banner perso

Some links about me. Many of my 3D designs are free. I also post on Google+ and in another blog, oz4.us

Tuesday, July 2, 2013

Extremely fat extrusion with a 1 mm nozzle

Going "nuts" and 3D printing with a 1.0+ mm nozzle!

A ridiculously big 1.1mm nozzle make it flow like hell !
After some testing with a 0.3 mm nozzle, I really had to try and post my quick and dirty tests with a huge one.

In fact, I had lost my smaller drill bit set made for smaller holes... So I took a brass cap nut, screwed it on a threaded rod, put it under my vertical drill press with a "regular" small drill, and here we go. I measured afterwards it to be something like 1.1 mm. Interesting...

Now for sure it was funny as a friend told me I just had made the first Scoubidou machine! But the first prints still were surprisingly good! Not only can I print very fa(s)t, but it still makes a nice print in the end.

Now, there were a few quirks I should have taken time to fix (see below), but very big nozzles are now part of my consideration when I want to prototype quickly and/or when the printed object needs no special accuracy: refining the ergonomics of a handle, or printing a toy and so on.






Testing it in real

Obviously with such a big nozzle, it makes sense to print fat also in the vertical plane, so I used a stunning 0.6 mm layer thickness. It also meant an amazing flow rate. I also asked for one pass walls: after all, they are already 1 mm solid as soon as the plastic is deposited!

With a printing speed of about 100mm/s, I never saw my spool of 3 mm filament turning so quickly... I first thought I would be wasting a ton of filament doing something silly, but then realized that I would be asking for the same quantity of filament in the end as whatever nozzle size printing the same geometry... except that it just prints ridiculously faster!

It just prints unbelievably fast... and fat ! But after the good laugh, it proved not so stupid in the end!

See? Of course you get coarser details than with a 0.3 mm nozzle where 10 layers may stack in each one of these,
but as for rapid prototyping, you just cannot beat it! And the result is far from a joke : robust and watertight.
And yes, this is yet another print of Cushwa's owl !

Now, the horizontal slight slopes are trickier, and I ran in trouble as my settings gave no chance to the printer to fill the required surface, as show near the feet below...

Horizontal planes need better attention, as with 0.6 mm layer height and one single wall,
it will not have a chance to fill gaps like those near the feet. Losing some material here also has
a drastic effect on the whole layer: I should use some filling at least, or a two-pass wall layout may be.

Some forgotten items, and a brief analysis

I had no flow issue at all beyond setting the correct value (I had no way to measure the nozzle diameter precisely).  The Ultimaker heater block handled the flow easily and it was able to melt all the filament without trouble (I even printed slightly below the usual temperature). The prints were watertight except on my obvious mistake with the horizontal planes. I did not try enough before switching back to a normal nozzle but I think I may have been able to fix the bridges (note that it may be a bit trickier because of the weight of such a big extruded thread...)

Also, I realize I should have tried retraction. But since I was printing with an almost viscous thread of plastic, here again I feel confident it can be made without special difficulty.

The other drawback, apart from the horizontal filling, was the cooling. As shown above, the part of the owl opposite to the fan produced wavy patterns because the filament had not enough time to cool down -- I think I reached my bottom limit of about 8 second per layer here, which was probably not enough. A second fan mounted to the opposite side would surely fix the issue.

The feel on the back of the owl is really nice, so it definitely can suit artistic goals also.
This nozzle would print tons of toys such as Lego brick very quickly, even though
I would give it some careful and pragmatic hand-made tuning with the help of some preliminary tests
because the software may probably be doing a few hypotheses not well suited to such a weird setup.


Conclusion

I printed the owl at about 100 mm/s and I was very surprised to see how regular a flow I got, the Ultimaker doing its job very well. With a faster setup to change my hotend or nozzle, I would definitely consider this as a viable solution for printing sketches or quick prototypes. The resulting object is watertight and very solid also, with more than 1 mm thickness of PLA, I do not think I could crush with my bare hands.

While printing hollow objects is attractive, the horizontal parts must be taken care of because it will have hard time to fill them with such a layer thickness.

As shown, I used super thick layers here to check how faster it is. This results in an almost round thread and wavy surfaces. With thinned layers (0.1 - 0.2 mm), the infill would still be crazy fast, but the impact on the outer look would be less noticeable compared to regular nozzles. The remaining impact would be in the loss of precision (e.g. the owl claws get blurred by the super large horizontal diameter of the deposited thread, whatever the squeezing that occurs due to reduced layer height).

Note that I printed only three items before getting back to a more serious business (such as tiny nozzles), but that was definitely worth the try!

I will try again just to look at the process: it reminded me my first days with my printer :)
More seriously, this is something to consider for special needs,
where speed is the key or when only the overall shape is what matters (e.g. ergonomics).

About Me

My photo

If you know me and you cannot tell exactly what my real job is, then you probably found the right Jeremie. Check zax.fr for some pointers.

I am self-employed and I help start-ups, research centers, small companies with their needs related to computers, sensors, data processing and mechatronics. If you have a project and know what "R&D" is, then you already sparked my interest ;)