July 26 4.30pm CEST. 'I'm a Material Engineer at Ultimaker for 3 years. AMA'.

[SandervG 1,521]

What's your favourite Cura setting? Why?

I know it is @TomHe's AMA, but what is yours?

[Tomhe 21]

@Tomhe, another question for you: How difficult was it to get PVA to work reliably, and what was the most difficult part of it? Speaking of difficult parts, what do you consider the most difficult part about your job?

PVA:

I spend over half a year basically only on PVA, and some of the work was already done before I arrived. It was hard to get it to work because there were so many variables:

- Software (firmware and cura) were unstable and missing essential features at that time

- The PVA grades improved over time, it was hell to work with them at the start

- Nozzle geometry wasn’t perfect yet, I started with 0.8mm nozzles so it was easier to remove the clogs all the time

I actually started this project on a modified UM2, with the old feeder that couldn’t handle PVA. Most difficult for me was thinking of how to change the Cura support strategy to make PVA prints without scarring.

Job:

Deciding/finding out what people want from our materials and profiles. People ask for quality, but quality can be a lot of things. Balancing visual quality, strength and print time is hard. And often companies will send models that they want to have printed with high strength (like production line fixtures), only to judge them purely on aesthetics when finally have the parts.

[nallath 1,120]

I believe the first material researcher at Ultimaker was @Nallath's dad, but I don't know exactly when he started working here.

I think that was about 4 years ago? I'm not quite sure.

[Tomhe 21]

What's your favourite Cura setting? Why?

My favorites (I am cheating a bit on this one) are the initial and final printing temperature settings, as those settings made it possible to print dual extrusion PLA without any need of a prime tower.

[SandervG 1,521]

@Tomhe; Which material would you personally prefer to see coming up next, to be used in 3D printing?

[Nicolinux 288]

Here is another question:

When you test new filament brands, do you have a favorite test model? Which one and why?

[Tomhe 21]

And I had a question from myself too

How far away are we from printing with conductive materials, what needs to happen for this to become widely available?

It’s already possible to print with conductive materials, but the possibilities are very limited. The resistance of Carbon-filled (Graphene, Carbon Nanotubes, Carbon Black) materials is still pretty high. It’s just enough to turn on some LEDs with a big battery. There are some issues on the process/printer side as well: Carbon-filled materials eat nozzles for breakfast, and Carbon nanotubes are about as hard as diamond… I think printing conductive materials for simple leads is already possible, but functionally prototyping devices will take some years of material, software and printer development.

[IRobertI 520]

I know that you were experimenting with different PVA blends long before the UM3 was announced. But I'm assuming that wasn't the only type of support material you tried. Are you still experimenting with new types of support materials or have you settled in on PVA and improving that as much as possible? And if you've rejected other types of materials, what has been the primary reason(s)? Nasty solvents?

[Tomhe 21]

Here is another question:

When you test new filament brands, do you have a favorite test model? Which one and why?

I often use the ‘OK Hand’ model and the 3D Benchy to test my settings, but I start out with simple shapes like cubes, balls, overhangs, etc. I like the OK Hand the most since it can also be used to test layer bonding (by cruelly breaking the fingers). I will use the Benchy for dual because it's a terrible model to print.

[Tomhe 21]

@Tomhe; Which material would you personally prefer to see coming up next, to be used in 3D printing?

I’m really interested in the metal-polymer materials, from which the polymer can be burned out to form a pure metal object. Unfortunately sintering the part is harder and more expensive to do than the printing… Using an Ultimaker to print metal jewelry would be awesome! Maybe it could be possible to use this technique for ceramics/glasses for even more possibilities.

[SandervG 1,521]

@Tomhe; Which material would you personally prefer to see coming up next, to be used in 3D printing?

I’m really interested in the metal-polymer materials, from which the polymer can be burned out to form a pure metal object. Unfortunately sintering the part is harder and more expensive to do than the printing… Using an Ultimaker to print metal jewelry would be awesome! Maybe it could be possible to use this technique for ceramics/glasses for even more possibilities.

Interesting! After burning out the polymer, are the remains (the metal parts) strong enough to form a model or does it turn out extremely brittle?

[Dim3nsioneer 557]

In addition to the test model you use for testing a new material, can you describe shortly how a new Cura material profile is produced? What kind of test and results are needed before UM can distribute the profile? If this exceeds this AMA, it might also be a nice topic for a blog...

[Tomhe 21]

I know that you were experimenting with different PVA blends long before the UM3 was announced. But I'm assuming that wasn't the only type of support material you tried. Are you still experimenting with new types of support materials or have you settled in on PVA and improving that as much as possible? And if you've rejected other types of materials, what has been the primary reason(s)? Nasty solvents?

We've played with PVAs, BVOHs (and blends of the two), HIPS and some more materials. We went for PVA because of the dissolvation speed and because it fully dissolves in water. Other materials, disintegrate into a milky mess (BVOH), need nasty chemicals (HIPS) or are unreliable to print. The nasty chemicals for HIPS also seemed to have effect on ABS, which we didn't like.

We are continuously looking into more support materials, as we strive to be able to have support solution for every build material.

[SandervG 1,521]

In addition to the test model you use for testing a new material, can you describe shortly how a new Cura material profile is produced? What kind of test and results are needed before UM can distribute the profile? If this exceeds this AMA, it might also be a nice topic for a blog...

Good one, and it links to another question I saw earlier;

Are the print profiles in Cura aimed to make prints that look nice, strong prints, or a middle ground of everything? What goal do you have in mind when making a profile?

Edited July 28, 2017 by Guest

[Tomhe 21]

In addition to the test model you use for testing a new material, can you describe shortly how a new Cura material profile is produced? What kind of test and results are needed before UM can distribute the profile? If this exceeds this AMA, it might also be a nice topic for a blog...

It could indeed be a blog post, but I'll try to give a shorter answer here:

First we start by finding the processing temperature by printing and measuring (caliper and scale) the extruded part. Then we try and find fan settings at which we have strong layer bonding, and a bed temperature at which we don't have warping.

With these three settings (printing temperature, regular fan speed and bed temperature) done, we can start looking at speeds. You want to print as fast as possible, but also want the overhangs to be printed nicely. So the outer wall must be  printed slowly, the infill faster. We use a lot of models to tune these settings, and we try to even out the flow with the inner wall lines.

And then there is the tuning for the z-seam, for top details, for top layer to walls connection, for the bridging, etc etc...

All the profiles for new materials are tested in the company by our testing team, and often tested outside the company by beta testers. Their feedback decides if a profile/material can be launched.

Edited July 26, 2017 by Guest

[Tomhe 21]

In addition to the test model you use for testing a new material, can you describe shortly how a new Cura material profile is produced? What kind of test and results are needed before UM can distribute the profile? If this exceeds this AMA, it might also be a nice topic for a blog...

Good one, and it links to another question I saw earlier;

Are the print profiles in Cura aimed to make prints that look nice, strong prints, or a middle ground of everything? What do you have in goal when making a profile?

The profiles are now made to be somewhat in the middle ground. PLA profiles have one less wall than ABS profiles so it’s slightly material dependent. I went for strong and nice at the same time, now we are trying (and succeeding!) to speed things up without losing these qualities.

[ghostkeeper 105]

What feature are you (in your opinion) in need of most in Cura?

[IRobertI 520]

Could you tell us more about how you determine things like layer bonding and the like. Is it purely by feel or do you also send parts off for "proper" ISO standards testing?

[Tomhe 21]

Could you tell us more about how you determine things like layer bonding and the like. Is it purely by feel or do you also send parts off for "proper" ISO standards testing?

Both. During the profiling we do it by feel and simple tensile test at first. We also have had a lot of our tensile bars tested with ISO standards, but there are no real ISO standards on how to print it (as no two printers are the same, and settings have a lot of effect on the strenght). We try to use data from 3D printed samples as much as possible in our technical data sheets.

[Nicolinux 288]

Do you use other slicers to see how they compare to your new Cura settings?

[Tomhe 21]

What feature are you (in your opinion) in need of most in Cura?

I really need Cura to know if I'm going to print single or dual extrusion on the UM3, so I can have specific settings for single extrusion and for dual extrusion. Setting up a single extrusion 0.8mm printcore PLA print at 'Sprint' speed (0.4mm layer height) while you have an 0.4mm printcore with ABS loaded in the other slot is leading to some issues. (ABS build plate temperature is used for example)

[IRobertI 520]

What is your favourite go-to material when you need something that will handle most of what you throw at it. And you can't say PLA

[Tomhe 21]

Do you use other slicers to see how they compare to your new Cura settings?

Personally I've used Slic3r before as I had issues printing PVA with Cura 15. It had some nice support features that we missed at that time. But it crashed at lot at the time, I couldn't slice complicated models with dual extrusion supports.

With S3D I'm not very experienced yet, when I tried it it didn't cool down and heat up the unused nozzle while the other was printing. So it was too slow for me at the time. Lately I haven't had much time to look at other slicers as we are implementing a lot of new features each Cura sprint, that we also need to find good setting values for.

Edited July 26, 2017 by Guest

[Tomhe 21]

What is your favourite go-to material when you need something that will handle most of what you throw at it. And you can't say PLA

Nylon if I need good supports/complicated models, and PP if it's an easier model. I've printed a big PP Ultimaker robot which we haven't been able to break yet (not even with a hammer).

I've also dried Nylon parts at over 90C without any issues, and I love the look of 100% infill Nylon parts printed at lower layer heights (0.1mm). They look injection molded.

Edited July 26, 2017 by Guest

[SandervG 1,521]

This is another question which I have seen a couple of times;

'What would you think of an engineering-profile, which is very bare boned and straight forward regarding printing speeds, temperature, jerk and acceleration? It does whichever command you give it, and decent surface quality is not the most important outcome? (Strong and dimensionally accurate being the goals). ?'

Edited July 26, 2017 by Guest