Prusa Research has done something few, if any, 3D printer manufacturers have done: develop their own in-house produced filament.
The Czech company has exploded in size over the past two years based on the success of their increasingly powerful open source i3 desktop 3D printing platform. Today they sell thousands of devices each month, sufficient to fund their massive staff of 320+.
And also apparently the development of a sophisticated filament line.
To be clear, many 3D printer manufacturers market filament along with their 3D printers. However, they don’t normally make it themselves. Typically it is made by other companies specializing in 3D printer filament and relabelled for the printer company. When the filament manufacturer is a well-known brand, like colorFabb, for example, it may be sold under its own brand.
This makes sense for most 3D printer manufacturers, who are (or should be) experts at building 3D printers, but not necessarily in producing high-quality 3D printer filament. The production of 3D printer filament is actually quite a tricky business, requiring considerable instrumentation and monitoring to effect proper quality control. This is why most desktop 3D printer filament recycling systems are unsuccessful: you need a lot of equipment to make proper filament.
But why do you need high-quality filament? Isn’t one plastic the same as another? There are multiple issues possible:
Filament chemistry dictates the precise glass transition temperature of the material. If the molecular composition changes, then the extrusion temperature may also need to change. Except that current 3D print slicing programs assume consistent material and therefore consistent extrusion temperatures. Inconsistent material chemistry means inconsistent prints.
Filament is made by melting raw thermoplastic pellets and extruding them through a fine nozzle. But coloration is achieved by mixing in a “masterbatch” of precision color dyes during the melt stage. If the masterbatch is varies (by choosing one from an alternative source, for example), the color changes subtly. If the amount of masterbatch coloration varies during filament production, the colors vary again. This effect is often seen in spools from different production runs and can be quite frustrating if both are used in one print.
Filament diameter is a constant issue with many 3D printer operators. The filament in theory should be perfectly round and of consistent dimensions. Again, the slicing program assumes a constant diameter and thus instructs the extruder motor to run at a consistent rate to achieve a specified volume of material extruding through the nozzle. If the diameter of the filament varies, there could be under-extrusion (gaps in the print) or over-extrusion (blobs). In some cases the diameter might vary so much that the extruder loses touch with it and the print stops.
Most 3D printer filament manufacturers declare a maximum variance of +-0.05mm. While this sounds decently small, Josef Prusa points out that this can result in a significant problem:
“When we started manufacturing Multi Material Upgrade v1, we found out that all brands, including the very premium ones, have trouble achieving a consistent diameter and preventing lumps. The MMU v1 featured precisely ground stainless steel tubes with inside diameter of 1.85mm. That should be fine as virtually all filament manufacturers claim they have the manufacturing precision of ±0.050 mm or better on their 1.75mm filaments. That should result in a maximum diameter of 1.80 mm. As you might already guess from these few paragraphs, it is usually bullshit.
Some manufacturers care so little about the diameter consistency, that they just check the string with calipers after setting the extruder and puller speed. Optimally, you want an online laser measurement system with two axes (to measure ovality) from a reputable manufacturer. Then, you just need to measure as often as possible to catch the lumps which are only a few millimeters long.”
Some filament production lines do include diameter checks. ColorFabb, for example, uses dual axial laser measurements as Prusa suggests, but certainly not all do. And this results in a great deal of bad filament on the market. Prusa calculates that the standard quality standard of +-0.05mm actually results in variances in extrusion volume of up to 11.4%!
Unlike most 3D printer manufacturers, Prusa Research prints many of the components for their production units using their own equipment.
This means they use an awful lot of filament, perhaps as much as 100kg per day. This gives them tremendous insight into filament quality issues, and their observations have led them to develop their own line of filament specifically to overcome the issues above.
I’m entirely impressed with the efforts made by Prusa Research on their filament production, which include features not often found – or never found – on competing filament production lines:
Dual axis laser diameter measurements continuously through production of each spool, with 4700 measurements per second. If a spool varies more than 0.02mm from the ideal diameter, it is rejected. In fact, Prusa Research will actually include a graph of the filament diameter in each package!
High-quality consistent raw pellets obtained from Natureworks, specifically their 4034D polymer, which is made in high volume for various industries.
Consistent masterbatch coloration injected into the production process by instrumented deposition. Many filament production lines do this step manually.
Robotic spooling produces tightly wound, highly consistent spools with no tangles. We’ve seen this from only one other manufacturer.
Prusa Research’s production is so highly automated and instrumented that they actually label each package with a QR code that leads to a web page – FOR EACH SPOOL – that shows the full production information!
Thus Prusa Research has launched a very high-quality line of 3D printer filaments. At launch they plan on selling only PLA material, currently listed in five different colors (Pink, Black, Silver, Blue and Red), but they have intentions of soon offering PET and ASA as well.
Here’s the best part: their costs per spool are actually quite respectable: only US$25 per kg. They’re soon to offer the materials via Amazon fulfillment, meaning you won’t even have to pay high shipping costs from Czechia.
This announcement ups the game for other filament providers, who now must reconsider their marketing materials and quality processes.
Via Prusa Research
New research could result in the availability of UV-resistant biopolymers.