicker traces

A the moment I'm sick and so I can't go to the lab and test our new designs but I will post updates as soon as I etch a new version.

Very nice. I'd considered using a heated bed but didn't think of using a PCB. Great idea. Please add this to the builders' wiki as soon as you can.

This is something I have been meaning to try for ages. I will my PCB.

To fix the resistance you could just attach a wire to the centre and link the other two connections to get 9R.


Edit:
On second thoughts that would of course give 4.5R, doh.

Edited 1 time(s). Last edit at 11/19/2008 04:57PM by nophead.

---

[www.hydraraptor.blogspot.com]

nophead Wrote:
-------------------------------------------------------
> To fix the resistance you could just attach a wire
> to the centre and link the other two connections
> to get 9R.
>
>
> Edit:
> On second thoughts that would of course give 4.5R,
> doh.

Yes it would be 4.5ohm. Still this could at least be driven with the PWM board (althogh the transistor would need a fat heatsink to drive that and survive). so this would actually work...too bad that i glued it together with another epoxy plate in an effort to make it more flat (which did not really help) so i cant access the centre anymore.

I'll just increase the trace width in the existing board and etch it next time i'm at the lab. It sucks to be sick and sit at home...but at least i have time to read and post to the reprap forum

---

greetings
wizard23
[reprap.soup.io]

With just a power MOSFET, an inductor and a Schottky diode you can convert the PWM to a current that will work with virtually any resistance.

You effectively create a small switch mode circuit, controlled by firmware, which you configure as either buck or boost depending on whether the resistance is too high or too low. If you are interested I can sketch the circuits.

I hope you get well soon.

---

[www.hydraraptor.blogspot.com]

Well for now I have a circuit that I can controll the heated bead with. We want to keep it as much "standard-reprap-parts" as possible. So for this it's not needed.

For other things or for people who want to try out the heated base plate as it is it might be usefull. Might also be usefull for our planned plastic recycler - which would basically be a modded old school meat grinder with lots of nicrome around it.

... hmmm ... you let me think ...

And then i remembered some old heating-elements i scavanged twenty years or so ago:


This are heaters with 115V AC and 150Watts each.

Connected sequential i can plug them direct to the line.

Mounted under an aluminium-plate and with an old but working temperature-controler and a PT100-resistor as temp-sensor this would be a perfect heating-bed.

I think with some heat-insulation beneath i should be able to set and control the average plate-temp between 30

here is what i am using. this is a 1/4 inch aluminum plate, about 8 inch by 4 inch. there are four power resistors jb-welded to the plate underneath. the reprap platform is wooden particle board - like the stuff inside of ikea furniture. the resistors just sit on the particle board. it has not caught fire yet but that is always a risk.

each resistor is 32 ohms. they are in parallel, which gives 8 ohm total resistance, chosen because extruder heater is about 8 ohms also. each resistor has a power rating of 25 watt. there is a 100k thermistor glued to center bottom of plate.

a stand-alone arduino running some simple code runs an extra pwm-driver-board to run the heater. this runs on a big separate 24V supply. 24V is needed to get the plate hot enough. a temperature of around 150C to 160C seems to get HDPE to stick pretty well. takes about ten minutes to warm up at full-blast 24V.

a word of caution to any of you folks who might want to try this at home: the plate gets HOT!!! you obviously need to be careful around it and not leave it running by itself.

so now the big question: does it help the plastic stick? well, yes. but the part also melts and oozes out a bit at the bottom as the build proceeds. to get the part off, turn off the plate and let it cool. with hdpe the part separates from the aluminum easily and is left with a nice flat surface. warping is also greatly reduced.

now, whether using a plate like this is better than using a raft like other folks, i'm not sure. both have good points and bad points - and i haven't tried using a raft. but i think the raft may be the better way to go, considering the hassle of the heated plate and the relative minor advantages...

Nice work,
Have you tried having it hot for the first layer to stick and then dropping the temperature a little below the melting point of HDPE (130 IIRC) while building the rest of the object?

---

[www.hydraraptor.blogspot.com]

Huh... a heated base plate reduces warping. Who'd have thought it? I wonder if that would make HDPE work?

I'm wondering if I couldn't rough up the back of an aluminum plate like you all are using above, apply a thin layer of Cotronics or Cerastil ceramic like Nop and I use to secure nichrome onto our extruder barrels, then lay down a length of nichrome and bury it in another layer of ceramic?

Mind, I've got a pile of power resistors that I'm not using for anything and a BUNCH of epoxy. :-D

I LOVE this project!

Edited 1 time(s). Last edit at 11/22/2008 11:43PM by Forrest Higgs.

Hi Wizard23,

Is your heated plate hotter in the center than the edges and if so has that been a problem?

If it is a problem, you could change the script so that a double loops starts around the outside with the edges close together, then as it gets to center the edges would be slightly wider apart so that there will be more heating at the edges of the plate to compensate for the extra cooling at the edges. Then the temperature across the plate could be constant.

Enrique Wrote:
-------------------------------------------------------
> Hi Wizard23,
>
> Is your heated plate hotter in the center than the
> edges and if so has that been a problem?

i think so. Although it's hard to measure this accurately. we baught one of the laser temperature measuring devices but it doesn't give good readings on shiny copper.

> If it is a problem, you could change the script so
> that a double loops starts around the outside
> with the edges close together, then as it gets to
> center the edges would be slightly wider apart so
> that there will be more heating at the edges of
> the plate to compensate for the extra cooling at
> the edges. Then the temperature across the plate
> could be constant.

A friend of mine als suggested this. And so i wanted to do it like that in the script. but as i did not have any real values on which to base the density distribution along the radius (and because the serpentine script was already finished I went for this design. But its on my todo list...still it will not be easy to figure out how to distribute the lines without one of these super expensive thermal cameras I don''t think the temperature gradient is a real problem though (but for bigger areas it might well be)

---

greetings
wizard23
[reprap.soup.io]

wizard23 Wrote:
-------------------------------------------------------
>
> i think so. Although it's hard to measure this
> accurately. we baught one of the laser temperature
> measuring devices but it doesn't give good
> readings on shiny copper.
>
You can correct those readings by plugging in the temperatures you're getting into the Stefan-Boltzmann equation and solve for energy assuming a 0.9 emmisivity, which is what most IR thermometers assume (check the product literature to be sure). Then use the radiant energy figure and plug it back into the same equation along with the book emissivity for polished copper.

In both the 1.6 um region you get an emissivity of about 0.03.

[www.raytek.com]

while for the 8-14 um band it runs about 0.02.

[www.infrared-thermography.com]

Mind, if you've let it get oxidized, which it will pretty quickly if it's spent siginificant amounts of time at temperatures > 100C, that emissivity is going to climb up to about 0.7.

Anyhow, plug those figures back into Stefan-Boltzmann and you can get pretty good corrected temperatures.

One thing that you need to keep in mind if you are using a non-contact IR thermometer is to get it quite close (maybe a cm) to the target, take the measurement and then pull it away. The cheaper IR thermometers typically have a fairly wide view factor (you can check in your product literature), so it's easy to either be measuring too big an area and, as a result, losing resolution, or measuring off the edge of your plate which gets you an average of the plate temperature and the background, too. They also tend to use cheaper IR sensors, too, which measure only the part of the blackbody radiation in the 1.6 um region and then basically guess about what's going on at longer wavelengths (8-14 um). This last, however, is really only a problem if you are measuring temperatures in the vicinity of room temperature. At temperatures ~100-150C the 1.6 um region sensors should work just fine.

I've done a LOT of this kind of stuff.

Edited 2 time(s). Last edit at 11/23/2008 11:39AM by Forrest Higgs.