Since BASF announced its Ultrafuse 316L A couple of months ago, we were eager to get our hands on this filament that promises all-metal objects using most desktop 3D printers. That’s quite a claim, so we snagged a few to try it out.
Affordable Metal Printing
For those who don’t know, BASF Ultrafuse 316L is a bonded metal (stainless steel) filament that requires post-processing of catalytic deagglomeration and sintering to become fully metallic. It is designed to make metal 3D printing accessible to almost everyone. Printed parts are initially in a “green” state and the debonding step removes most of the polymer binder that makes the filament printable, bringing the part to its most brittle “brown” state. The final sintering step removes the remaining binder and fuses the metal particles into a fully dense metal part. BASF is creating a network of downstream companies to handle these steps.
The first thing I notice about this filament is its weight. The product is, like its price, considerable. A reel of his comes in at three kilos and costs $465. It’s a little surprising that BASF doesn’t offer a smaller, more affordable reel, but I don’t pretend to understand their production decisions.
Getting started with Ultrafuse 316L
The product is professionally packaged and looks very much like a giant Moon Pie in its vacuum-sealed foil packaging. It is also very tightly wound with no overlaps or knots in the filament. It arrived a day before the Dimafix bedding adhesive that is recommended to keep 316L parts down, but I was too excited to wait for it. I’m going through this stuff, so why not try printing it on painter’s tape, right? Exactly, I knew you would agree.
First, the nozzle on my LulzBot had to be changed to an extra hardened steel E3D v6 nozzle because 316L filament is 80% steel powder and therefore very abrasive. While it is possible to print this material on a standard brass nozzle, it will quickly wear down the brass and affect the diameter of the nozzle, eventually leading to poor print quality and failed prints. It is also recommended to use a dedicated nozzle for 316L to ensure that no other material gets into your metal prints; foreign materials can cause objects to explode in the debinding and sintering process. If you don’t use a dedicated nozzle, be sure to run a good amount of cleaning filament through the nozzle to completely purge all old material before printing with 316L.
Loading the filament was just like any other filament; it’s a bit softer than ABS but not as soft as TPU, so it should work fine with Bowden extruders. It extrudes without problems at 240°C on my E3D hotend. I was pleasantly surprised that the calibration cube adhered well to the painter’s tape, but my surprise was short-lived as the corners started to peel away about halfway through the print. I canceled the print because I was just testing bed adhesion and print settings. So painter’s tape doesn’t work., lesson learned. The image below shows that the infill is not 100% recommended (voids can cause parts to fail in post-processing), but I was still surprised by the weight of the small object. This image does a good job of showing the level of detail that can be achieved with the material.
The Dimafix arrived the next day so I applied it to my glass bed and printed the calibration again with 100% infill. Prints with this material take a long time due to the solid fill, the thin layers that are recommended to increase the density of the parts, and the slow print speeds that are necessary to achieve smooth walls. Objects should also be scaled 19% on the X and Y axes and 21% on the Z axis due to anisotropic shrinkage that occurs during the post-processing steps. This 20mm cube took almost four hours to print; if I were to print it on plastic (mostly hollow) it would take about 20 minutes. But that’s a small price to pay for metal prints. A tiny calibration cube weighed a whopping 60 grams in its “green” state.
From green pieces to metallic pieces
Packaging everything to send to the post-processing company was a bit stressful because the “green” parts are quite fragile; they feel like a heavy clay and are very flexible. Processors ask that a form be included with each shipment that lists all parts, as well as their weights and dimensions. They also ask that each part be individually wrapped. I found the task less tedious if you pretended to be a museum curator cataloging priceless artifacts for safekeeping.
The most difficult part of the entire process was dealing with the uncertainty of whether the parties would survive both legs of their postal journey. That and the wait, which actually wasn’t that long; I shipped the parts on October 31st and received them on November 25th, but keep in mind that I am in Sacramento and they are in New Jersey. Fortunately, processors inform customers when they receive their parts and email a report of the parts before and after processing.
Opening the package certainly had a ‘Christmas morning’ vibe as my curiosity and excitement were in overdrive. The metal parts are just what I wanted. They are solid, hard and shiny. They even make that unmistakable metal ‘clink’ when put together. Every part came out fine without any noticeable warping. But what about the contraction? How accurate are the scaling recommendations?
Look at that shine! Sorry, let’s get to the important numbers. These are the dimensions of the 20mm test cube after processing: X: 19.82mm, Y: 19.91mm, Z: 19.46mm. That’s a variance of 0.5% to 2.7%, which is pretty good considering a shrinkage factor of around 20%. It follows that the Z axis has the largest variation since the layers can be crushed by gravity during the sintering step. BASF had already adjusted their scaling recommendations once before I worked with the material, so they may make another small change after more parts have been processed.
The details in the prints appear completely unchanged, which is good to know. Here is a Bender head which I will now use as a paperweight; I took it outside to bring out the brightness of it and it is so reflective that it oversaturates the photo.
It is important to note that the thin antenna survived the return trip even though it was not individually wrapped. That is a testament to the strength of this material.. And about that harshness: wow! One of the parts she wanted to try is a carabiner that is printed as two pieces that snap together. But this material is simply too hard for models that are designed to be assembled like plastic. It took two sets of pliers and some serious grunting to bend the metal enough to assemble the carabiner. Bending parts with pliers is unthinkable if they’re printed on plastic, so this really changes things. Here is the before and after:
The layers of material in the contours create a nice wood grain effect not normally seen in metal. To test its strength, I compared its charge limit to a solid PLA version of the same model; the plastic one broke at 105lbs (which is pretty impressive) and the metal one didn’t even bend with my brother’s full weight (125lbs) hanging from it. That’s super scientific, I know. I’m the best I could do with a simple hanging scale and a brother.
I also printed a one piece carabiner that uses a motion pattern as a spring mechanism. The spring does not work on this material as it stays in whatever position it is pressed on; it looks like the metal will eventually break if it is repeatedly bent back and forth. Therefore, I would not recommend this material for parts that are meant to be bent regularly.
The other functional part tested was this multi-tool that includes a bottle opener, which successfully opened a bottle, albeit with some flex at the thinnest point, which is just 1.5mm thick. If the bottle opener was a bit thicker, it probably wouldn’t have bent. Hex sockets work fine.
The functional and strength tests were very informative and encouraging, but the cosmetic tests were just as enjoyable. To see how these pieces polished up, I put the Lantern ring on a rock tumbler with steel screws for a couple of hours. The results are absolutely brilliant.. This thing really shines. Parts of it are even like a mirror. Here’s a before and after:
Going around with screws is just a basic form of polishing, although it’s obviously effective. Tumbling for a longer period or using wheels and buffing compounds would probably produce even better results. This is sure to catch on with jewelers already using 3D printers in their business.
Most of the parts I print have geometries that would be difficult and expensive to CNC machine or mold cast. Machinists are increasingly using 3D printing, so they are probably adding it to their repertoire as well.
my verdict
BASF Ultrafuse 316L is a very exciting material with a lot of potential. It prints relatively easily on most 3D printers and the results are impressive: tough, detailed and shiny, just as stainless steel should be. The tensile strength is not quite that of forged steel, but it is very good nonetheless. It is by far the strongest material I have ever tested on a 3D printer. And the ‘wow factor’ is high with this one; everyone who sees the parts has the same reaction: “Did you print this?”
With this material, metal 3D printing is finally accessible to almost everyone. Small businesses, designers, and artists will benefit from this filament, whether they’re making custom gear, robots, or jewelry.
related story
Hands-on review: Ultimaker S5
related story
Hands-On Review: HI-TEMP Filament
related story
Large-Scale 3D Printers Big Enough To Print Furniture
related story
How to Get Stronger FDM 3D Prints
related story
Leave a Reply