Future or Fraud: Conveyor-Belt 3D Printers

Photos courtesy of BLACKBELT 3D BV                                   Sketches are original.

New conveyor-belt styled 3D printers have recently made their debut in the 3D printing world. They seem to offer what we've been missing out on all this time: unlimited print volume, potential for mass production, unsupported overhangs, and so much more. Question is: do they work?

This is Future or Fraud.

A Little Bit of Background

This idea of continuous printing isn't new: it has already been seen in the early days of the RepRap project, and Stratasys has even patented the Automated Build Platform by Makerbot. But never before has there been a full-fledged, fully self-reliant belt printer that you can buy -- until now.

The first major breakthrough with these belt-styled printers came along when Stephan Schürman, previously a 3D printing engineer at legendary filament producer Colorfabb, announced the Kickstarter campaign for his new Blackbelt 3D printer. It was a pretty (but not wildly) successful campaign, finishing off with €101,970 pledged of a €50,000 goal.

What's surprising, though, is what happened during the funding period: the renowned 3D printer company Printrbot paired up with Polar 3D and "'bust[ed] [their] butt[s]'" to get the Printrbelt model launched -- before the Blackbelt campaign could end (quoted from this video). Patents and legality aside, this raises a few eyebrows for any 3D printing hobbyist, especially considering the glaring design similarities between the Blackbelt and Printrbelt.

Was this just coincidence, or did Brook Drumm, esteemed CEO and founder of Printrbot, see something in this technology?

 

On the surface

Both the Printrbelt and the Blackbelt feature a conveyor belt for the print surface, allowing for incredibly long prints and/or an endless line of models constantly rolling off the printer. They also have a unique motion gantry for the printhead, positioned with a 45o angle of elevation from the print surface (can't really call it a normal print bed, can I?). This allows for unsupported overhangs (more on that later) that *might* save you support material with a better surface finish.

Throughout this post, I'll be going through what each of these new designs has to offer, as well as the potential worries they may bring. Yes, the new gantry design has nothing to do with the conveyor belt, but since we have yet to see any other commercially available belt printer with a different design, I thought it'd be fair enough to include both in this analysis. Don't worry, I'll make sure it's clear which of the new designs is the source of praise or worry.

Because of how much I must cover, this post may get a little long, so here is an outline (with links) that will help you jump around:

 

Introduction

Background

Overview


What's good:

Unlimited Print Length

Unsupported Overhangs

Potential worries:

Build Plate Adhesion: First Layer

Build Plate Adhesion: Print Surface

Print Quality: Surface Finish

Print Quality: Layer Resolution


Final Verdict

Acknowledgements

Reference

The Pros 

Unlimited print length

Thanks to: conveyor belt design


This is the whole point of using a conveyor belt: having an "unlimited" z-axis. Blackbelt 3D market it as the perfect solution for printing prosthesis, large format furniture, signs, extrusions, sheet fabrics, etc. They even sell a roller table add-on for insanely long prints, like this massive Blackbelt sign they printed:

The Printrbelt doesn't feature such an add-on (yet), but I'm sure many people will come out with custom, DIY rollers to take advantage of this feature.

But there's more.

 

Perhaps the most groundbreaking revolution this can bring is series production. For years, 3D printing has been a form of rapid prototyping -- you can rapidly roll out prototypes and iterations without having to wait on external suppliers and constantly invest in new molds. When it comes to mass production, however, injection molding is unrivaled. Though entailing a hefty investment, it can produce parts near instantaneously and continuously. Relatively speaking, a 3D print takes forever -- and parts need to be manually removed before a new print job can begin.

This conveyor belt design opens numerous doors to mass production. It removes the need to constantly clear the build plate and start a new print: if you need 50 copies of something, just go ahead and print 50 copies. Come back in a day or two, and everything will be sitting in a bucket at the foot of the printer. It's that easy. (Assuming prints don't fail, of course).

No, this new design can't resolve the fact that prints take an extravagant amount of time, but for small startups not ready to invest in molds, it might be the perfect solution.

I speculate that conveyor belts are the future of 3D printing in series production. Plus, you can print some pretty long models if the need ever arises.

Unsupported Overhangs

Thanks to: 45o gantry design


Another attractive feature of these new printers is their ability to produce sharp overhangs  (even 90o!) without any sort of support material. No more messy undersides, forceful support removal, or wasted plastic! How exactly does this work?

A common guideline for printing overhangs is 45o: it's an angle that most printers can manage, while those with better cooling setups and more finely tuned settings will be able to go up to around 70o. However, by turning the entire printhead gantry so that it's 45o from the printbed, each layer comes out backed against the previous. Thus, angles are formed by controlling how far down each subsequent layer goes. Because of this unique approach, overhangs don't print like overhangs anymore.

Seems innovative, right?

 

Turns out, there are a few gaps in the logic -- it doesn't always work! In fact, this "magical" trick can only be pulled off when the model is designed and oriented such that the underside of the overhang is facing the printhead. If carefully considered, that's the only scenario in which the overhang is actually produced by the method previously described. Anything else and the plastic will be printed into thin air, not connected to the build surface or backed against a previous layer. This means that only specific models with specific orientations will actually be 100% support-free.

In fact, Blackbelt 3D explains this in the design guidelines area of their site, using the following pictures:

Plane's head prints into thin air.

New gantry design (left), traditional setup (right)

 

As can be seen, though less support material will be required when compared to traditional gantry designs, models aren't always unsupported.

Does this mean that all this is just a fake? Not really. In some cases, you will be able to get away without support, while other times, it'll only be able to reduce the amount needed.

Either way, I think this design is respectable: it's not perfect and doesn't necessarily live up to high expectations, but it does its job.

The Worries

Build Plate Adhesion: First Layer

Source of worry: conveyor belt design/45o gantry design


Note: this section goes over how well the machine can lay down the first layer. For how well the model can actually stick, please see the Bed Adhesion: Print Surface section.


The first layer of every 3D print is arguably the most important one. It will make or break the print -- and I have my qualms about these new designs.

Both the conveyor belt and 45o gantry present separate challenges for a successful first layer. If you look at a "traditional" FDM setup, the nozzle points straight down towards a stiff build plate and squirts out a layer of plastic. The height between the nozzle and build plate is adjusted so that the plastic squishes down for a solid connection. This lays a firm foundation for the remainder of the print.

With the new setup, however, the nozzle approaches the build surface at an angle. This means that the machine can't obtain the same, secure "squishing", leading to a flimsier base that may detach. The sketch below provides an exaggerated view to emphasize my point:

The printer lays down a line of plastic down the length of the bed.

 

As can be seen, the plastic cannot be fully pressed onto the printbed the way a "standard" machine allows. Yes, some squishing will occur, but as the circled areas of the header image show, it's not perfect. This may or may not ruin your print; it boils down to the details of the engineering.

The conveyor belt may also present its challenges with the first layer. As with any belt that bends over rollers, it's flexible: so it's not exactly easy to print onto. Of course, both the Printrbelt and the Blackbelt have a raised platform under the belt, holding it taut. This should, in theory, resolve the issue, but Blackbelt goes one step further and makes their belt out of a carbon fiber composite with "intrinsic stiffness and low thermal expansion" (quoted from their Kickstarter campaign). This challenge isn't impossible to overcome, and I believe that Blackbelt 3D has done an excellent job, but once again, it comes down to each product's design.

 

So, is this a show-stopper? I honestly don't think so. With some careful design and finely tuned settings, this setup could perform just as well as the printers we are accustomed to. However, performance will rely on product design more so than "traditional" setups.

Build Plate Adhesion: Print Surface

Source of worry: conveyor belt design


Note: this section goes over how well models can stick to the print surface. For how well the first layer can be produced, please see the Bed Adhesion: First Layer section.


After you get the first layer down, you need to make sure the print will stick. And this proves to be yet another unknown of this new technology.

Most printers with this point in time come with some sort of aid for bed adhesion: a heated bed, BuildTak, blue painter's tape, PEI, etc. With conveyor belt printers, the fact that the first layer isn't very secure only magnifies the significance of bed adhesion.

 

Based on my observations and research, the Printrbelt and Blackbelt both incorporate adhesion solutions. The Printrbelt does have a heated bed (more on that later), as Brook Drumm reveals here, with what appears to be Kapton tape on top. The Blackbelt, on the other hand, has the aforementioned carbon fiber belt with "the necessary surface adhesion to print on". I'm not sure how well that carbon fiber will work, and it shows some very slight warping in the header image above, but this looks convincing:

 

A print being held up with only the edge (impressive surface adhesion!).

 

So surface material should work fine -- but what about high-temperature or warp-susceptible plastics? If you've printed ABS, polycarbonate, or nylon before, you know that printing onto a cold surface is guaranteed to fail (unless you have some magical coating coupled with incredible luck). So that's where a heated bed comes in . . . until you realize that belt + heated bed = failure.

Okay, maybe the word "failure" was a little harsh. But there is a serious flaw in the logic: a heater can only (with easily available technology) heat the top section of the belt. That means that most of it will be unheated -- so when new belt material rolls on, it needs time to heat up. With the belt constantly moving (albeit slowly and in steps), the entire front section of the bed won't be up to temperature. Layers laid down on this area will have a high chance of popping off, especially with high-temperature plastics.

 

Thus, it looks like high-temp materials may be a stretch. Only time will tell if the bed adhesion will suffice, but my prediction is that it won't be the best.

Print Quality: Surface Finish

Source of worry: 45o gantry design


Note: though layer resolution plays a key role in surface finish, this section focuses on the printer's ability to produce crisp lines, not how fine the lines are. For information on layer resolution, please visit the Print Quality: Layer Resolution section.


The crocodile above is pretty darn clean. In fact, it's really clean. So why is it that this seemingly Ultimaker-level print quality is included in the "Worries" section?

Well, once again, there's an engineering hurdle that once again boils down to product design - top surfaces. With the "usual" setup, the nozzle is parallel to the build surface, so top and bottom surfaces tend to be pretty flat. A gantry at 45o however, alters the print process such that flat tops are produced by ending layers at a constant height from the bed. To highlight this, I've (once again) created an exaggerated sketch:

The blue line is a constant height from the base. Inconsistent layer length leads to an inconsistent surface.

 

This is a pretty blatant flaw that is inherent to the gantry design, but it is by no means impossible to overcome. With careful designing, quality components, and refined firmware, it shouldn't be much of an issue. Blackbelt 3D seems to have done a fine job.

 

So, I'll be honest: this flaw is pretty picky. Printers we have now deal with the same issue when it comes to lateral (side) surfaces, and results have been fine. However, the redesigned gantry doesn't resolve lateral surfaces; it only adds top surface inconsistencies, making prints that much more unpredictable.

Print Quality: Layer Resolution

Source of worry: 45o gantry design


Note: this section focuses solely on layer resolution, meaning how thin each layer can be produced. For surface finish in general, please visit the Print Quality: Surface Finish section.


Layer resolution is key when it comes to print quality. Sure, 200 micron prints can come out just fine -- amazing, in fact. But 50 microns is going to look better.

Unless, of course, you haven't tuned your settings.

 

So what is it about this new gantry that may throw off layer resolution? First, consider how a "standard" printer controls layer height: the printbed is moved down or the printhead is moved up with either a lead screw or belt/pulley system -- both of which are precision-made. The lead screw has a carefully machined thread for constant pitch, while the belts and pulleys have the same careful machining applied to the grooves. This makes for an incredibly precise and consistent motion system.

The conveyor belt and new gantry switch things up because layers are no longer parallel to the build surface: they're at 45o. This means that layer height (or now thickness) is controlled by the motion of the conveyor belt; a 50 micron motion forward creates space for a 50 micron layer. And that's when things start getting absurd.

You have a massive, flexible sheet wrapped around unprofiled rollers, held in place only by tension. A single slip of engineering will wildly throw off that 50 micron motion. The belt wrapped too loose, and it will induce slippage. Poor firmware control, and the 50 microns will never come out exact. Cheap stepper drivers and motors, and 50 microns would be impossible in the first place.

 

As with many of the previous points, this can be "easily" taken care of with meticulous and thorough product design. However, it does bring a whole slew of complications, which could mean two things for you: inordinate inconsistencies or excessive price tags.

The Verdict: Future

In retrospect, there have been quite a few worries with these new designs. So why is this labeled as Future?

All things considered, this new conveyor belt technology is opening so many locked doors in the 3D printing universe. It is finally enabling additive manufacturing to grow out of just rapid prototyping. It is finally enabling this technology "'to revolutionize the way we make almost everything'", as President Obama once put it. When Henry Ford started his assembly line, not everyone was impressed -- but it turned out to be the birth of mass production as we know it.

This is the birth of mass production for 3D printing.

 

Of course, there are uncertainties. But going back through, many, if not all, of these uncertainties boil down to attentive engineering. Yes, they bring along complications we don't see with orthodox setups, but with the proper designs, products will come out fine. They are hurdles that can be overcome, just like with a standard printer.

So far, we've only seen the conveyor belt coupled with the 45o gantry, and the latter seems to be the source of most of the worries. The belt itself, the one that's finally opening doors to mass production, is sound. As these designs progress, I'm sure new and improved gantries, either 45o or standard, will make their appearance. And when these 2017 belt printers seem obsolete, they will be regarded as the dawn of mass production in additive manufacturing.

 

This is an emerging technology.
It carries its qualms, but it has a future.

 

 

 

Acknowledgements

Every article has contributors, and they deserve recognition. I would like to thank Blackbelt 3D for making this article possible by supplying the photographs used. All photos are courtesy of Blackbelt 3D BV (sketches are original).

All text in this article is published under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. To request usage, please fill out the contact form.

14 thoughts on “Future or Fraud: Conveyor-Belt 3D Printers”

    1. Thank you so much for stopping by! It’s truly an honor to have the founder of Polar 3D post here.
      If you have any additional feedback or words, please feel free to leave another comment or contact me.

  1. Could some of these issues be mitigated by taking this concept to eliminate the issues (but still have the support problems) of a straight up/down gantry? In other words, could the conveyor belt still be there but fit it to a standard printer I imagine a Prusa I3 type printer could do something similar. The Y axis could still move in the direction it does, but it is a conveyor belt. Everything else is the same mechanically. The belt could move forward and backward (which should actually give a smaller footprint) and when a piece is complete the belt can just roll forward for the part to ‘pop’ off. I do imagine that heating will present challenges, but I am thinking that there is even a market to retrofit existing printers in this fashion so long as they have a bed that does not move in Z. What do you think?

    1. Hello Samer,

      Thanks for stopping by!
      Having such a design would take care of some of these issues, but if you think carefully, it would also make “infinitely” long prints very difficult and restraining. Each layer must be comprised of only one part (no split sections, i.e. vase-mode models), and the bed would have to roll back and forth, constantly popping the first layer off the conveyor belt before winding it back for the next layer. This would most likely fail entirely. Thus, that design would only enable a constant production of size-constrained parts, just as you described.
      Hope that helps!

      –Hironori Kondo

  2. Hmm it looks like your site ate my first comment (it was super long) so I guess I’ll just sum it up what I had written and say, I’m thoroughly enjoying your blog. I as well am an aspiring blog blogger but I’m still new to the whole thing. Do you have any recommendations for novice blog writers? I’d genuinely appreciate it.

    1. My apologies; I’m not sure how that happened. I’m glad you like the content!
      If you’re an aspiring blogger, try to stay fresh on ideas, and offer your unique perspective on things. There are so many sources of information out there, and without your view, you aren’t really bringing new information to the community. Even if your information or project is not entirely new, try to give it your own twist!
      Hope that helps and good luck with your blog!

  3. Generally I don’t read article on blogs, however I would like to say that this write-up very compelled me to take a look at and do it! Your writing style has been amazed me. Thanks, very nice article.

  4. I like the valuable information you provide in your articles. I’ll bookmark your blog and check again here frequently. I’m quite sure I’ll learn many new stuff right here! Good luck for the next!

  5. Thanks – Enjoyed this article, can you make it so I get an update sent in an email every time there is a fresh article?

Leave a Reply

Your email address will not be published. Required fields are marked *