Tuesday 5 January 2010

Hot metal and serendipity

I couldn't get to work today because we had seven inches of snow during the night and a couple more today, so I had an extra day of RepRapping.

So my extruder is back working after re-fixing the thermistor with some RTV silicone. I get a degree or two more temperature swing with silicone compared to Cerastil, so not ideal, but it is workable. I think the plastic has such a high specific heat capacity and thermal resistance that it probably averages out the temperature swings anyway.

I switched to ABS to make a change from PLA as I am now able to use my 5kg spool of oval ABS that has always been two wide for my previous extruders. The bore of this one is 3.6mm, which is actually a bit on the big side for 3mm filament. I think about 3.3mm would be the best compromise.

My first experiment was to see if I could extrude directly onto my heated aluminium bed. My initial attempts failed to stick, even at 110°C, but I found that I could lay down a raft. I always cool the raft before applying the first layer of the object (I also drop the temperature of the first layer to 190°C), otherwise it welds too strongly to remove. When I cooled the raft it detached from the bed, presumably because it shrinks.

I reasoned if I could get the raft to stick then I should be able to get the object to stick. The difference is I do the first layer of the raft at 4mm/s and have the head lower than I would normally, so that the filament is squashed more. I tried making the first layer of the object at 4mm/s and a little lower than it should be. It almost worked so I upped the temperature to 120°C and tried again. This time I was able to make one of Zaggo's whistles.



When it came to making the pea it got too hot and started moving around.



Normally I would use a fan on ABS to get small items to hold their shape, but obviously blowing cold air onto a hot base is going to waste a lot of power. The fix I have in mind is to blow a very small jet of air at the same temperature as the base and aim it just below nozzle. Hopefully by keeping the jet small I can avoid the sort of power that hair dryers use. Adding the heated bed has increased the power consumption of my machine by about 50W, which has more than doubled it.

When I cooled the finished object and the bed to 40°C, by running the fan, the object simply lifted off. At 120°C the ABS is like a soft rubber or gel. It clings to the aluminium, but will peel off with very little force. When it cools it becomes completely detached.



The bottom of the object is smooth and shiny and perfectly flat. I can actually see part of one of the swirls that are on my bed if I catch it right in the light. That means the plastic takes the texture of the base, so you could pattern and texture it in the same way as injection moulds.

The next thing I tried was a Mendel vertex bracket as these are big enough to warp. It managed the outline, but when it started doing the outlines of the holes the filament failed to stick so I aborted that build.

The obvious way to get more grip is to use a sheet of acrylic as many people report that works well. I have a couple of problem with that though. Acrylic is a good insulator so the temperature control becomes more difficult. It tends to warp unless it is held down at the edges. I don't have any bolts long enough to mount it on my bed with the frame on top. I ordered some 2BA studding last year, but all the post from just before Christmas has gone missing.

I looked around for a piece of metal with some texture and found some aluminium with a satin finish painted with metal primer, from a very old experiment. It looked promising to start with: -



But it soon snagged and started ripping it up again: -



However, as you can see, I held the plate down with Kapton tape and by accident part of the object was extruded onto the tape. It stuck well to the Kapton but was peel-able. This looked extremely promising. Kapton on top of aluminium could be the perfect bed material for ABS. It looks like it will be reusable many times, as masking tape is for PLA.



The bracket stayed perfectly flat during the build. I cooled it with the fan to 40°C. It was quite difficult to remove. In the end I put a penknife under one edge and tapped it with a hammer. It came off cleanly and with a perfectly flat base with a glassy appearance.



The only blemishes are the gaps in the tape, what looks like an air bubble in the tape, and the dent from my penknife.



The base is a slightly golden colour and that extends up for the first few layers so I think the bed was a bit too hot. I had it at 120°C and the first layer at 4mm/s, so I will have to back track a bit and see if I can get away with a lower temperature and faster first layer, but this is looking very good. No warping, no raft, a cheap reusable bed material and a mirror finish.

29 comments:

  1. Hello :-)

    I am being very happy because after reading your last message I just used 3mm acrylic sheet on top of the aluminium heat build platform, and it sticks perfectly!!!

    I wrote a blog message, using pictures:

    http://casainho-emcrepstrap.blogspot.com/2010/01/perfect-build-platform.html

    ReplyDelete
  2. This has great promise. Kapton also comes in a liquid form. I wonder how smooth it could be made.
    I'm also wondering if there is some sort of spray on substance that would work.

    ReplyDelete
  3. i am learning allot lately, sounds like people are starting to get allot of bugs worked out in the past few months, i hope i can learn allot and put it to use in the one i am trying to build

    ReplyDelete
  4. Hi Nophead,

    That really is an amazing finish. It even looks like the whole part has a bit smoother finish, or it's just that I'm comparing with my own results. :-)

    -Geert

    ReplyDelete
  5. Your finishes just keep on improving sort the tape marking marks and your quality of print from the photos looks like that from an objet commercial printer.

    An inspiration to all.. Rep~Rappers

    ReplyDelete
  6. wow! perfect printing geometry. very good job! carried out to perfection...as always :)

    ReplyDelete
  7. The improvement in dimensional stability alone is a huge bonus.

    ReplyDelete
  8. Thanks for the positive comments guys.

    @Geert,
    The finish on the rest of the part is quite shiny. I think it varies with the exact type of ABS and the temperature it is extruded at. This reel of filament is more shiny than the white ABS I got from MakerBot for example. IIRC lower temperatures give a more shiny finish. I dropped this from 240C to 230C as there is no problem bonding when the object is at 120C and I get less fumes at 230C.

    A lot more experimenting to do. I expect the bed could be lowered to about 80C, and maybe the plastic to 220C. The object I made is extremely hard and well bonded. No way is it going to delaminate.

    @Casainho,
    Do you get a similar shiny finish on your objects with acrylic?

    ReplyDelete
  9. That looks really good, it almost looks like the pictures Curlrup took of his objet printed parts (http://dev.forums.reprap.org/read.php?4,27999).

    Is there anyway to make the remaining sides as smooth with the same finish, like possibly putting some Kapton tape on the sides and pressing an ironing iron against it for a short while or something.. (Hehe that sounds pretty stupid though, but still).

    ReplyDelete
  10. WOW...

    Hope the day off isn't too painful with work.

    ReplyDelete
  11. I agree with the other posters -- the builds look very nice, even at the blown up full size JPGs :) How far did you plan for testing warping limits -- on larger objects?

    Filiph: One thought I had for getting better exterior surface resolution (I.E, it's smoother') I posted a few years back.

    Have a "spoon", a piece of steel wire (1mm diameter) with a half sphere sanded onto the business end, near the print head (solinoid activated), or on a different tool. The extruder lays a layer of filament (1mm!). The solinoid activates, and the system spoons stuff around to remelt and reshaping it to the resolution of the positioning system.

    There are a few issues to this system: inverted surfaces don't get details without a bent spoon and 5th rotational axis, and there may need to be a generation or two to get extrusion predictability up; There are also great advantages -- extrusion speed (1mm is 8x faster than .5 mm), and the details possible on the sides and top of objects on a Mendel can now approach .1mm, the thickness of writing paper or human hair diameter.

    ReplyDelete
  12. Where did you get the Kapton tape nophead?

    Laszlo

    ReplyDelete
  13. @Filiph

    I have melted Plastics, PP PE ? directly on a warming plate from a disasembled coffee maschine which was covered by a teflon like coating.
    The part could easyly removed after cool-down.
    The surface looks very good and shiny.
    Maybe you could lay down and/or press a flat reprapped part on this hot plate, switch off the power and let it cool down or quick cool with a hair dryer / fan.
    Maybe you could use a teflon coated pan on your kitchen oven too.

    Erhard

    ReplyDelete
  14. @all , @Filiph

    I have found an awesome finish to all your flatsided reprap parts:
    You need a Teflon pan made of aluminium and your kitchen oven plate with thermostatic control.
    Warning : Overheating Teflon/PTFE over 200 degree Celsius can release highly toxid fumes !!! Look inside the manual of your pan.
    Starting to heat up the oven plate with the lowest temp and rise it very slowly to the melting point of your plastic type. After each increase of the temp wait for 2 or more switching cycles of the thermostate. If you find the melting point with a crap piece of your plastic read the value of the knob-scale for future jobs.
    You will find different settings depend on used plate, pan size/type and of course your plastic type. I also prefer to do the fine cleaning, boring, tapering jobs after this procedure.
    Now press the flat side of your reprap part on the bottom of your pan for some seconds until you see some melted plastic come out the gap between and then instantly cool down the bottom of the pan with cold water and than the inside too.
    Depending on result start again or take the next with the dryed pan heated up for a couple of minutes. Repairing your mismatched parts could work by add some plastic with an soldier iron to part itself or melting it directly in the pan and then press your part on the molten plastic.
    The finish depends on the finish of your pan.
    Tested with bottlecaps (PE) and PP.

    Erhard

    ReplyDelete
  15. Hi, You can check out our take to heated print bed at our site prusadjs.cz/2010/01/heated-reprap-print-bed-mk2/ We are working on this thing nearly for two months with I hope pretty impressive results. We are using RepMan.

    ReplyDelete
  16. What about printing ABS onto masking tape? We've found it to stick better to tape than Acrylic.

    ReplyDelete
  17. Is our python source open, and if so, where can I find it?

    ReplyDelete
  18. What I meant to ask is if "your' python source is available for download? I am doing something similar to the "cooking with hydraraptor" and would like to see your complete source if possible.

    ReplyDelete
  19. @BeagleFury,
    Nice idea with the sculpting "spoon" but I think the software would be a nightmare as you are only moving plastic around so would have to deposit the right amount to start with.

    @Filiph,
    Erik be Bruijn showed me a Yoda head that was made on a commercial machine and then exposed to a solvent, in gaseous form IIRC, and that was perfectly smooth on all surfaces.

    @prusajr,
    It looks good but, with Kapton I can use a higher temperature to get the ABS to have a smooth glossy surface. I think Acrylic warps too much above 100C and would stick too well to the ABS.

    @Geeky Gnu,
    I don't know if the glue on masking tape will handle these temperatures and the surface is not smooth. I have used it for PLA, which will only need a bed temperature of about 40C, so it might use it for that, but I will try PLA on warm Kapton first because it is smooth.

    @twgray,
    No I haven't published the rest of my Python code because it is only relevant to HydraRaptor and a very old version of Skeinforge. I also change it most days as I am continually experimenting, so it is not a very coherent piece of code!

    ReplyDelete
  20. Thanks for your reply, do you have any more information about this solvent? Sounds dangerous to do this treatment near a reprap since a lot of the parts would be printed in the same material as the solvent would be used to smooth, and it didn't deform the head at all?

    ReplyDelete
  21. I don't really know any more, perhaps Erik does. It was a commercial sample so I don't think he witnessed it being made.

    I think it would be done in a sealed container and yes it is probably dangerous to do at home. It would be quite easy to vaporise acetone but it would be highly explosive.

    Using liquid solvent would be a safer alternative.

    It was completely smooth and glossy, like an injection moulding but with no split lines, etc.

    I don't know if it deformed as Yoda's head is not a geometric shape that you could notice deformation.

    ReplyDelete
  22. Great discovery Chris! I hope to follow in your footsteps soon and build a platform too. I already cut and bended an aluminium sheet (1mm is easy to cut with a Stanley knife or any sharp blade, for that matter). Being so thin, it will require even heating to start with.

    @ Filiph: I justed posted about the sample part:
    http://blog.erikdebruijn.nl/archives/125-Smooth-I-am,-yes....html

    ReplyDelete
  23. NopHead, I got some shiny finish but not as perfect as yours.

    Today I will get a piece alluminium and I hope to start building a new heated build platform. After I will try to follow your steps and validate your work :-)

    ReplyDelete
  24. Do you have a supplier link for 'RTV silicone' you use to stick thermistors into heaters, I thought when making my heater it would be a bit neater than slapping some kapton on it.

    ReplyDelete
  25. Hi Giles,
    I used oven door glue that I got on eBay: http://cgi.ebay.co.uk/OVEN-GLASS-DOOR-GLUE-High-Temperature-Silicone-ADHESIVE_W0QQitemZ400042548358QQcmdZViewItemQQptZUK_Home_Garden_Kitchen_Ovens_Hobs_Cookers?hash=item5d2464dc86

    ReplyDelete
  26. Ah, great! I actually came across that stuff but wasn't sure it it was "The right stuff"

    Something that has just crossed my mind is what about Exhaust repair paste as a possible alternative, something like Holts gun gum for example. I don't know what its made of but I am guessing its Sodium silicate based. Furthermore I can't find any specs on max temp.

    ReplyDelete
  27. My Dad found some Stovax high temperature rope adhesive, I couldn't find a spec on max temp for it but other brands go up to 850 oC So wondering if any of these might be a suitable cheap alternative: http://www.google.co.uk/products?q=high+rope+adhesive&hl=en&aq=f

    ReplyDelete
  28. That looks promising and not expensive. Duralco 133 high tempeature thermally conductive epoxy looks like the ideal stuff to use but I can only find one source of it in the US with no prices.

    ReplyDelete
  29. I had been doing some tests by putting a blob of stovax on alu and placing it in the flame of my cooker hob. I found that it started to burn when put in direct contact of the flame, but couldn't get it to melt when heated in-directly (i.e with the block of alu between the flame and the glue) No Idea how high a temperature it was going up to as I don't have a thermometer. Now I have just got a reply from stovax customer services:



    Dear Giles Bathgate.

    Thank you for your enquiry.

    The Stovax Thermic seal adhesive withstands heat up to 1250 degrees Celsius.

    Regards,

    Customer Services

    Stovax Ltd

    ReplyDelete