XSTRAND GF30 PA6 Test Print

[Gabriele 8]

Hi, I just received the new XSTRAND GF30 PA6 filament and the CC print core. 

I did a test print with official parameters, no modification at all and this is the result.

It seems there are stringing problems especially with the support. 

Any idea on how to fix it? 

 

(P.S: I printed it using a Polybox with the filament at 22% humidity)

 

 

 

 

 

Edited November 26, 2018 by Gabriele

[Gabriele 8]

Same print at a bigger scale.

 

 

 

 

[JohnInOttawa 104]

Wow.  Hopefully Owens-Corning can engage.

 

That's just painful.

 

John

[NBull 14]

How is you Polybox set up/connected?

Not saying it's the deal breaker but i would try and print whit out the Polybox.

My experienc wit the Polybox is that when i use it with 2.85mm fillement it tends to grip (friction) the filement to much so the feeder has a hard time pulling and pushing the fillement back in the box.

[Gabriele 8]

The main problems is very bad stringing and support consistency.

At the moment I'm attempting the same big print with these tunings:

 

Limit support retraction: disabled

Support density 25%

Support pattern: Triangles

Support overhang angle 52 (with some support blockers)

Build plate adesion: Brim

 

I attach an image of the the result (19 hours left).

At the moment I can see:

• The support is much more consistent where there are large support areas, small support areas are still problematics. 
• Stringing is huge especially in longer travel moves, I checked and despite the retraction a fast travel move result in a 4cm string puring from the nozzle. 

@NBull I print lot's of Nylon and PVA with the Polybox and never had a problem. 

I checked carefully and the retractions are clearly visible. I saw during some travel moves with correct retractions that at least 3-4cm of material pur as a string from the nozzle. 

 

I'm thinking about lowering the nozzle temperature but since the material is made from Owens-Corning itself with same hardware I'm using it seems strange the temp is not correct. 

 

 

[Gabriele 8]

I lowered the temperature 10° during the print shifting from 260° to 250° and despite the retraction happens in this video you can clearly see the string of filament flowing from the extruder during an horizontal move.

 

 

[P3D 46]

Before attempting other things, maybe you could try to dry the material (real drying, for at least 4h, preferrably more, at 80°C) not just putting it in the Polybox). This way, humidity could be ruled out as the cause once and for all. I had these kinds of issues with other filaments, but only from wet material. While I do hope that the manufacturer isn't as careless as Ultimaker is with packaging their TPU filament, there is still the chance that the material got wet some way or the other.

[SandervG 1,521]

I also suggested this in another thread. It's actually recommended to dry the material before you use it, which could influence stringing. There is a note about this in the 3D printing guidelines: 

"* IMPORTANT NOTE: GF30-PA6 is very sensitive to moisture. For optimal printing results, the product must be dried at least 4 hours in the oven at 80°C prior to use and stored in a dry, Pelican like case, even during printing."

 

[Gabriele 8]

@SandervG @P3D The material has been dried at 80° for 4 hours as suggested and immediately placed in the polybox.

Today I will try a new spool, I hope is just a material defect.

 

[AD_XSTRAND_OC 21]

5 hours ago, Gabriele said:

@SandervG @P3D The material has been dried at 80° for 4 hours as suggested and immediately placed in the polybox.

Today I will try a new spool, I hope is just a material defect.

 

Gabriele, We are sorry for the initial experience with X-Strand. Our team is working hard to get the corrected Print profile updated in CURA. For the time being can you mimic these settings?

 

Retract 12 MM at a sped of 40 mm/s

first layer height of .12mm for 0.1mm layers

Adhesion solution: Blue tape, Dimafix, Elmers Glue Stick

printing speed 25-40 mm/s

Print temp-240°C

Regular Fan Speed - 50%

Build Plate Temp - 60°C

Infill Speed- 50 mm/s

Initial Layer speed 20 mm/s

Minimum Speed 5 mm/s

Minimum Layer Time - 3 s

Build Plate Adhesion -skirt

[Gabriele 8]

Hi @AD_XSTRAND_OC, thanks for the support. 

I tried the small scaled print with your parameters and the outcome is way better than the original one (first set of photos posted).

There are still a small amount of strings especially in zones where the layer area before a travel movement is small.

Another problem I still notice is the adherence between the support and the print. You can see it in the first image attached, the support did't stick the inclined printed wall.

 

I will attempt now to print it full scale with a designed support base on the inclined wall an I'll share you the result tomorrow (about 25/30 hour printing time)

 

 

 

 

[Gabriele 8]

After about 6 hours of printing the UMS5 started to pause with a "empty filament error" each 5 minutes although the filament was in the feeder correctly and the spool was half full without any impediments. I disabled the filament sensor and after a couple of hours more the nozzle was completely clogged with the filament weared out completely by the feeder.

I had to disasseble it completely to unclog it, totally I printed one and a half spool of XSTRAND with that CC nozzle. 

 

Now I cleaned everything with a cleaning filament, moved the Polybox closer to the machine shortening the bowden tubes (despite I never had problem with other filaments and the Polybox) and started the print again.

 

Maybe the 12mm retraction distance is a bit too much and the feeder continuos movements wear out the filament?

Maybe the temperature is a bit too low and the nozzle start to clog?

 

[P3D 46]

Thank you very much for this thread! It makes me think if the CC core is really worth the investment, if there are so many problems with one of the primary materials I intend to use it for...

[SandervG 1,521]

2 hours ago, P3D said:

Thank you very much for this thread! It makes me think if the CC core is really worth the investment, if there are so many problems with one of the primary materials I intend to use it for...

 

I don't think any of the difficulties Gabriele is facing has anything to do with the CC core. I think it's great that @AD_XSTRAND_OC has joined us and is working with Gabriele to help out. 

[P3D 46]

21 minutes ago, SandervG said:

 

I don't think any of the difficulties Gabriele is facing has anything to do with the CC core. I think it's great that @AD_XSTRAND_OC has joined us and is working with Gabriele to help out. 

 

Actually, this was not what I intended to say, even if it can be read that way. I was really referring to problems with the material, not so much with the printcore.

 

Anyway, I will wait for these teething problems to be sorted out, don't want to be a guinea-pig for those things, whatever the problem cause turns out to be (and now that you mention it, a defective CC core isn't out of question yet). Those prints still look pretty bad, and nowhere near the quality level I'd expect.

Edited November 30, 2018 by P3D

[SandervG 1,521]

Oh, my bad if I mistakenly read between the lines. Apologies. I'm sure we'll sort this out and start producing quality parts soon! 

[P3D 46]

8 minutes ago, SandervG said:

Oh, my bad if I mistakenly read between the lines. Apologies. I'm sure we'll sort this out and start producing quality parts soon! 

 

No problem, reading my post again, I noticed that indeed it could be read like you interpreted it. Looking forward to getting great, strong, reinforced parts out of the S5 :)

[Gabriele 8]

11 minutes ago, P3D said:

 

Actually, this was not what I intended to say, even if it can be read that way. I was really referring to problems with the material, not so much with the printcore.

 

Anyway, I will wait for these teething problems to be sorted out, don't want to be a guinea-pig for those things, whatever the problem cause turns out to be (and now that you mention it, a defective CC core isn't out of question yet). Those prints still look pretty bad, and nowhere near the quality level I'd expect.

 

I don't think the difficulties are related to the print core itself, I think the issues are caused by the material profile. 

I do a lot of research with materials as I use 3D printing for prototypes but also for functional parts. 

The CC core itself unlock interesting opportunities, from what I saw at the moment the XSTRAND PA is an incredible opportunity. 

The material when printed is really stiff and strong, it has very high softening temperature and zero shrinkage (a difficult combination to find). Despite the filament is very sensitive to moisture printed parts left 24 hours submerged in water are still rigid as they where before immersion (at the moment tested just on small print). 

 

To be honest I though that the material profiles where lot's more accurate, this material is dedicated to professional use where "working out of the box" mean avoiding tuning time related costs. Material cost is quite high, at the moment I almost used 3x 500g spool just for tuning purpose. 

 

I think the parameters provided by @AD_XSTRAND_OC gives huge improvements on the predefined Cura ones and are acceptable for low sized prints (if you consider a couple of things as you model, for instance avoid having support growing on inclined walls because it doesn't stick) as you can see on the images.

 

The bigger print is still stringing and two support already broke as you can see in the image attached.

In the first image what seems stringing on the left part of the print is actually a broken support resulting the printer printed in free air.

The printed part though at the moment is almost perfect.

Walls are consistent and clean both in corners and straight line, ripples and layering problems in small section are completely gone. 

@AD_XSTRAND_OC At the moment I think the zig/zag infill pattern with 0 Support Wall Line Count is too delicate the material is very stiff and head movement brake support tower when they start to grow vertically. 

 

The next print I will go again with the same settings (if you don't have other suggestion) but with triangle pattern that gave me more consistency with the support in past tests. 

 

 

 

 

 

[P3D 46]

38 minutes ago, Gabriele said:

The printed part though at the moment is almost perfect.

 

Great that this works for you! I think my idea of "almost perfect" is a bit different, as I compare these to the parts coming out of my companies MarkForged printer, where any stringing/worm-like artifacts are usually an indicator something is amiss, either with the material (the most probable case) or with the print head/feeder/bowden tubes

 

Thank you very much for keeping us all posted, though! I hope Owens Corning will in some way compensate you for (re)doing their testing work.

Edited November 30, 2018 by P3D

[Gabriele 8]

4 minutes ago, P3D said:

 

Great that this works for you! I think my idea of "almost perfect" is a bit different, as I compare these to the parts coming out of my companies MarkForged printer, where any stringing/worm-like artifacts are usually an indicator something is amiss, either with the material (the most probable case) or with the print head/feeder/bowden tubes

  

Thank you very much for keeping us all posted, though! I hope Owens Corning will in some way compensate you for (re)doing their testing work.

 

"Almost prefect" was referred only at the printed part.

I will post the detail when the print finish but at the moment the printed layers looks really clean if you don't consider the string (that you can easily eliminate).

 

Maybe comparing the UMS5 with a MarkForged printer is a bit unfair, I think they have completely different prices 😉

I agree (as other forum users) that that Cura profiles are far from "plug and play" especially if you do engineering work, at least on my experiences with the UMS5 where I created custom profiles for every material I use. This is a perfect example, the nozzle temp provided in Cura is 20 deg over the printing temperature shared By Owens Corning in this post that is providing the best results at the moment.

 

[P3D 46]

57 minutes ago, Gabriele said:

 

"Almost prefect" was referred only at the printed part.

I will post the detail when the print finish but at the moment the printed layers looks really clean if you don't consider the string (that you can easily eliminate).

 

Maybe comparing the UMS5 with a MarkForged printer is a bit unfair, I think they have completely different prices 😉

I agree (as other forum users) that that Cura profiles are far from "plug and play" especially if you do engineering work, at least on my experiences with the UMS5 where I created custom profiles for every material I use. This is a perfect example, the nozzle temp provided in Cura is 20 deg over the printing temperature shared By Owens Corning in this post that is providing the best results at the moment.

 

 

A OnyxOne goes for about half the price of an S5, though. Of course, it is not nearly as versatile as an S5, but still.

[JohnInOttawa 104]

Thank you for this thread.  I'm paying close attention and learning from your experience.

 

I have to agree though.  That is a lot of expense for testing.  I hope you get a 'pioneer's premium' for doing this beta work.

 

John

[Gabriele 8]

I just managed to finish the first acceptable big print 🎉

Here are the results and some considerations, just for reference this is a 27 hour print, the object bounding box is around 150z; 180x; 150y. The profile I used is attached.

 

Stringing

Stringing is heavily reduced but still present especially on support when a long travel move is required.

Strings that start from support are much longer respect that start from the printed object.

Pay attention to the images, in some part what seems to be strings is support printed in air since some support towers broke during print.

 

Support

Two support towers broke completely.

Since in the past print i never managed to have reliable support growing from inclined walls on print I edited the model to have some horizontal base for support.

I will try next print with triangle pattern.

 

Layering

Overall layering seems good, there are problems on smaller details, like the wall next to holes, as you can see from the closeup that there is a sort of under-extrusion. 

In some part horizontal lines are visible.

Gray colored gradients are visible along the Z axis

 

Dimensional Accuracy (see the image to localise the part on the print)

Side walls expected: 3.65mm printed: 4.15mm

The open part near to the top expected: 20.2 printed: 19.59mm

Side to side distance of the lock lever expected: 30mm printed: 30.4mm

Distance from support base to lock lever expected: 20mm printed: 19.58mm

Thickess of the top surface expected: 4mm printed: 4.2mm

Hole diameter expected: 5.2mm printed: 5.0mm - 5.2mm slight ovalized

 

Additional Consideration

Although I had this filament at 80 deg for 4 hours and placed it immediately in the polybox where humidity is 16%-22% and where it stayed about 48 hours (print time included) I heard in some part during printing a noise like a "shshshshsh", could be moisture in the filament? I never heard the classic pop sound that usually wet filament produce.

For the next print I will bake the spool at 80 deg for 6 hour and then start the print immediately.

 

 

 

 

 

0.2 XSTRAND PA (WIP rev5).curaprofile

[JohnInOttawa 104]

That's encouraging progress!

 

I know the manufacturer is recommending XSTRAND for the supports as well, but given all of the issues, I'm wondering if another material that is known to do well as a support would work better.   I haven't worked with Breakaway and I'm not sure if PVA could tolerate the temperature environment without just degrading.  Would PLA be a possibility?

 

John

[AD_XSTRAND_OC 21]

Hips works great as a support for PA6