Assessing Nozzle Wear In 3D-Printers

How worn are your nozzles? It’s a legitimate question, so [Stefan] set out to find out just how bad 3D-printer nozzle wear can get. The answer, as always, is “It depends,” but exploring the issue turns out to be an interesting trip.

Reasoning that the best place to start is knowing what nozzle wear looks like, [Stefan] began by printing a series of Benchies with brand-new brass nozzles of increasing diameter, to simulate wear. He found that stringing artifacts, interlayer holes, and softening of overhanging edges and details all worsened with increasing nozzle size. Armed with this information, [Stefan] began a torture test of some cheap nozzles with both carbon-fiber filament and a glow-in-the-dark filament, both of which have been reported as nozzle eaters. [Stefan] found that to be the case for at least the carbon-fiber filament, which wore the nozzle to a nub after extruding only 360 grams of material.

Finally, [Stefan] did some destructive testing by cutting used nozzles in half on the mill and looking at them in cross-section. The wear on the nozzle used for carbon-fiber is dramatic, as is the difference between brand-new cheap nozzles and the high-quality parts. Check out the video below and please sound off in the comments if you know how that peculiar spiral profile was machined into the cheap nozzles.

Hats off to [Stefan] for taking the time to explore nozzle wear and sharing his results. He certainly has an eye for analysis; we’ve covered his technique for breaking down 3D-printing costs in [Donald Papp]’s  “Life on Contract” series.

15 thoughts on “Assessing Nozzle Wear In 3D-Printers

  1. I immediately identified the sprial profile in the cheap nozzle! My old Makerbot Cupcake prints that shape when making cylinders – its caused by the threaded rods on which the nozzle rides being non-straight. My guess in this case is a cheap chinese CNC or lathe having non-straight threaded rods or even hardened rods.

  2. Not sure if this is what is happening here but in the process of boring out a block of steel in the past I’ve seen similar screw patterns from using excessive force and inadequate lubricant being applied.

    1. That was my thought exactly. I could notice this pattern when drilling holes with a hand held drill and not when using a proper drill press with adequate advancement speed. It can also be what @Jim McCracken says, the CNC machine is wobbling.

      1. VERY unlikely. I doubt they’d be making these nozzles in a mill. Makes no sense to do so and the nozzle is VERY thin to be using an end mill that long in a plunge cut that deep. Easier to start with brass hex stock in a auto-feed lathe with a revolver/turret tool changer and crank them out ..
        Face, center drill, nozzle size drill, throat size drill, shape and size threaded portion, cut thread, shape and cut free the front of nozzle in a single operation (Use the centre drill as a mandrel/stake to stop the part flying away or have a powered tailstock grab the part to stop it prematurely departing). Since this is china they might also skip the drilling steps, then hand load the undrilled blanks in a second lathe (or even just a column drill with a jig) for the drilling steps

        The spiral pattern in these nozzles is the result of using a badly ground twist drill with slightly asymmetric cutting edges at the tip with high feedrate. The assymetric nose grind causes the nose of the drill to go off-center and then it just bounces around in that spiral as the hole progresses.

        1. +1 I was about to point out the same likely combination of off center drill ground and high feed rate. As a note, grinding off center can be used intentionally to get a drill to bore a hole larger than its nominal size. It is possible that the nominal bore size is one that the machinist did not have at hand and offset the center grind to get it to pass tolerances. I would suggest getting some more from the same manufacturer and see if they are still make that way. My guess is that you will find a new batch to be different. If the off center drill ground was unintentional than I would expect to see that batches would come out with slight differences as the machinist might well be hand sharpening the drills without any jigs or specialty drill grinding tools.

  3. The nozzle geometry is going to have an impact on back pressure, which impacts how uniform the flow is.

    Generally high pressure is caused by narrow nozzle diameters, longer throats, sharp changes in angle prior to the throat and poor surface finish.

    The chamfer (or lack of) at the exit from the nozzle will impact post die swelling and ‘die shear lip stick’ which will impact your surface finish, both of these phenomena are also impacted by the nozzle back pressure.

    In the past when experimenting with paste extrusion I found that too much pressure was bad for consistency, but too little pressure could also give really poor results (we were running between 15-100bar, i have no idea what a modern hot end would typically work at). For a given nozzle geometry and extrusion rate the properties of the input material could make the difference between garbage and good quality extrusion, so I suspect that “best settings” will vary pretty dramatically for different materials; a lower viscosity material (either changing the material, moisture content or the temperature) will react differently to changes in nozzle geometry than a high viscosity material.

    It would be interesting to see the effect of running a worn nozzle and a good nozzle at +/- 5C either side of normal running temperatures with the same filament and comparing the finished prints.

  4. The spiral is probably just the bit flexing as it cuts due to being unevenly sharpened, or just unevenly dulled after popping out thousands of cheap nozzles. A machine shop going for quality will swap out the tools before the results start getting really horrid, but a lot of those cheaper manufacturers don’t really care about precision and will run the stuff until it breaks.

  5. As for the odd internal hole shape, I’m guessing it’s “because they’re cheap”. Not in-and-of-itself a cause, but….

    Brass is used because it’s cheap and easy to form. Easy to form does not mean “cutting”, it usually mean “shmooshing”…aka “cold-forming”. My guess is that the outside nozzle profile (on the cheap ones) is actually formed on a multi-ton (100-ton?) press. The blank is drilled to the size, and then pressed to achieve the conical tip. That force compresses the entire block of material, resulting in the odd wavy walls you’re seeing.

    Well, that’s my thinking. What’s funny is that the E3 nozzle, with those long straight lines down the bore is also indicative of a punching/press forming operation.

    Not saying definitively that’s what their processes are, but cheap brass is more often formed molten of through cold-forming than with a carbide or other cutting tool.


  6. Offhand, I’d say that the boat tests with larger nozzles is not a useful test. Say What? I say this because throughout the rest of the video, you come to the conclusion that absolute nozzle diameter isn’t actually the way that the wear presents itself, until the nozzle is well-and-good on the verge of useless. To be useful, the analysis would need to be able to find the tunable region of active in-use wear, with experiments showing correlation in that range. Instead, by assuming that nozzle diameter was a proxy for wear, the results don’t necessarily tally: excessive stringing in a non-changing nozzle has to do more with retraction settings precisely overcoming the gravity-fed plastic leaking out, held partially in by surface tension. Using the same GCode on progressively larger nozzles only proves that stringing is related to the surface tension being less effective as the nozzle diameter grows. It is not, however, causatively linked to wear. So while the stringing effect is true, it doesn’t appear to be genuinely representative of the wear modes being investigated. No?

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.