Friday, March 31, 2017

For Those of Us Still Considering a 3D Printer

Ten months ago, I ran into a friend and fellow engineer I used to work with.  It was at a restaurant/bar not far from the beach where I had gone to hear a band play, and I hadn't seen this friend since I retired.  He was playing keyboard for the band (another friend was the guitarist).  Among his very first words to me were, "do you have a 3D printer, yet?"  He knows I'm a home machinist, so I must be the logical guy to ask for input.

I still don't have one.

This week, Machine Design ran a series of articles on the topic of entry level 3D printers; all three by author Cabe Atwell.  I had never really noticed this guy as an author, but looking at his author's page at MD, I see I've read some of his stuff before, and also that he's a principal at a company called Gunhead that has two divisions: one that makes CNC lathes on the ubiquitous "7 by" platform while the other makes paintball products.  Which makes me think we'd probably get along fine.

The first one that caught my eye was "Are 3D Printers Overrated?"  In the first couple of paragraphs, he mentions something I've only heard as hushed whispers among the hoopla over the technology.  The bitter truth is that you can buy a 2D printer (better known as an inkjet or laser printer) set it up and get great results instantly.  3D printing is still a much, much less established technology.
You’re probably groaning by now, but hear me out. Yes, prices for the maker/desktop/DIY category of machines have dropped, the quality of their output has increased, and promotional materials abound with printers creating beautiful and delicate objets d’art.

What isn’t shown is the effort, the sweat, the tinkering, the trial and error, and the screaming to the gods themselves to please, please let it work this time.
He then goes on to list eight trouble spots that the home level printers have to address before they can really be thought of as useful appliances.
  • Warping.  Caused by excessive sensitivity to temperature changes.  Enclosing a printer works better than going with an open frame.  On a similar note; printers must have their beds aligned carefully; leveled so that print head is at right angles to the part, and the gap between the print head and object has to be set. 
  • Porosity.   Even when printed with 100% fill, parts leak too much for many applications.
  • Prints that won't stick to the print bed.  As Atwell puts it: "Reapplying treatments or specialty surfaces is too much like inking the letters on a manual printing press. There is no equivalent operation necessary before you print out your great aunt’s Bundt cake recipe on your inkjet."  
  • Prints that won't come off the print bed.  They stick too well.  
  • Prints don't get out of the way.  Partly an extension of the last issue, partly due to there not being something like the "out tray" that holds the prints coming out of a 2D printer.  3D Prints don't get ejected and left for pickup.  This makes the 3D printer require more babysitting than a paper printer.  
  • 3D Printers are way too slow.  For a variety of reasons, it takes hours to a day or even more to see if the print came out right. 
  • 3D Printers are way too manual.  Think of a laser printer: you buy it, and you buy a stack of paper.  You stick the paper in the printer, follow the setup instructions, and it prints.  As he puts it, 3D printers are, "more like MP3 players that get songs loaded onto them via Morse code. Oh, and the Morse code tapper thing costs extra."
  • And finally, the BIG one... (say it with me) Software.  "A decent selection of software is available, and we’ve all seen the impressive results. The problem is that the software, if it’s user accessible, requires a lot of tuning. Adjusting variables and trying to correlate cause with effect can be maddening."  
While this makes the printers seem exasperating, things like this need to be said.  If people aren't feeding back to the industry that they expect more out of the products, the pace of improvement will be slower.  The developers need to see the products through the eyes of their customers.

The next piece of the three almost seems to contradict this one.  Almost.  This one is entitled, "Ten 3D Printers that are Near Click and Print Capable Right From the Box".  In the first piece, he mentions two printers that are close to his ideal, the Tiertime UP-BOX and Afinia H series of printers, both of which are rather expensive.  In this piece he lists another ten, from a couple of relatively cheap consumer-grade printers, starting with the Tevo Tarantula I3 at $237 and the QIDI Technology X-ONE at $449.  The first thing you'll notice about the Tevo Tarantula is that it's not enclosed.  See his first item "Warping".  There is a Prusa USA i3 printer at $699 which is also not enclosed, but the selection runs the gamut to an Ultimaker3 at  $4295 and only three out of the 10 are not enclosed.  For those, he says, "build an enclosure".
(The Tevo Tarantula)

If you're seriously looking for a 3D printer, you might well find one in these 10 that will work for you.  The prices above seem to run the range; you'll also want to consider that they might not print in all the same materials.  That Tevo Tarantula prints in ABS, PLA, PETG, wood and PVS filaments, while a $699 (enclosed) XYZPrinting daVinci Pro 1.0 will only print in ABS and PLA (although those are the most popular filaments).   
Finally, the third column, "Essential Equipment for your Filament Based Printer" covers just that: 11 essential accessories that (chances are) you don't even know you need for your printer.  From BuildTak, a mat that goes on your printer's bed to help you avoid the problem of prints that fall off too soon, to a spatula-like print removal tool, to more mundane things like Rubbermaid sealing containers, and Super Glue.  Don't forget a set of feeler gages.  

Two of the main reasons I don't have one yet are the phase he describes above ("the effort, the sweat, the tinkering, the trial and error, and the screaming to the gods themselves to please, please let it work this time") and the long wait times to get something simple printed out.  It's cool stuff, but we're not really ready for a "3D printer in every house" world.  We're not ready to download files of plastic things we need from Thingiverse instead of buying them from Wally World, although that world is coming.  A fast working 3D printer that address the user friendliness issues that Cabe Atwell raises would be a good approximation of a Star Trek replicator.  Need a replacement part for something that broke?  Maybe you tell the replicator what you need and it goes to Thingiverse (or something like it), looks up a design, and prints it for you.  To borrow a phrase from Sci Fi author William Gibson, The future is already here—it's just not very evenly distributed.”

12 comments:

  1. I've got some 3D printed reproduction parts for my Supra made by one of the guys on the celicasupra.com forum, and they're pretty good.

    It took him numerous tries to get a file he could use that accounted for dimensional differences, and the surface finish isn't "factory perfect", but for parts that are 30+ years old, and no longer available, they're a God send. The originals in the car from Toyota get dried out and very brittle, and your chances of having the literally fall apart if you have to remove them is very real.

    I'll take some snaps and post them this weekend.

    Jay Leno is very good at keeping on top of this for his automotive restoration shop, so I'll embed one of his recent videos from "Jay Leno's Garage" on YouTube.

    ReplyDelete
  2. I never saw 3D printing as ever approaching plug-and-play, even a caveman can do it. Just as a fully equipped machine shop doesn't make one a machinist, truly revolutionary 3D printing takes experience and understanding.

    Sure, there will likely be a day when printers are capable of relatively mundane tasks with little skill but the days of taking a CAD file and "figuring out what you mean vs what you say" are likely pretty far off.

    I say this because I work with a guy who is deep into additive mfg. He constantly struggles with people who don't understand why they have to provide him with a charge code to tweak their model prior to printing. They get frustrated when he comes back and says, "This isn't going to work as drawn but, if you describe what you're looking for, we can fix it.". They think he's just the monkey that feeds the machine and removes finished parts.

    ReplyDelete
    Replies
    1. I guess that's what machinists of all levels get. People think they can put a drawing together and "all ya gotta do" is turn the machining center on. There's no appreciation of the fixturing and everything else. Irish ran a parody quote response for machine shops the other day with about 20 things that are always called out incorrectly or unclearly on drawings.

      Still, Atwell brings up some good points about user friendliness. Every step in the right direction helps.

      Says the guy who actually blew a bonus check on a laser printer for $600. Today's $60 printers are better than that one.

      Delete
  3. " Just as a fully equipped machine shop doesn't make one a machinist, "

    dang! Now you tell us!

    ReplyDelete
  4. In my opinion, if you 3D print gun furniture on a low-end machine you will be disappointed with the layer adhesion and surface texture, and on a high-end machine you will be disappointed with price.

    Was it you who split an Aluminum 1911 frame into haves to make the magazine well shallow and easier to machine? You could do the same to CNC cut wooden furniture. Your oiling system might make removal of sawdust easier than it otherwise would be.

    ReplyDelete
    Replies
    1. No, that wasn't me who did the 1911 that way, I just did a blog post on the original story.

      I would think the oil on everything and sawdust would translate into an unholy mess, not better. I guess you mean it would trap the dust and keep it from blowing around?

      Delete
    2. I was thinking the oil system would make it easier to remove wood dust from between moving parts by displacing it with fresh oil. Pump oil, run the gcode to move the table around quickly in a big oval, done. The rest of the sawdust paste is just cosmetic. Sometimes I think the metalworking industry was a sinister plot by the paper towel industry. At least we're no longer using milk for soluble oil.

      Delete
    3. Sometimes I think the metalworking industry was a sinister plot by the paper towel industry. You got me to laugh out loud with this one. I have what feels like a constant flow of those blue paper "shop towels".

      I see your point. I guess I'd have to try it and see what it looks like. Right now, dust from cutting wood seems to settle into a thin layer over everything, over the course of a few days.

      Delete
    4. Right now, dust from cutting wood seems to settle into a thin layer over everything, over the course of a few days.

      The Shopsmith dust collector does a great job of keeping fine woodworking dust out of the air. It's a more sensible packaging/arrangement than everything else I've seen. It's not a shop vacuum, it's tuned for more flow and less suction. I wouldn't feed it oily metalworking chips but for woodworking you'll like it. Here are links for pictures, recommend finding it used on ebay and picking it up in person. The replacement clear hoses are much more resistant to being accidentally stepped on than the original black hoses.

      http://www.shopsmith.com/ownersite/catalog/dustcollection.htm

      http://www.shopsmith.com/ownersite/catalog/dc_hoses.htm

      Delete
  5. To me, 3D printers are like other 'hot' technologies - the people who want you to buy them exaggerate the benefits and underplay, or downright ignore, the disadvantages.
    I've worked with 3D printers, autonomous vehicles, wind turbines, solar power, and other new technologies and all of them have major issues that their supporters don't like to discuss.
    These, and many others, are improving technologies but are NOT ready for the widespread use that their supporters claim they are.
    I'd be more likely to look at Ghost Gunner's desktop CNC than a 3D printer for serious work.
    But then again, I can do what it does with a drill press and a cross slide vise ...!

    ReplyDelete
  6. Oil and wood = big problems. For instance, you can never finish the part, because the oil will prevent adhesion of any coating, or it will prevent the absorption of other oils. Oil plus sawdust equals a horrendous mess that is more abrasive than you can possibly imagine, and horrible for machine parts. And you damned sure don't want to breathe any of that.

    What 3d printing is very good at is making investment casting stuff. You can buy wax filament and make, say, small replacement gun parts, for instance. And then lost wax cast them. http://www.machinablewax.com/product.php?product=52

    ReplyDelete
    Replies
    1. What you say about oil and sawdust is pretty much what I've always heard. Back when I was building the shop, someone was advising me to hang curtains between the wood and metal working areas to try to keep the dust off the metal tools. I have dust collection in the wood space, but it doesn't get everything. The fine, flour-sized dust still escapes.

      I've done investment casting before, and have always thought of using a printer to do that.

      Delete