Effect of Nozzle Size on Strength

[rsterne]

I made a 0.5mm nozzle by drilling out a 0.4mm, and then I tested those sizes, plus a 0.6mm, all in hardened steel, printing my Tensile Test Coupons in Kimya CF-PETG.... I printed the coupons using a 0.25mm Layer height, and a 0.75mm Line Width, for all coupons.... Here are the results (I did two coupons in the Vertical Z axis for each nozzle diameter)....



There wasn't a lot of difference, but using the 0.4mm nozzle for printing that size bead was not as good, either in X or Z.... The 0.75mm width is 188% of the 0.4mm nozzle diameter, which is larger than the 150-160% recommended maximum.... There was almost no difference between the 0.5mm and 0.6mm nozzles in either X or Z, which to me means that if you need to print a 3:1 bead that is theoretically perfect for a given nozzle size (eg. a 0.5mm diameter for the 0.25L x 0.75W bead), you are better to print that bead with a slightly larger nozzle than a smaller one.... This also confirms not going larger than 150-160% of the nozzle diameter for the bead width as well....

I am not able to purchase a Diamondback nozzle in 0.5mm, and printing a 0.25mm x 0.75mm bead is a great choice for a lot of the things I make, so this indicates that the 0.6mm nozzle I have coming will probably produce a stronger part than the 0.4mm using that bead size....

Bob

[TorqueMaster]

Excellent test, Bob. 
I've been using a 0.5mm nozzle for most of my prints, mostly at 0.60mm line width x 0.20mm layer height -- quite similar in ratio to your 0.6mm nozzle findings, so good to know!    PETG, not CF-PETG.

I believe it was Prusa that recommends a line width 10% (or more?) larger than nozzle size -- not necessarily for strength but because the flow naturally tends to be a bit wider than the hole.  So your findings are in good company, even if not for the same reasons.

The choice of layer height is tricky -- I believe that you are going with 0.25mm because the CF-PETG adhered better at that thickness?  But if printing fine threads (x1.0 or x1.25, for example) 0.25mm layers are borderline acceptable.   Might be good to use Cura's variable line heights to print any threaded layers thinner.

I do not notice much loss of visual quality with a larger nozzle, but sure do notice the increase in throughput!  I wonder how many folks plod along with 0.40mm nozzles because that's what the printer came with...?

[rsterne]

Some CF filament manufacturers recommend a 0.25mm layer height, to allow enough flow with the filled filament.... I have found that you can use 0.20mm if you use a 0.6mm width because that increases the pressure inside the nozzle.... For detailed CF parts, I use 0.20mm x 0.60mm.... For large parts, or where I don't care too much about surface quality I use 0.30 x 0.90mm, which prints twice as fast.... I find that with CF, you can hardly tell the difference in appearance (from a distance) between 0.20 and 0.25mm layers, so a good choice for that, if you don't have a 0.50mm nozzle, then using a 0.6mm to print 0.25 x 0.75 is a good choice....

Bob

[TorqueMaster]

I looked at my purchased, but not yet used, hardened steel nozzles, and found the same situation -- I have some 0.4 and some 0.6, but no 0.5. 

I think some CF filament recommends at least a 0.5 nozzle-- have you experienced any clogging doing CF with a 0.4?

[rsterne]

It doesn't print well with some CF filaments with the standard 0.4mm line width if you only print 0.2mm high (other filaments are OK)..... but if you print at 0.2mm high x 0.6mm wide it's fine, or you can print at 0.25mm high x 0.4mm-0.6mm wide.... I have actually never had a nozzle clog with any filaments....

I haven't had a lot of luck with CF-Nylon (very fussy, hard to dry), but tons of success with CF-PETG.... It's my standard go-to now for CF filled parts....

Bob

[TorqueMaster]

Ok, very good, thanks.  My thought was the CF fibers, however thick they are, when trying to get through the 0.4 nozzle, doubled up or shoulder to shoulder if you will, might eventually cause a clog.  It wouldn't be so much due to flow rate, just random chance of too many poops clogging up the tiny pipe, if you will.

[subscriber]

Thanks for reporting on the drilled out nozzle, Bob.

[rsterne]

The CF bits are 0.005-0.01mm in diameter (5-10um).... I don't think they can double over (too stiff), and so small I think all they do is make the melted filament more viscous.... They do apparently make the surface of the layers somewhat rougher, according to the chemist at Matter3D in Victoria the 15% CF has better layer to layer bonding than the 10% because of this.... In my tests, it is a bit better, and used with a 0.6mm nozzle with 0.3mm layer and 0.9mm line width, the "Z" strength is over 75% of the "X" strength.... That's pretty impressive, IMO....

A quick estimation, using MS CoPilot AI says there could be about 5800 CF strands of 5um diameter packed into a 0.4mm nozzle.... If the filament was 15% CF (by volume) there would be, on average, about 870 CF strands inside the diameter of the nozzle, and the other 85% would be molten PETG.... I don't think clogging would be significant, but certainly the pressure required to extrude would be higher than for pure PETG....

Bob

[TorqueMaster]

Ahhh, MUCH thinner strands than I had expected.  Did they mention the length of the strands?  Even at 10:1 length:width, they could go through any nozzle sideways, so probably not an issue either.  Maybe the recommendation I saw for larger nozzles was to reduce backpressure back to "normal" values, to give the user a print experience similar to plain PETG.   

[rsterne]

In a Google AI search, it says the length of the strands is 0.05-0.2mm.... However, in other places I have found 50-70um.... So, a 10:1 length:width ratio is probably about right.... Most filament makers, if they say anything at all about orientation of the fibres, claim that the fibres orient themselves parallel to the flow, whether that is true or not I have no idea.... but they are short enough to go through the nozzle sideways anyways....

Bob