Super hard nozzles - Test pilots

Discussions concerning the Ultimaker 2 / Ultimaker 2+ series of printers, including the Ultimaker 2 Go
Anders Olsson
Reactions:
Posts: 415
Joined: February 8th, 2016, 8:37 pm
Location: Uppsala, Sweden
3D Printer(s): UM2 UM2E, UM2Go, UM3, Delta Tower, Form 1+, Form 2
Contact:

Re: Super hard nozzles - Test pilots

Post by Anders Olsson »

So, I should write some more details about ruby/sapphire nozzles:

Background
I originally made these for printing my boron carbide plastic for neutron shielding, but soon found out that they are useful for abrasive materials in general and that they tend to print a bit better than other nozzles. So it became a project on it's own.

Development
We have gone through several designs of fixing the jewel and several suppliers. It proves to be a challenge, but we are getting there I think.
Nozzles has been sent out for testing, but most of them were 0.5 mm sapphires, looking like this one:
2015-09-09-6236.jpg
For various reasons that design was not optimal, so I focused on the 0.4 mm version seen in the first post.

Performance of current batch of prototypes
The prototypes I have sent or will send are made at my jewelers lathe, which means the precision is not what it should be.
My intention is to further test this concept of assembly and the performance in general, which is normally best done by the users.
It means you get a rather expensive nozzle before everyone else and for free in return for some feedback :)
The current nozzles might for various reasons have a bit lower performance than the final product, like lower maximum printing speed, but they are all tested by me before delivery and should perform as expected under normal conditions.

How to handle a ruby/sapphire nozzle
The prototypes are a bit fragile and must not be overtightened when mounted as they might break.
I recommend my torque wrench: https://www.youmagine.com/designs/nozzle-torque-wrench when tightening the nozzle.
EDIT: To avoid confusion I erased some text here.
Don't tighten the ruby nozzle more than necessary. The torque wrench will give about 0.25-0.5 Nm which is in my experience always enough.
Ultimaker recommends 1 Nm, which I think is a bit more than needed. Going much higher may destroy the nozzle.

Sapphire/Ruby is fragile, which means that you should be careful not dropping the nozzle on a hard surface.
I also recommend that you try to avoid using rigid metal objects when cleaning a clogged nozzle.
Use a flexible wire from a wire brush rather than a surgical needle and be careful not to bend it!

Motivation and future
So my motivation to try to take this thing beyond the lab is to encourage filament manufacturers to make more interesting filaments filled with various functional materials (like highly electrically conductive, x-ray shielding and such). If the manufacturers has a solid solution to prevent nozzle wear it should be more likely that they produce that kind of stuff.
This is also why I posted information about this early on in the UM-forum: To prevent anyone from locking the technology to by patents and such.
The improved printing quality is also interesting, Neotko has most experience in that area and can probably tell more.

Current design is not optimal for certain reasons as I mentioned, so if everything goes to plan the a final product will look a bit different.
Those details I will keep to myself though :)
Last edited by Anders Olsson on May 16th, 2016, 4:48 pm, edited 1 time in total.
User avatar
ivan.akapulko
Reactions:
Posts: 389
Joined: February 11th, 2016, 11:27 am
Location: Ust-Kamenogorsk, Kazakhstan
3D Printer(s): Ultimaker 2
Contact:

Re: Super hard nozzles - Test pilots

Post by ivan.akapulko »

What tool to make the holes in the stones? Drill with a diamond coating of such small diameter, laser or something else?
Anders Olsson
Reactions:
Posts: 415
Joined: February 8th, 2016, 8:37 pm
Location: Uppsala, Sweden
3D Printer(s): UM2 UM2E, UM2Go, UM3, Delta Tower, Form 1+, Form 2
Contact:

Re: Super hard nozzles - Test pilots

Post by Anders Olsson »

To be able to afford these for the research project, I started off with jewels intended for water jet cutting.
They happen to come with suitable hole sizes, but then needs to be grinded into a volcano shape to be suitable for 3D-printing.

If these nozzles make it to some kind of commercial product however, I will custom order something with more optimized dimensions.
User avatar
ivan.akapulko
Reactions:
Posts: 389
Joined: February 11th, 2016, 11:27 am
Location: Ust-Kamenogorsk, Kazakhstan
3D Printer(s): Ultimaker 2
Contact:

Re: Super hard nozzles - Test pilots

Post by ivan.akapulko »

Anders, what method do you use for the diagnostics of the channel of the nozzle? X-ray, microscope, or some other method of inspection? I agreed on Monday with the Department of metallurgy at the local University, they promise to help with the analysis. Given the fact that not all the nozzles came out perfectly (at yesterday's photo you will notice the offset of the outlet relative to the Central axis) I think, possible artifacts of the surface of the output channel.
User avatar
LePaul
Reactions:
Posts: 3966
Joined: February 7th, 2016, 10:26 pm
Location: Bangor, Maine USA
3D Printer(s): 24 - Yes I have a problem!
Contact:

Re: Super hard nozzles - Test pilots

Post by LePaul »

Roboduck wrote:I've been using E3D hardened steel nozzles with XT-CF20 for a while now and while they hold up much better than the brass they still see a lot of wear on the tip. Mostly the tip edges get rounded which can cause bad layer adhesion and filament curling up onto the nozzle.

I'm excited to see for myself how well the ruby nozzles hold up and how they print. They look really promising!
Can you share the settings you used for good prints? I just bought some of this material for my UMO with heated bed. If you want, there's a thread on this topic here viewtopic.php?f=14&t=14&start=15
Anders Olsson
Reactions:
Posts: 415
Joined: February 8th, 2016, 8:37 pm
Location: Uppsala, Sweden
3D Printer(s): UM2 UM2E, UM2Go, UM3, Delta Tower, Form 1+, Form 2
Contact:

Re: Super hard nozzles - Test pilots

Post by Anders Olsson »

ivan.akapulko wrote:Anders, what method do you use for the diagnostics of the channel of the nozzle? X-ray, microscope, or some other method of inspection? I agreed on Monday with the Department of metallurgy at the local University, they promise to help with the analysis. Given the fact that not all the nozzles came out perfectly (at yesterday's photo you will notice the offset of the outlet relative to the Central axis) I think, possible artifacts of the surface of the output channel.
I have only used microscope for the ruby nozzles.
For brass nozzles we have cut a few nozzle in half in a milling machine to inspect the internal shape, like this:
2015-11-25-6426.jpg
The internal shape of the nozzle in the picture is not exactly according to specifications by the way.
User avatar
Izzy
Reactions:
Posts: 604
Joined: February 12th, 2016, 2:29 am
Location: England
3D Printer(s): Ultimaker 2

Re: Super hard nozzles - Test pilots

Post by Izzy »

Wow, sounds like the ideal nozzle for Carbon fibre in the future, will keep my eye on this for future testing, I'm doing my Open University course at the moment, so have little spare time to tinker, then I want to do some better testing of the Kai Pathy lay filaments, before testing a few of the carbon filaments on the market
User avatar
Dim3nsioneer
Reactions:
Posts: 221
Joined: February 19th, 2016, 2:13 pm
Location: Zurich Area, Switzerland
3D Printer(s): Ultimaker Original, 2, 2+, 2 Ext+, 3, S5, Form 2
Contact:

Re: Super hard nozzles - Test pilots

Post by Dim3nsioneer »

Anders, do you still have test nozzles left? I happen to have a spool of xt cf20 here which I need a reasonable nozzle for. And it would be interesting to work with one before I have them in my 3dSolex flagship store... ;-)
User avatar
Izzy
Reactions:
Posts: 604
Joined: February 12th, 2016, 2:29 am
Location: England
3D Printer(s): Ultimaker 2

Re: Super hard nozzles - Test pilots

Post by Izzy »

Looks a beautiful nozzle, unfortunately I haven't had a need to use carbon fibre filaments other than an initial test of ColorFabb's XT-CF20. I do want to test some other manufactures filaments with carbon etc but won't have time for a couple of months until I finish my OU assignments. If your still looking at that time I'll put my hand up and join in, in the mean time I'll watch your results with interest.

I though I had seen and replyed to this topic before but couldn't find my reply on any of the pages. I replied again and then my previous post pops up and appears,
Last edited by Izzy on February 20th, 2016, 11:55 am, edited 1 time in total.
Anders Olsson
Reactions:
Posts: 415
Joined: February 8th, 2016, 8:37 pm
Location: Uppsala, Sweden
3D Printer(s): UM2 UM2E, UM2Go, UM3, Delta Tower, Form 1+, Form 2
Contact:

Re: Super hard nozzles - Test pilots

Post by Anders Olsson »

I will see what I can do :)
User avatar
Neotko
Reactions:
Posts: 1142
Joined: February 7th, 2016, 7:02 pm
Location: Madrid
3D Printer(s): UMO+ x2.5
Contact:

Re: Super hard nozzles - Test pilots

Post by Neotko »

On my months of tests I did observe the following thinks with the olsson-ruby-nozzle

- Overhangs. Amazing difference from e3d to ruby. Probably because the tip area had (on the model I did test) the tip with less 'flat' area around the hole.
- Temp on the very tip of the ruby was a bit coller with my cheap-thermal camera (probably just because a transparent material and infrared don't work well, so most probably a bad reading). Also, interesting, Ruby has a good thermal conductivity (google).
- Top layers quality did seem worse, but after many many many many tests (anders knows this with photos) at first I though that the less flat area was the problem, but the more and more tests that I did it started to show that actually the think it's that the ruby nozzle it's way ahead on precision. So, if you have the most minuscule bad think aligned, feeder error, bad esteps or changes in preassure of bowden, then the flat layers start to show 'cracks' very minuscule (but I'm obsessed with perfect flat layers for my items/shop) so a Inmediate solution it's to lower the speed, but that's not the sollution.

About this stuff about flat layers...
E3D nozzles, flat area works like a 'blur' effect, just like if you apply a soften effect on photoshop, it hides the problems. With the ruby nozzle you see everything, changes in flow, bad filament, or any problem you have. So at first with my maaaany tests I though that the ruby was the problem but no, I found out, after 3 months (specially when I did some digging into the real esteps of the ultimaker and the 'phantom' 10% (https://ultimaker.com/en/community/1933 ... missing-10) I did found out that my esteps where not 'exact', and that there's no real way to get an exact Esteps on a bowden, except when you stick to a temp/amount of extruded plastic) The distance of the bowden.. blaa bla bla (read that link for more info on this).. Soooo.. After doing all the stuff I talk about on that link I found out that my top layers where just perfect (if I adjusted the temp to keep the same pressure/speed that with e3d).

So if you have read to this point you must think 'I WANT PHOTOS' ok... This are photos of the many I send to anders by msgs.

This one it's e3d LEFT ruby RIGHT (same gcode) this was soo long ago and I was also on the process of doing a better dual fan for my MHC.
Image

Other interesting image. RIGHT RUBY, LEFT E3D (same gcode) This one shows how the ruby removes the vibration on the print by a quite noticeable % (the vibration from the belt/motor/nozzle shake) (remember this it's umo+ hotend modded)
Image

One of the imgs showing the amount of tests done:
https://www.dropbox.com/s/uuypnmngzid88 ... 6.JPG?dl=0
I wasn't able to post this on ultimaker at the time because I was designing the brooches for my shop and the shop wasn't open yet

This it's also a interesting image, this one was taken 4 months ago or something like that. It shows the little 'cracks' that Ruby shows vs E3D, but the source of this 'underextrusion' isn't because the ruby, but the ruby actually shows that the extrusion isn't correct. This was fixed later by doing a real esteps correction. But at first I though that the culprit was the ruby, indeed it wasn't.
https://www.dropbox.com/s/0nr5jpha87kzl ... 7.JPG?dl=0

This is also another interesting test I did OCT 19 2015. It also now it's fixed by doing a propper estep calibration. E3D hidden better the problems that's for sure. But the problem isn't the ruby. Here it's a fragment from the chat with Anders on that OCT 19:
Some img of some prints. The % number sometimes it's for infill or outlines so that part can be confusing. But you can see what I mean on the photos. Each printed has written an E for e3d or R for rby
https://www.dropbox.com/...28zz1s/imgs.zip?dl=0
A img from the nozzle tip (again it must be because the infrared and ruby doesn't work well and I don't have other temp sensor to do a test)
Image

So, all that imgs are very old. This was taken a few weeks ago, after finding the magic missing 10% on the esteps. Ruby does a perfect top layer:
Top layer 1, And top layer 2 closing
Image

Top layer 3 (0.2 size layers)
Image

Top layer 3 and top layer 4 covering)
Image

So... yeah, ruby nozzle kickass. Btw my top layers are printed at 60-70mm/s 0.2layer 210C 1.75mm filament.
User avatar
drayson
Reactions:
Posts: 254
Joined: February 15th, 2016, 5:04 am
Location: Graz, Austria
3D Printer(s): UMO with mods :-)

Re: Super hard nozzles - Test pilots

Post by drayson »

@ Anders. sorry that you had so less response i the UM forum.
I saw the post, but was not aware that it fit´s also the UMO as nearly 90% of the posts there are just about the UM2 (but this is just my impression...)

I would also be interested but I have no filament abrasive enough to test. So I will wait until you will start to sell them ;-)
User avatar
drayson
Reactions:
Posts: 254
Joined: February 15th, 2016, 5:04 am
Location: Graz, Austria
3D Printer(s): UMO with mods :-)

Re: Super hard nozzles - Test pilots

Post by drayson »

@ Neotko, where have you found the 10% in ESteps?
User avatar
Neotko
Reactions:
Posts: 1142
Joined: February 7th, 2016, 7:02 pm
Location: Madrid
3D Printer(s): UMO+ x2.5
Contact:

Re: Super hard nozzles - Test pilots

Post by Neotko »

@drayson read the link posted, it's all there. It was a very intense week of tests :D and reading info, but Ultimaker never answered the post.

https://ultimaker.com/en/community/1933 ... missing-10

My finding report that 10% for 2.85 filament seems to work, but for 1.75 I had to use a 5.5% increase from the actual math esteps.
User avatar
LePaul
Reactions:
Posts: 3966
Joined: February 7th, 2016, 10:26 pm
Location: Bangor, Maine USA
3D Printer(s): 24 - Yes I have a problem!
Contact:

Re: Super hard nozzles - Test pilots

Post by LePaul »

That's a very nice write up

I'd be really curious how a hardened nozzle versus the ruby perform on something like the ColorFabb Cabon Fiber filament I just received.
Post Reply

Return to “Ultimaker 2 / Ultimaker 2+ / Extended & Go”