« Posts tagged FDM

Extrusion Printing – A Quick Review of Strengths and Weaknesses

A stacked build with 47 parts for a 3D printed outfit.

When someone says 3D printing nowadays, they’re almost certainly referring to desktop extrusion printing of a single material, usually PLA or ABS. I recently did some technical edits for a book on 3D printing, and was surprised to see that this idea is so deeply ingrained in media and culture.  One of the points I emphasize in my 3D printing classes is that most of the weight put into this perspective is due to the marketing from companies trying to push printers or filament. As a consultant that puts me in an odd position when dealing with the consequences. Additive manufacturing has great potential, but it is important not to make statements that lead to false conclusions. Media and marketing often strive to present things with just the right level of evasion that will allow them to avoid taking responsibility for the confusion they create. The purpose of this post is to put, in a nutshell, the most important information that printer companies are not excited to tell you, but are essential to make a proper decision.

1.) Books about printing already assume you have decided to buy a printer, so look at other options first.  In reality you need to look at the size, material, and number of things you’d like to print. Home printing can be fair quality, but if you value your time your are probably not going to see a return on your investment unless you print several object per week for a year or so.

2.) Home printers are great for education, but not mission-critical business applications. DIY printers are designed to be inexpensive and repairable at home. Every business I know using a home model for continuous production has at least two, and a technician to operate and repair. If the printer is in regular operation, this is a busy job, since each printer has an “up-time” of about 60-90%. They are great in a support role for larger machines. When the industrial systems are occupied, inexpensive satellite machines are great for test prints and radically increase efficiency. Most small businesses I see using a single small machine have it sitting cold, for whatever reason, about 90% of the time.

3.) Single extrusion is limited. You can print anything…that doesn’t have an overhang more than about 45 degrees from vertical. Extrusion printers must always print on something, either a base plate or plastic on a previous layer. Many times it just isn’t possible to orient a design to print well. An unsupported section can lead to distortion, bad surface finish, or even a crashed build where your output looks like a pile of plastic spaghetti. Using auto-generated support mars the finish is labor intensive. Sharp tools are often used to remove support, making this method unsuitable for children.  Dual extruder machines often use a soluble support material that eliminates these problems. In fact, while it can waste some material, I have had great luck with stacked builds using these machines. One single build can output dozens of parts, saving tons of labor and set-up time. The 6″x8″x6″ build shown at the start of the article contains all the parts for an entire 3D printed outfit (she’s a petite 4’10, 93lbs).

To lay out and print the parts individually would have taken at least five overnight builds. The above stack was printed in about 40 hours. The pieces were roughly separated afterward to accelerate the chemical support removal process, so the whole job was done in 48 hours. Laser sintering works this way as well, with stacked builds being the norm, but easier cleaning (support is powder, removed with compressed air instead of chemicals).

4.) Outsourcing is usually easier. How many parts are you printing, how large are they? If your answers are “many”, and “small to medium”, do it with extrusion. If it’s a one-off project or cannot fit on a printer you might consider a composite of many prints, but accuracy will be low and labor high, so this is really only a solution if it must be done on a very limited budget. Desktop extrusion can be outsourced, too. Printer networks like 3D hubs enable you to connect with anyone in your neighborhood with a printer, and at $0.25 per cc, they make the creation of small but bulky parts very affordable. Note that by “bulk” I am speaking of overall density and the proportion of volume to surface area. Extrusion has a sweet spot in the middle. Beyond a certain size, outsourcing can overcharge for thick parts because they charge for the space inside, which is often sparsely filled. If you have deal with an individual, try to negotiate pricing based on material and print time.

I use extrusion printing all the time, and it is convenient to have a machine at home if you have the time and space to deal with it. Its great to wake up in the morning and having the part you modeled last night ready to go without missing a beat. If you want that feeling several times a week, or if you just want to tinker with the machine itself, DIY is the way to go. If you don’t have a clear idea of what you would make, and especially if you have not yet developed your modeling skills to make whatever you think of, I’d strongly suggest looking at other options – even other manufacturing methods – that might be better suited to your use and lifestyle.

[Edit – July 2015 – 3D Systems has now released a water-soluble filament for home printing. This type of material has been a challenge to create because home printers are not typically enclosed, making it hard to control the temperature, and therefore expansion, of the two different materials (build and soluble support). Another limitation was the caustic chemicals required. Makerbot has been selling a material that was soluble in limonene (that orange-smelling cleaning solvent), but hadn’t been marketing it loudly since it was tricky to work with. The new “Infinity” material apparently needs only regular water and washes away in about 15 minutes! A huge change from the 10-hour cleanings that have been standard.

This new material is available only in a cartridge for the Cube Pro. I do have access to one, and I am interested to give the new 3D Systems material a try in the coming months.]

The Pace of Advancement in Extrusion-based 3D-Printing

There are some wild claims being made about what kind of advancements we might see in 3D-printing, particularly extrusion-based desktop printing during the next couple of years. I’d like to take a moment to examine why I think the predictions, while they might have a hint of truth in what will be accomplished, are not accurate in terms of the timeline.

Before I address the level of innovation and progress, we also have to define what constitutes innovation, and there are some slightly varying interpretations floating around. One recent view makes a very sharp distinction between innovation and invention. Invention being the point where the seed of an idea is brought into reality through experimentation or prototyping, and innovation being defined by the impact it has on us. A big example would be the iPhone, which was clearly innovative in many ways. There was not a lot of invention though, since nearly all of the components (touch-screen, mobile internet, etc.) were already available, and it was just very well designed, built and marketed, leading to incredible impact that most people agree was a great example of innovation. The definition in some peoples eyes then becomes purely defined by impact, which I don’t entirely agree with. I think the innovation was in the integration of the technologies in a conscious and harmonious way, which to me is just an example of Design that simply goes outside of the scope of what most people think of design – “how stuff looks”. The iPhone was innovative because the technologies were mature enough to perform predictably, which made it possible to create a design without glaring flaws, which is what was needed for adoption and impact. So by the ‘impact’ definition, if one of the technologies you choose is not mature enough, or some aspect of culture made it hard for people to accept your creation, or even if some random thing happened like an earthquake that prevented your product from being successful, then that means you did not contribute to innovation. It also would mean, for example, that if a factory that you contracted to produce your device simply stole it, they would get credit for the innovation. I don’t think that definition holds water. In that case both the inventor and the integrator worked together, consciously or not, to produce the innovation.

My definition of innovation is a combination of Invention and what I’m calling Integration, where the invention is gracefully married to the whole ecosystem that allows it to exist and flourish. This is essentially turning something rare and delicate into a robust commodity so that people can conceive of it, access it, and modify it, and so business people can model it, predict changes in its use, and so will be willing to help it succeed. I have put more than enough into this aspect of what I wanted to say, so for further information I suggest reading the works of Simon Wardley. Those ideas spin off into fairly dense economic theory which is not my bag, but you can get a nice intro to his ideas about cycles of innovation in this appropriately themed blog post: Spoiler alert: 3D Printing.

Is 3D-printing innovating at light speed? It may seem that way, but I think it’s been exaggerated. Let’s look at extrusion printers and quickly review their development. Scott Crump invented the whole thing in 1989. CNC machines were available, but had been limited to removing material instead of adding it. His company, Stratasys, came up with the extruder bit, and then a whole wave of other things that flowed from that, like the sparse fill, support, heated build chamber, etc. that were needed to make the output accurate and repeatable. His patent expired and along came Rep-Rap, which beget MakerBot and everyone else. Making printers a commodity is a part of “3D-printing innovation”, but in my mind we should not separate it from the invention stage because it is entirely dependent on it. Without invention there will be no innovation. How many really strong examples of invention are there in open-source extrusion printers? Ones on the level of in-fill, support, and others like Statasys added?  There were a few things with tool paths (mostly from Joris at Ultimaker), and the Thing-o-matic conveyor belt was a seriously awesome idea, but that’s about it. There are thousands of people working on them day and night, and all I see is people making them bigger or smaller or cheaper, or sticking weird materials in them. That fulfills the “adoption and diffusion” portion of innovation, but adding another extruder or changing layer thickness or whatever is not going to bring us to 3D-printed cell phones as people are predicting. There has been virtually nothing done in 20 years! Why, because people outside of industry suddenly care, will the actual technology suddenly leap forward? I’m not saying it won’t accelerate rapidly, but I want to see some examples.

I think it is more likely that new innovations will come as a result of inventions which are not 3D-printers, but create an ecosystem where the printer becomes more useful. Obvious technologies to integrate would be 3D scanning and digital object transmission. Both of those things are already here, but it is the integration of them that is innovative and has value. Both of those things need robust software (let’s call it an “information ecosystem to live in”) to reach the level of maturity needed for integration. 3D scanning software must align, clean-up, analyze and transform the data to make use of it. A digital object distribution system needs to take all kinds of situations into account and allow for economic activity to spring from it’s use. People in industry know these things, which is why we have Geomagic and Shapeways. Is it possible that one might come up with a fantastically brilliant way of doing things better while they’re tinkering with their printer? Sure! Kids who are now using one at 10 years old are going to have the perfect mindset to do exactly that when they’re 20. But it will be because they connect the existing ideas with something else, where no one saw the connection before and didn’t see how the new relationship would be beneficial. When it comes to the invention portion though, a lot of that requires some hard-core science and observation of things that haven’t already been observed for hundreds of thousands of man-hours by really smart and creative people who were working to come up with a solution because their livelihood depended on it.

If it is true that we will see the massive innovation that has been predicted within the next few years, we first need to see the inventions that completely up-end the way we think of 3D-printing. A completely new deposition method. A new way of supporting the build material that doesn’t have to be cleaned off and doesn’t harm the surface finish. We could see nearly microscopic building blocks (voxels) that link together and have different properties produced in mass quantities as proposed by Hod Lipson at Cornell. Build material might be supported by electromagnetic fields, a stream of particles, tiny robotic arms that either directly support or place temporary fixtures, or any other exotic method that will seem like science fiction right up to the moment when it becomes science fact. Since we have not seen these advancement move past the “what if?” stage, we are going to have to be patient, but more importantly, work extremely hard, to make it happen.

I understand why the change seems to be happening at a frenetic pace. There has been an explosion of media coverage, and those who have been quietly working on their own advances have suddenly been given a good reason to make a lot of noise about what they are doing and the potential it has. By the time their message filters through media, their perhaps optimistic claims can be blown wildly out of proportion. If you just heard about 3D-printing last week, and this week you read a story about printing of living cells, it is easy to get the impression that the progress had been made in one week, when in actuality the printer had been around for 20 years and the doctor had been working on that application for 10 years. Boiled down into newspaper article, then hastily reviewed in a short blog post, then summarized in a tweet, there really isn’t much reference point to give the reader. It’s just disembodied “Wow” that drifts completely out of context, and certainly does not contribute to actual advancement except where, by chance, excited investors throw so much money at it that some of it lands it the right spot.