Quest for a clear print

[pm_dude 27]

Triggered by the request of TMrevlje http://www.thingiverse.com/thing:175065. Cutting the top I inserted a blue or red filament to see it from outside.

Using Ultimaker 2.

First I started with Colorfabb XT and used the setting Colorfabb recommend:

- 240C/70C

- 0.27mm height

- 0.8 thickness

- 0.8 top/bottom

- 0 infill

- 30mm/s

- Flow at 110%

- No cooling fan

- Cooling time per layer was at 5 sec making it go slower and slower near the end.

- Cool head lift was on but was only triggered on the last layer.

All the following prints will detail only the changes made to the PREVIOUS print settings.

Model was printed standing up so layers are mostly aligned with the tube. Supports done in Meshmixer.

1:

- XT (all above settings)

I found it a bit too white and not clear enough

2:

- Use Ultimaker Natural PLA (clear)

- 225C/60C

Difference between PLA(left) and XT(right)

 

All the following are using Ultimaker Natural PLA

 

 

3:

- Try with the model like shown in Thingiverse (sideways) + supports using Cura

 

Difference between print 2(left) and 3(right)

 

 

4:

- Going back standing but with less supports to remove as much as possible the head jumps.

- Remove cool time per layer

 

The print was too hot with no cool timer and no fan

 

 

5:

- 210C

- Fan 100%

- 20mm/s

 

Difference between print 2 (left) and 5 (right)

 

 

6:

- Removing top/bottom (set to 0.0 instead of 0.8mm). I remove it because it add infill everywhere the shell thickness of 0.8mm per layer doesnt create the required top/bottom thickness. It cause the head to jump to go fill the gap once the shell is complete.

 

Difference between print 5 and 6

 

 

7:

- 190C

 

Too cold

 

Difference between print 6(left) and 7(right)

 

8:

- Back to 225C

Difference between print 6(left) and 8(right)

9:

- Back to 30 mm/s

Too fast and too hot (even with fans)

Difference between print 6(left), 8(center) and 9(right)

 

 

 

All the prints from 1(top left) to 9 (bottom right)

https://drive.google.com/file/d/0Byhp9wWpyMGWcDJRamtiSUVQTTQ/edit?usp=sharing

 

 

So overall my findings are:

- print 6 is the most clear print

- print 8 is second but is more constant than 6

- printing at between 210C and 225C at 20mm/s with fans is good.

- having the cool layer time will bring change in transparency due to print speed changes.

- every single time the nozzle extrude filament you will reduce transparency. If infill gets added you see it. Depending on the print it might be possible to remove top/bottom if not try to keep it lower than shell thickness

 

- I need to test again with other transparent filament. I have 2 translucent (blue and green) from Faberdashery. They are very light so it may give interesting results

- I need to test layer height (smaller and bigger).

 

[tmrevlje 1]

Thanks for this extensive testing of the model with Colorfabb's XT material, just as you said the number 6 and 8 do seem best, I would go for number 6. Will use your settings and try it with the "Transparent Red" I received from Colorfabb.

Will also most certainly invest in a spool or two of the XT, it seems a material to be played with, capable of achieving some really cool results.

Looking forward to your results with the translucent green and blue.

 

[DidierKlein 729]

You can do some cool stuff with XT, it prints more clear when heated up (250°c with no fan) and more opaque when lower 235°c with 100% fan.

You can see the difference here:

 

I think that spiralize helps with transparency also.

 

 

This is a print in Ultimaker Natural PLA:

 

It's more opaque because of the shell thickness

 

 

But put a Led in it:

 

Overall regarding transparency i think that Natural PLA from Ultimaker does a good job

 

[tmrevlje 1]

Well that's encouraging since I have 3 spools of Natural's on the shelf and I won't have to wait for the XT, but this is translucent vs. transparent I'm thinking off to testing, will post results when I'm done.

 

[DidierKlein 729]

I think you have to keep with big layer heights also to have more transparency

 

[tmrevlje 1]

Thinking like 0.1 or even up to 0.2 and higher?

 

[pm_dude 27]

higher. for XT colorfabb recomment 2.7 and up. I'll try another print and test at 0.2 and 0.3 to see the difference.

 

[jonnybischof 60]

You should take a look at this.

1. The material: Taulman T-Glase. Should be a good material. XT is probably very similar to T-Glase, although I haven't directly compared the materials myself.

I love XT, but mainly for it's stability and ease of printing. Prints get somewhat transparent, but I believe XT is more optimized towards strength and printability than optical aspects.

 

2. The information on how to print stuff to get the best "transparent" aesthetics: Print with a big nozzle (if possible), maximum layer height and as few shells as possible. You can see that they get very nice results with their tests.

If you want to add another material to your list: BendLay is another clear filament. Afaik it's a PET-G material, much like XT or T-Glase, but I'm not sure about that.

I've got a roll here, but haven't printed anything with it yet o.O

 

[pm_dude 27]

Ho yeah! I seen this one before. Its true T-Glase could be a good test to run too.

I ran the test between 0.2mm and 0.3mm layer height and the bigger the better for transparency.

 

[eldrick 65]

I had some decent results using transparent ABS and then vapor-bathing it in acetone. It came out mostly clear, almost like glass, with some cloudy parts where the plastic was thickest.

 

[pm_dude 27]

Interesting. Acetone could be the only 'real' solution to light bouncing around in so many thin layers!

I'm scared of that s*** lol

[eldrick 65]

Acetone is pretty mild stuff. Just use it with good ventilation and try not to get it on your skin.

 

[xeno 108]

I had some great results with XT with a thicker layer hight and 0.8 nozzle

 

[jonnybischof 60]

Ok, so it's not so much the material that matters, but how you print it

Large diameter nozzle and thick layers = nice transparency.

ABS and acetone is probably the only solution that doesn't need a larger nozzle.

And I bet that you get the best results when you combine both large nozzle & layer height with ABS and acetone finishing

 

[pm_dude 27]

Gr5 mentioned somewhere that Ethyl Acetate worked on PLA like Acetone on ABS.

http://www.printedsolid.com/smoothpla/

I wonder if it would work on XT or T-Glase.

 

[jonnybischof 60]

It might work, but it is considerably more dangerous to handle than acetone!

Acetone vapour actually doesn't "explode" when you light it on fire, but burns more like a candle (still I wouldn't try it, but you can look it up on youtube). Whereas many other flammable gases will light up in a fireball, or will just be unhealthy when inhaled in the first place.

 

[printedsolid 3]

Gr5 mentioned somewhere that Ethyl Acetate worked on PLA like Acetone on ABS.

http://www.printedsolid.com/smoothpla/

I wonder if it would work on XT or T-Glase.

 

Actually, acetone works on XT, T-glase, and PET+. Similar to the Ethyl Acetate with PLA, results are much more dependent on technique and settings than they are with ABS. I've found that a very short exposure works incredibly well, but if you overdo it, the material starts to whiten. I suspect that the amount of moisture in the air either when printing or polishing may also be a factor.

There is a lot of room for optimization and deeper understanding, but it's a starting point. Also, this blog post was done with the original lower temperature XT. Not sure if I've tried it with the newer XT yet.

http://www.printedsolid.com/vapor-polishing-colorfabb-xt/

 

[themark 0]

Thanks for this testing method of the model. It's very helpful for me.

 

[DidierKlein 729]

I stumbled upon this method:

http://taulman3d.com/t-glase-optics.html

T-Glass filament from Taulmann + XTC3D coating, result is amazing

 

[Labern 775]

very cool. but does this mean we now also need a spinner?

I wonder how well it works with XT.

 

[pm_dude 27]

Nice!

[selmo 0]

OK, I'm going to deviate from conventional wisdom here. I've gotten some nice clarity with t-glase, trans ABS, etc. by using thin layers. The lack of clarity comes from air bubbles breaking up the light's path through the material. Using thick layers with a single perimeter to get round extrusions of solid material is one way to avoid air gaps, although you get a side wall that's ribbed (the new 2-part epoxy Smooth-On coating or vapor smoothing addresses that issue).

I've experimented and had some good results by using thin layers (below 0.1mm), increasing temperature, increasing extrusion flow, and printing slowly (without a fan, if the material doesn't droop). By smashing down heavy layers of hotter-than-usual plastic, it flows and conforms better, minimizing air gaps. I've gotten some clear parts with multiple perimeters that way. It's not perfect, but I think it looks better than the recommended 0.3+ single wall results. YMMV.

 

I have this sticker on my machine, under the bed.

 

[pm_dude 27]

Good idea. Got some pictures to share?

[selmo 0]

I'll look around and see if I have a good example. One thing I've found is that t-glase has a lot of pigment, such that the transparency is kinda lost when you have multiple perimeters (at least the red and blue, which is what I've got). I'm guessing they designed with single wall prints in mind.

 

[solid-print-3d 30]

I had a lot of luck just using regular clear PLA, a drilled out old nozzle (1mm micro bit from harbor freight), .7mm layer height (70% of nozzle width). So, 1mm nozzle, .7mm layer height, 25mm/s print speed

Tried same setup in T-Glase as well. T-glase wasn't as clear in my case as the clear PLA. Keep in mind I was using 2 perimeters (2mm wall thickness).

Will be trying the Clear ABS and acetone vapor next.