Multiple Material 3D Printing

[youtube=http://www.youtube.com/watch?v=D4Yq3glEyec]

It’s no surprise that we’re wild about 3D printing, especially [Devlin]. Now we’re absolutely out of our minds for this multi-material polyjet machine that is featured in the video above. Before we go any further it’s worth mentioning that this post is not advertising, we just think this machine is unbelievable.

It is capable of printing 600 dpi in 3D using multiple materials at the same. Two types of rigid material, one like ABS and the other like polypropylene, as well as seven levels of a soft material all exist on the same print head. They can be deposited along with a support material at the same time. In the video you can see enclosures that come out of the printer with rubber-like padding already mounted in the hard plastic shell. They even show a bicycle chain that is fully assembled after printing. Cost for these machines? We don’t want to know, it’s just fun to dream about having unrestricted access to one.

[Thanks Karl via PDDnet]

60 thoughts on “Multiple Material 3D Printing

  1. Hold on I’m still trying to find my jaw… Holy Sh!t that is amazing! I like the rubberized material you can print into objects. That just makes it even more epic.

  2. To be honest I am starting to dis-like hackaday more and more every day. It is like they ran out of information to post on hacks so now they post information videos. The videos are nice, but its not a hack. It is like eBay posting news, they just don’t do that. I hope hackaday can pull it together and post hacks like they use too, and good hacks at that, I loved watching the old videos. Now its like oh.. yeah.. a informational video, no not another one (“yawn”).

  3. It’s news for me alright! Never thought make such good quality parts. I like probably many others here have only seen crappy makerbot parts with terribly thick layers. Just can’t stop thinking of making rubberized custom electronics enclosures!

  4. Sparkfun so needs to get one of those with some of that 10 million they make and set up a batch3Dprint.

    That machine is so cool. I didn’t realise that 3D printing was possible in such a high resolution…

    Mowcius

  5. I run an invision hr and an old solidscape patternmaster II. Both of these printers work on inkjet tech just like the eden…youll never get reasonable surface quality with rep-poop’ers. Dont get me wrong their cute…but not terribly practical.

    I am suprised that the DIY crowd hasnt put more effort into something like the envisiontec perfactory system…zaxis stage, UV cure acrylic, and DLP projection.

      1. The DIY community HAS…now…2 years later… yes. Check the dates of this thread….note the final comments date relative to jr’s first posting on the subject in the yahoo group.

  6. @ techartisan

    I think it’s because the materials cost too much. AFAIK all two part polymers components degrade over time, which isn’t cool for the home/hobbiest market.

    @ all the “this ain’t a hack” folks. Freaking build something and submit it to hackaday. Otherwise shut the fuck up.

  7. Could you do a post dedicated to 3D printing services? The only ones I’ve heard of to date are shapeways & growit3d, but there have to be more, in different price ranges (and buying parts locally should reduce shipping costs)

  8. @barry99705 @techartisan but the materials sure can’t be THAT costly? (somebody know a reseller?)
    the technology looks awesome and quite cheap, one motor for the Z-axis, a modified video projector, adding a new layer of material could be done manually to make it even simpler… looks like a “doable” project, provided one can get the material

  9. printapart uses the invisionHR great for high detail small object printing…jewelry etc. I used them for a few test parts before we bought the printer. Good service and reasonable price…instant online quotes.

    Shipping costs arent a huge portion of the expense.

    Most commercial printers are tied to proprietary materials…I burn through $1500 in material every month or two with light personal use.

    These commercial services also have to concern themselves with build time and volumes as thats where their margins of profit are squeezed from…

  10. RepRap? Not Practical? Surely you must be joking.
    What the RepRap lacks in surface finish it makes up for by being cheap. A reprap (or makerbot) costs many orders of magnitude less than any commercial machine to set up, and the objects it produces cost just 2.2 cents per cubic centimeter (less if plastic is bought in bulk). Try to find a commercial machine that can even come close to that.

    [5lbs of abs costs $50 in the makerbot store. 5lbs=2.26 kg, so 45.2 g costs $1. From what I’ve been able to find, the density of abs seems to be 1.02 g/cc, so (45.2g/$1)/(1.02/1cc)=44.3cc/$ =2.2 cents per cc.]

  11. My personal reprap derivative cost me under $300 US to construct from scraps, and gets between 0.10 and 0.35 mm thick layers. That’s nothing even comparable in resolution to this machine, but not only is mine open source and self-upgrading, it freaking costs less than the print material that this monster uses. I printed mother’s day gifts yesterday.

    Maybe I don’t make the kind of money you reprap-critical people do, because I barely manage to support being a “maker” and being a student.

    Hackaday, go back to posting interesting hacks.
    This is impressive, but also
    A) Expensive.
    B) Proprietary.
    C) Old news.

  12. @pascal…. star technology $150 a liter for one component UV acrylic… UVA4101-9

    @nick Not joking at all.
    If you want to print low resolution large parts and dont mind stairsteps and long build times then parts from a reprap or a professional extrusion machine, like the dimension uprint a $15k reprap then perhaps they might be good enough for you.

    Id sculpt a blob of polymer clay by hand if I had such loose tolerances…Id do just as much work cleaning up and tuning a reprap model postprocess.

    If you want any degree of mechanical complexity or precision interface between surfaces…This technology is not practical. I guess ultimately it depends on what your after.

    Inkjet 3d printers might be a bit more difficult to manage but I seriously believe that the DLP cured system is a practical project to attempt…if I didnt have prototypers already Id think about it myself.

  13. Hmm…

    That basicly uses an inkjet printer and some UV-setting polymer.

    I wonder if I could find a UV setting polymer that would pass through the standard thermal heads of my inkjet printer at home. If I could, then this could be truely home-made.

    I’m thinking of some Epoxy Resin and Acetone mixture…

    Need to think of some similar way to do the support material. What about a wax and solvent mixture?

  14. My invision hr uses a UV acrylic build and a paraffin support both “shot hot”..prints layer…then cures the layer..then a rotating heated drum skims the surface off reducing layer thickness…then repeat.
    My patternmaster, which was the predecessor of the solidscape t66, uses a dissolvable wax support and a jewelers thermoplastic wax build material…again both preheated and served hot via inkjet.

    A homebuilt system could easily use more then just the two materials…as the connexx500 does.

    the perfactory system doesnt use a support material it uses a stage that it lowers to near contact with a wet (with uvcure acrylic)glass stage then raises as the “layers” cure. So a relatively precise Z axis and modifying a dlp projector is really all there is to it. The lack of support material creates some design limitations that discouraged me from pursuing one when I was purchasing printers…but the simplicity, precision and continuous layering are definitely pluses to be considered.

  15. @spyder_21
    did you even read the rest of the page? I mean really! Theres 1 other informational post on a process which many ‘hackers’ will find useful (toner transfer), there is another post asking for our (the readers’) input, which you seem only to eager to give.
    The rest IS hacks.

    Plus you said yourself the post was interesting… stop your bellyaching!!

  16. “I think it’s because the materials cost too much. AFAIK all two part polymers components degrade over time, which isn’t cool for the home/hobbiest market.”

    Wrong. These are one part materials. Technically, the inks have acrylic/epoxy/urethane/polyester chemistry (monomers and oligomets) plus multiple photoinitiators, stabilizers, surfactants, etc. Nothing TOO fancy but their inks contain quite a number of components, components you can’t just buy one at a time in small amounts. We are talking about 55 gallon drum feedstocks.

    I own multiple Objet printers and the materials are crazy expensive. $500 – $800 for a single cartridge, which is only good for 2 – 3 large models and by large, I mean under a cubic foot. More like 8x8x8.

    The materials don’t degrade over time once printed although they do have a shelf life due to the reactive nature of the materials. However, there are certain things you can do to improve that.

    Also, the printheads are expensive. The construction is a bit robust as they are industrial printheads that are designed to run at 65 – 75 degrees C. This machine has 8 printheads (as do most of their machines except the Alaris which has 1 and Eden 250 which has 4) and each printhead costs $1100 or so and is only good for a few thousand hours.

    However, the quality is fantastic. EnvisionTEC has better quality potential but they can’t print support material – so they have to build scaffolds out of the model material to support the material. Since the scaffolds are made out of the same material, you have to manually remove (with a knife) those and it leaves marks where you do that. On large projects, that can take a LONG TIME to clean the model up enough and on some geometries, it isn’t even possible.

    All that said, I wish the DIY market were to go after polyjet technology over makerbots. By their very design, makerbots are limited. You can’t make the “glue gun” printhead much smaller and you can’t fill in “pixels” with a round nozzle. Delamination is an issue and ultimately, quality is not likely to get anywhere close to polyjet anytime soon, if ever. But they DO make what can be called “parts”. Kind of. No offense intended.

    If you were able to marry room temperature UV inks (that Objet doesn’t have – all of their inks are too “thick” to be able to be printed at room temperature) with off the shelf printheads (or a gang of printheads) then you would be on to something huge. Because then you could use LED UV lights, cheap printheads and presto – ultra cheap, ultra high quality printer with very cheap consumable prices. Trust me, those inks have a HUGE markup. If you were able to buy them at cost, I imagine, actually I take that back, I know that pound for pound, they would be on par with the ABS plastic used in most makerbots.

    “What the RepRap lacks in surface finish it makes up for by being cheap. A reprap (or makerbot) costs many orders of magnitude less than any commercial machine to set up, and the objects it produces cost just 2.2 cents per cubic centimeter (less if plastic is bought in bulk). Try to find a commercial machine that can even come close to that.”

    I can’t go too far into the details, but if you know what you are doing, you can create inks for polyjet technology that rival those prices.

    “I wonder if I could find a UV setting polymer that would pass through the standard thermal heads of my inkjet printer at home. If I could, then this could be truely home-made.”

    You will most likely not be able to. Traditional piezo inkjet heads jet at room temperature. Most polyjet ink would never flow at all at room temperature, you would have to thin it out with monomers, which lower the quality of the part – usually so much so that the part you printed would have no mechanical properties at all. And off the shelf printheads jet far too low of a quantity per jet. You need ideally around 60 picoliters per drop. 30 would work. Most inkjet heads might manage 1. You would also need to print more than you need and then scrape a small amount off because you can’t deposit EXACTLY what you need.

    You would also need 1 – 2 250 watt UV bulbs and requisite power supplies along with some algorithms to compensate for blocked jets, an accurate to 16 micron Z axis, etc. Not that it can’t be done but the ink and the ink delivery system would prove to be a rather large challenge.

    I spent 2 years and six figures developing a custom formulated room temperature polyjet ink that has multiple patents pending. And I would very much like to see DIY polyjet printing not only happen but become the next big thing. Honestly, there is no compelling reason why this can’t happen. All it takes is inks that can be run through inexpensive inkjet heads. And I know for a fact that something like that is possible in a machine that could be assembled for under $1000 that costs the same as a makerbot to actually run.

  17. @PolyJetter: why does EnvisionTEC need support structures? I thought it was like SLS, they just replaced the single laser beam by a DLP UV projector — and that SLS didn’t need support structures, because you always had a bed of non-fused material that holds the structure.

  18. Because you have to hold the material you are curing and prevent it from moving so when it does the mechanical recoating cycle at the end of every exposure (to build the resin up), it doesn’t move the model you are trying to make. If it moved even a micron with each pass, it would very quickly ruin the model. So you have to anchor it. Also any overhangs past a certain amount need to also be anchored. At MINIMUM, you need to anchor the model’s base (whatever you you designate as the base because you tell it what face the base is) to the build tray. Depending on the overhang angle, you may be able to get away with just that one support “material” geometry using EnvisionTEC’s technology.

    EnvisionTEC is really more akin to SLA then SLS. They just use MEMS DLP and a projector bulb instead of a UV laser. SLS in plastic powder does not generally need support material but in metal it does (because the density of sintered metal is far heavier than the density of loose metal powder).

  19. Blocked Jets: Not a problem – we just interleave the layers so a single blocked jet affects different places in the model equally.

    Accounting for volume differences: Attach a tiny piece of sandpaper after the print head. As it prints, any excess will be “worn” down. The amount of excess here should be tiny, provided the print head is pretty consistent in volume.

    Too small droplets: Well this just slows down printing. Some inkjet designs seem to be able to do drop sizes from 2 to 50 pL. A less viscous material with a lower boiling point will also increase volume.

    Arrays of print heads: Sure – I guess you could have a 4×4 array of print heads to print at 16x speed – aligning them would be a pain.

    Material Cost: non-issue, most of the ingredients are produced by the barrel.

    Accurate XYZ positioning: Well X and Y are done already in a home inkjet. Z only has to move one direction, slowly, so not too hard. One could begin with only X and Z, and print tiny models no bigger than the length of the print head.

    PolyJetter: Feel free to contact me with the contact form on my website.

  20. “to thin it out with monomers,”
    Hmm – what about thinning it with a solvent which evaporates. Considering the tiny drop size of an inkjet, the solvent would probably evaporate within a few hundreds of miliseconds, so by the time the UV light cures the resin, it ends up being just as strong.

  21. Ketone / alcohol solvents tend to be flammable, noxious and tend to warp material in random ways as it evaporates. You can use water (which is a great solvent) but you have some of the same problems – and water has to be heated up a lot to evaporate. But then you add in the miscibility problem. UV resins tend to not be water soluble. Only a few PEG based hydrogels, modified BPA epoxies and the like are actually what one could consider water soluble. And very few are 100% soluble.

    Also, keep in mind that the head is printing a new layer every few seconds, curing it instantly with UV light. Solvents would not have time to flash off fast enough, even with the small droplet size. I would have to calculate how much heat would have to go into a droplet in order to flash it off fast enough but for various reasons I believe that to be highly impractical to deposit 20% “solids” or even 50% “solids” and then reduce it down. It is FAR better to simply print 100% solid, ultra low viscosity (almost as thin as water) material.

    The average polyjet resin is on the order of 400 – 800 centipoise at room temperature. So roughly 400 – 800 times thicker than water. Not quite molasses but starting to get up there. Too thick to jet with most printheads. Even the ones on their machines until you heat it up to 65 – 75 degrees C. Nearly 170 degrees F.

    To add insult to injury, at that temperature, you also run the risk of thermally curing UV if you keep it at that temperature long enough, no matter how many inhibitors you put in it. With the right material combination, it isn’t a big issue but just another thing to think about.

  22. Sounds like you know lots more chemistry than me – my approach was going to be just mix things at random till it worked…

    What about something like superglue (cyanoacrylate) – it looks like you can get it down to 5 centipoise, so it should pass through an ink cartridge.

    Obviously you have a big problem with it’s potential to block jets, but I guess as long as you find a way to keep it from setting while in the head it should do the trick. It has a 2nd advantage of not needing a UV light..

  23. So how do you cure cyanoacrylate in seconds without it curing in the heads? It cures with moisture. Which means it would cure at the printhead. Assuming you could prevent that, how would you cure it instantly (or near instantly) once deposited?

    If you really boil it down, UV inks are basically non moisture curing (UV) superglue (cyanoacrylate) that cures with cationic (less common) or free radical (most common) chemistry. Hit them with the right kind of UV light (which spectra you need depends on what photoinitiators you use) and they harden just like superglue.

    I love the innovative thinking but I worry that you are going to very quickly wind up with very clogged printheads or a vacuum chamber to print in, which has its own issues (mostly cost). Then again, many UV inks do not cure well in the presence of oxygen but that can be overcome with various techniques. I am not sure how you would overcome the moisture curing problem in normal atmosphere. Even deserts have 10+% humidity (moisture).

  24. Could a peltier cooler/freezer be used? If not as an enclosure than as a dehumidifier/airtreatment process?

    ->Freeze the moisture right out of it.

    I say this because I assume a desiccant filter wouldn’t be cost effective?

  25. You would have to have NO humidity. Or darn close at the air/printhead interface. Then, 100% humidity (or at least a good amount) to cure the cyanoacrylate. How do you separate the atmosphere of the printheads from the model you are printing, which is only a few millimeters away?

    Maybe smarter people than myself can figure it out – it very well may be possible. I just find there to be easier ways (like UV curing) to go from a liquid to a solid given the process requirements involved (fast cure time, moisture present in normal atmosphere, ink jet clogs being VERY expensive and highly likely, etc.

  26. “That’s pretty freakin’ cool. I wonder how hard it is to get the support material out of the tight spaces in more intricate moving parts?”

    Not terribly difficult. The support material is basically a hydrogel, similar perhaps to a contact lens if you want to think of it like that. It crumbles as you can see in the video. High pressure (500 – 1500 PSI) water is used to break most of the support material. Then you can soak it in lye/water to remove the last residue. Rinse and that’s it. Harder than say melting away a wax but not as bad as SLA or metal SLS (which involves a bandsaw to remove the printed item from the metal base and then dremel work to remove remaining support structures, if any).

    Average time to clean up a polyjet part depends on the complexity of the part. Big, highly complex parts may take 20 – 60 minutes. Especially if there are very, very small and therefore brittle parts. I would say an average polyjet part takes no more than 10 minutes of post finishing time.

  27. @PolyJetter:

    I’ve been very interested in the idea UV curable resins for quite some time now (both for a home-built SLA or polyjet system), but I haven’t had much luck finding inexpensive resins that one could use for testing/tinkering. I imagine that there are also a number of fellow reprappers who are quite interested in learning more about these materials (probably even a few lurking on the reprap forums).

    A few of my 3d printer projects have been featured on hackaday in the past few months — both a makerbot/reprap-style fused deposition machine, as well as a prototype for a SLS machine. I’ve been interested in developing home-built stereolithography or polyjet projects, but the material is something that I just don’t have a lot of experience with or much luck researching. If you’re up for it, would you be able to either drop by the reprap forums, or to contact me through my website, to share a little more about the kinds of UV curable liquids that are required? (and, potentially, where one might find inexpensive versions of them?)

  28. A HUGE amount has been written about UV curing chemistry. Books and publications and journal articles for over 30 years. It is, as with most things, complex to deal with but fairly simple to get the basics. The world of UV inks though is GIGANTIC. The amount of raw feedstock numbers in the thousands. That’s off the shelf raw material products available right now. Most are not suitable for use in UV inkjet chemistry for various reasons. But some things that make no sense in UV chemistry are what you need that only becomes apparent later. There are epoxy (BPA and BPF), urethane, polyester type chemistries for the oligomers as well as all sorts of different kind of monomers. Each one imparts a whole host of different properties on the final material and sometimes they combine in unexpected ways. Brittleness, adhesion, surface energy, viscosity, stiffness, tensile strength, elongation at break, surface tension, etc are all controlled by the selection of monomers and oligomers.

    Then there are fillers, pigments, suspension aids, dispersion agents, etc. They have to match the density or have the right functional end groups added to ensure stability… so you can see how the chemistry can start to get a bit complicated pretty quickly.

    The problem is sourcing the materials. Most vendors (and there are not too many out there) want to sell you a drum of material at minimum. You can’t just call them up and order a few pounds, although some vendors will offer samples at nominal cost. In general, you have to source at minimum, 3 different products (monomer, oligomer and photoinitiator) although realistically, the number is 5 – 7 or more to make an actual workable ink.

    Then, assuming you can come up with an ink, decide to buy 220+ gallons of it (at the cost of several thousand dollars), you have all sorts of other issues to contend with.

    You have to filter the material down to 1 micron or less, or else you will clog your nozzles. You have to be able to test and control viscosity across very exact temperatures. You have to be able to do real time droplet and satellite formation visualization to make sure the drops you are forming are accurately falling and recombining. You have to make sure the materials interface together properly. You have to source the industrial inkjet heads and build an interface to drive them.

    Once you source thousands of dollars of chemicals, you have to actually physically mix the ingredients together, which isn’t always easy. Some are powders, some are as thick as liquid glass and others are as thin as water. It’s not as simple as just pouring stuff together and stirring. Did I mention everything is UV sensitive?

    Then once you have it all together, testing is difficult because you need rigs to do UV exposure testing and you have to avoid UV exposure to your skin and eyes. I once got a nasty sunburn from only a few seconds of exposure. As bad as if I was out in the sun all day. You don’t even realize it until a little while later.

    So, in short, there is a fairly good reason why people have not been experimenting with UV chemistry just yet. Like most things though, it IS possible. It just isn’t easy, something you can do on a shoestring budget or something you can just start doing without getting at least a bit serious about it – investing in the right equipment and doing the right research.

    Which explains why a large X,Y,Z gluegun (makerbot, repraps, etc) are becoming so popular. The “ink” is a single, inexpensive component, which cures by cooling.

    If we can get a recipe list of where to buy UV inks (or a reseller who is willing to supply small quantities of either bulk purchased raw material or reasonably priced finished inks) so the guesswork is taken out of the ink process (AND the inks can be used in off the shelf printers due to room temperature UV jetting being possible) then maybe the rest of the issues could be tackled in due time as they are more of the basic details that have already been solved by cheap desktop printers and reprap style work.

    BUT, somebody has to invest the time into creating inks AND then decide to sell them for pennies. Not that I agree with the manufacturer’s decision to sell their inks for 20 – 40 times what it costs to make them (or more), I understand that the work that goes into their ink creation is substantial. Last I heard, Objet had at least 15 full time chemists on staff working on creating new inks if that gives you any idea as to how generally complicated formulating inks can be.

    Which, again, isn’t meant to be a discouragement. I repeat that I would absolutely love to see an open sourced UV printer and if the ink hurdle is tackled (which I believe I have done) then I see no compelling reason why something that would rival $150,000 machines in output quality could not be built for the low 4 figures range and run for very reasonable consumable prices.

  29. To the “where are the hacks” people: read the comments, they’re building something in here.

    Something epic it sounds like. Anyway, when you guys are done I think we should compare your nickels to the nickel in this video to see who prints the better nickel.

  30. While I believe that the inkjet technology is possible to homebrew…the “inks” are definitely a barrier.

    Build times with my solidscape are dreadfully long in comparison to my invision…The solidscape uses one jet per material scanning the print head in x and y lowering the stage (z).
    Whereas the invision uses multiple jets to print its field in a single Yaxis sweep raising the printhead (z).
    So the two technologies are not unlike inkjet vs laser printer in speed.
    A reprap with a pair of inkjets instead of a extrusion head wouldnt be far off of the solidscape system.

    The DLP system having a single axis and a complete field of exposure seems like it would be the simplest to DIY. Suitable UVcure materials are available in a number of different durometers.. http://www.star-technology.com/cgi/htmlos.cgi/003382.1.2245418094414069950/epoxies/star-tech-catalog.html
    while you would lose the continuous build that gives the perfactory system its continuous z-res blending… lifting the part after exposure then lowering onto another stage loaded with a different material could possibly allow a support material or even a material of a different durometer….not unlike a screenprinter doing multiple colors.

  31. @mikey…they didnt print a nickel. the polyjet machine doesnt print metal the one “metal” part which shows the reflection of the nickel….has been hand sanded and then plated…its rather funny that the narrator makes a point of the surface quality…you can sand and plate a zcorp just as easily…and it would be just as shiny and smooth.

    It is also important to consider is that many of these multimaterial parts are useful as aesthetic prototype or one offs….however the cost of the materials and speed of these printers do not make Rapid Manufacturing economically feasible…This might not be of great issue to the tinkerer but many of these parts cannot be reasonably reproduced through traditional manufacturing techniques.

    Until these technologies reach a state of ubiquity, when I can fill a room with units for the price of one of my machines…designing with the inherent limitations of injection molding and machining, while not as fun, is a valuable and necessary practice.

  32. “very expensive heads” – well I can get 3 inkjet heads (ie. 3 colour) for $5 now… If those cheap heads can be used, then risk of damaging the heads isn’t a problem. Even if you could only print 1 part per replaced head, it would still be pretty cheap.

  33. The stock replacement heads for my solidscape run $1200 each.
    inkjet might describe the technology used….but The jets themselves are far more robust then anything youll find in a printer.
    You wont print 1 line let alone one part with a $5 printhead unless you make a huge breakthrough in the materials.

  34. heh.
    nice work.
    as for other technologies, you can indeed thin down epoxy with a solvent but it needs to be nonreactive or it will mess up the curing process.

    interestingly it is possible to premix epoxy and hardener and “water it down” so to speak, it will still react but far slower.
    Should give enough time to deposit, as the solvent evaporates (may need applied heat) the components react and set solid.

    the same trick also works with silver paint, if you want a wide distribution of particles.
    as the solvent evaporates they will “clump” together and become conductive.
    supposedly UV light can stimulate the particles t form conducting networks through epoxy and generate interesting structures.

  35. I can’t believe people are actually against posting this.
    This device would be a godsend for making any sort of casing or parts you needed for modding.
    This is highly relevant to the finalization process of most hacks: housings and support frames.

    I loved that rubber-dotted casing, that was pretty amazing the detail you can get on it.

    Eventually, we will get our replicators, maybe not as advanced, but still pretty cool regardless.
    I’m hoping for the food type myself. But there we could see a whole new form of attacks on people, making them download food packs with horrible parts in it. (like high salt content)
    Will we have some sort of AV for 3D printers? Would be pretty nasty building a board and literally building virus binaries on the chips…

  36. Well I’ve managed to acquire, for ~$50 a selection of samples of potentially useful adhesives and resins, 100 used/blocked printer cartridges, and a few inkjet printers for parts. I’ve also got some micro-heaters to try to reduce viscosity with a hot-head. All the stuff will be delivered over the next 2 weeks, and I’ll post back if I make any progress.

    My aim for this project is to make something that is super-low cost to operate, so I can leave it running 24×7, and be worrying about it’s electricity cost more than it’s materials cost.

  37. @hunnter….its a 5yr old machine its hardly news.

    if the machine had just come out..
    if they had just made some advance in resolution…

    Id much rather see a post about an automated pizza printer some one built then to see what amounts to an infomercial (he even refers to growit as his sponsor)….

    As it stands this discussion thread…largely taken on by those of us annoyed with fluff….has been the only content of relevant value….

    oh and I checked with objet there no arduino in the connexx series printers ;P

  38. Sadly, Objet’s control boards are all custom made. There are actually several boards controlling them, including an embedded computer with a servo controller board. It runs windows and of course, the software is closed source. I am not sure if the arduino is able to control linear encoders and brushless servos well enough to be able to control something like this.

    Honestly, if you want to go DIY with as little fuss as possible, don’t even bother jetting materials. Just do the EnvisionTEC route and use a single DLP chip, an off the shelf projector bulb (which is what they use only they modify it slightly to make it harder to use – although it is still possible). Software like Magics can generate support structures for you, although it is closed source and rather expensive.

    If you go the DLP route, then you don’t have to jet anything. In fact, EnvisionTEC’s materials are better because they can do nano materials, with improved heat deflection, clearer materials, stronger materials, etc because they are not limited by viscosity like Objet is.

    Nobody out there is going to say this but I will. Objet’s dirty little secret is that while they offer a variety of materials, the materials they have are all “watered down” with monomers to push their viscosity into the jettable range. They can’t use the materials they would like to because they can’t jet them. Some of their materials are rubbery but the rubberyness is actually a side effect of lowering the quality of (by adding “thinners” aka monomers to (in this case) urethanes oligomer “strong” material. It doesn’t take much effort to rip or tear their rubbery materials.

    If the DIY community went DLP, just expose a layer, coat, expose, coat. No jets to break or clog. No fancy control for X and Y – an Arduino should have the ability to power it. Cheaper parts, especially given that the inks could be made outside of Objet’s ink patents. Faster (build the entire model layer at once vs do strips), larger build envelope (you could even trade off quality vs speed if you wanted to), higher resolution (seriously, the resolution in X and Y is better than Objet’s and has the ability to scale better as higher resolution DLP becomes available).

    The only downside? The single resin and lack of support material, which means supports have to be removed by hand. Not too terrible really – you can always sand off the support connection points – but if the geometry is complex, you can’t get around the fact that it is going to leave SOME marks and it is going to take SOME time to do so.

    But in the end? For Ultra High Quality, fairly cheap materials, cheap to build machine? Totally worth it. If it were me and I wanted to advance the state of the art in 3d printing, I would build an open source DLP based machine using UV resins. Not sure if a reprap would be able to make parts for it, but a machine like this would quite literally have the quality and resolution to print most of the parts for other second generation machines. Slap in an arduino, DLP chip, projector bulb, a Z axis motor and a control board and you just copied a machine.

    And if you bumped the quality down and the thickness up, you could make something with straight walls VERY fast, even if it was fairly large (larger than most commercial build envelopes, save for very large format SLA machines). You could even speed it up when making things where stairstepping wasn’t an issue (like a case that just needs to be X inches deep) and then slow it down for curves and detail work. I don’t even think the commercial machines do that.

  39. I’d just like to say thanks to PolyJetter for sharing all this information. I can only hope that someone goes ahead and uses it.

    I’ve always felt like the rep-rap type machines seemed a bit silly. The idea is great but I just always thought the actual product they spit out isn’t worth the effort. If the method PolyJetter is suggesting would create higher quality objects (as in better resolution) it would be awesome to see the open-source community run with it.

  40. The key patents covering plastic extruders are expiring. The key patents covering this sort of resin curing are going to hang around for a little while yet.

    Manufacturing printheads for resins is sufficiently difficult that it is out of reach of hobbyists at this point. Some repurposing of inkjets is possible; some simple DIY versions of the z-printer using water in the printer and a powdered feedstock which is glued together by the water do exist, but they’re fragile, slow and simple.

    SLS is conceptually and mechanically waaaay simpler than this sort of UV resin curing, yet you don’t see a whole lot of that at the DIY level either. It’s hard. There isn’t an army of highly skilled and totally altruistic engineers out there who are just too lazy to make these awesome things for you.

    Reprap quality will continue to improve. It’ll catch up the likes of stratasys given time… you can either assist or pay up for a commercial offering, but whining about it accomplishes nothing.

  41. Very bad ass. Yes we know it’s not a hack but these machines were born from hacks. Someone had to take a CNC machine and make it function as a printed instead of a lathe.

    This is still cool regardless of it not being some DIY hack. Sometimes there just isn’t anything for them to make articles about people! Don’t freak out so much, when they have a hack worth posting – THEY WILL DO SO. Relax….breathe…relax…now breathe again. It’s going to be alright.

    Thanks Hack-A-Day for at least having new material for me to read! You guys are doing a great job.

  42. Replace the materials with cell cultures – skin, muscle, nerve, blood vessels, etc and you can print someone a replacement limb or other organ from their own cells.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.