Rough surface with PLA+PVA print

[phi90 1]

Hi all!

 

I am trying to print a PLA+PVA (AA0.4/BB0.4). The PLA has a partly rough surface, which I have never seen when printing on UM3 without the PVA.

After dissolving the PVA, the print looks like this:
 

 

This is printed with the standard Cura "Fine" profile (without prime tower). 

 

Has anyone encountered a similar problem and has been able to fix it?

 

Cheers!

T

 

 

Edited March 15, 2018 by phi90

[gr5 2,026]

Unfortunately, while your photo was just fine yesterday it appears you have deleted it off your invisioncic server.  Please repost the picture.

[phi90 1]

I don't seem to be able to edit the original post anymore？ Anyway，here it is！

[JohnInOttawa 103]

I've had similar issues when trying to print a threaded coupler with PVA support.  Sadly I can't offer an explanation as to why this happens, but to my eye with my model, and perhaps yours, it seems as though there is a bit of interplay between the support model and the primary one.

 

Not the same as trying to print a two color model with two AA cores, but something along the spectrum from totally integrated to totally separate.  In my case, the layers in the threads seemed to come out 'un merged' where it looked like  interleaving between PVA support and PLA model layering, when I expected just PLA.

 

I know your model is pretty complex and clearly benefits from support. Have you tried printing a partial version of it without PVA support to see if the surface texture is better (that is, in the areas that don't sag without support)?

 

Also, and pardon the dumb question, but is there any possibility that age or moisture level of the PLA or PVA are a factor?  In my case, I put both the PLA the PVA in a dryer for a day or so and tried again, though not perfect it did make a difference.

 

John

Edited March 11, 2018 by JohnInOttawa

[phi90 1]

Thanks for the advice!

 

I haven't dried the PVA yet as it was fairly fresh out of the packaging. By that I mean it was out for a few days. I have no feeling for how long PVA can be outside before moisture becomes a problem. I guess it depends on the ambient moisture. But how long can I go without worrying if the ambient moisture is "average"? It is in a semi-sealed box to protect it from dust in the workshop. It gets pretty warm in there. Since there are no water sources inside that box, I would assume that the high temperature would decrease the relative humidity and hence be good for the PVA. But I have no experience with this at all. 

 

I heard that PVA turns from translucent to opaque once it has a higher moisture level. My support structure was still translucent, so I am assuming humidity was not an issue here.

 

The extrusion temperature was 215˚C, which is on the lower end of what Ultimaker recommends. However, the internet usually talks about lower temperatures like 190-200˚C. I haven't tried playing with this yet. have people found this to affect the PVA support quality?

 

Ultimaker says that PVA should be stored below 28˚C. This is clearly not the case for me.. However, the glass transition temperature of PVA is 85˚C, so I don't see where the 28˚C is coming from (except for the humidity concern?). 

[gr5 2,026]

The giveaway that you have too much water in the filament is that you can hear the steam popping and crackling as it prints.  And sometimes you can see steam.  And sometimes you can see water condensed on the cooler parts of the print head.

 

PLA doesn't need to be kept dry.

 

You should keep the pva sealed in a bag with dessicant even if you aren't using it for just an hour.  

 

Nylon is much worse - if you leave it out for an entire day it definitely needs drying.

 

I doubt any of this is related to your issues.  I don't do enough PVA prints to be able to comment on your print.  I don't know if that's normal quality or bad quality and I don't know how to fix it if it is bad quality.  

 

I'm curious if the PVA was making a solid layer just before the PLA was printed on top.  There's an option related to that in cura somewhere.

 

@kmanstudios - what do you think of that quality?  Is that typical for you?  I get the impression you tend to paint everything so I don't know if you worry about increasing surface quality all the time or just enjoy printing away all your creations without worries.

[phi90 1]

2 hours ago, gr5 said:

I'm curious if the PVA was making a solid layer just before the PLA was printed on top.  There's an option related to that in cura somewhere.

 

No, it didn't. I had the brim printed in PVA, but the first layer of the model itself was still PLA. 

 

I am not certain that the PVA is what is causing the imperfections. I tried to print an arch with PVA support on one side and no support on the other and both sides came out perfectly fine. If anything, it looks like the "no-support"-side has a similar roughness artifact on the non-overhanging side:

 

That being said, I tried to print my complex model again and it came out worse. This time, instead of just having the PLA "bumps", it also had holes of similar size and shape. I feel like this is where PVA was inside the printed geometry, similar to what I imagine JohnInOttawa was talking about.

 

I will keep experimenting, but it would be super helpful if someone else has come across this previously  

[kmanstudios 1,120]

PVA can be a tricky thing. I have run into issues with it, but mostly because I made the PVA too loose/not dense enough. I also made sure that I put the resolution of the PVA surface to as high as I was comfortable with.

 

I will try to get some screenshots together to show what I mean. But I just got back from half a day at the VA getting all antibioticed up and am about to go to bed for a while.

 

To answer @gr5's question, "sometimes." LOL

 

Seriously, I do a lot of abstracty experiments where I want the nature of the 3D printing to have an effect as part of the piece, but, not always as you can see in this post. Really depends on what I am doing and what I am looking for as far as finish.

 

Page 2 of this shows where I wanted as little 3D artifacting as possible especially with the PVA materials.

But some of my stuff Is really out there as far as being textured goes. like this:

 

[kmanstudios 1,120]

OK...got delayed, so I did this first then bed

 

Check that your profile and support resolution are at least at the same settings. But, I have put my support resolution at a lower than profile setting to try and smooth things out.

 

 

The blue area is where you want to make sure the densities are 100% to be as solid and smooth as possible. Also, check the interface pattern to see how it works for you.

 

Also, some of your stringies are able to be taken care of with the wave of a heat gun. Finally, without any size comparison I am looking at the line widths you say you sliced at (Fine = 0.1mm) so I am guessing that the thinness of the edges are not much larger than 0.3 - 0.4mm in thickness. So these could be really tiny imperfections.

 

As for the PVA inside the PLA parts, yeah, I have had that happen. I try to cook my PVA a bit by putting temps up a bit to harden it. Also depends on the brand used and moisture content. But I see a lot of difference on what would be assumed the underside of the arch print. In the orientation of the pic, it would not need support. But if it was an arch with the flat sides down and the arch going over, the side you listed as having support is straight and consistent on the corners (from underarch to side walls) while the other side is wiggly and loose until there is enough density to merit a surface to build off of and the corners between the underarch and sides are straggly as well until it has enough density to build upon.

 

Dats whut I see......

Edited March 14, 2018 by kmanstudios
typelexia

[kmanstudios 1,120]

9 hours ago, phi90 said:

However, the internet usually talks about lower temperatures like 190-200˚C. I haven't tried playing with this yet. have people found this to affect the PVA support quality?

This is very brand dependent. I cook mine at about 230°C when printing.

PVA coloration will vary according to brand.

PVA time spent before absorption is depending on local temp and relative humidity.

 

Annnnnd, sometimes, it can be stored or packaged in such a way before it even got sent to you that it got moisturized, just a not so perfect batch. That can happen. I have opened brand new PVA and put it on immediately and it sizzled and popped immediately when printing.....

[kmanstudios 1,120]

I am curious, if you printed with the UM2 without PVA, what were you using for support? Also, have you played with your retraction and temp settings for the PLA as well?

 

Can you show a comparison print?

16 hours ago, gr5 said:

@kmanstudios - what do you think of that quality?  Is that typical for you?  I get the impression you tend to paint everything so I don't know if you worry about increasing surface quality all the time or just enjoy printing away all your creations without worries.

I will have to say that now that I am back to printing some deterministic models (Not Abstractish or fractals) I am revisiting my own settings. I do tend to paint everything, but I am trying to also cut down on post processing. And with the new filaments coming out with the  new diffuse properties, flaws are more prominent now than layer lines and I am experimenting with them at the moment.

 

Right now one machine is dedicated to this type of printing while the other is doing long term, weird stuff. I will know in 2 days on an engine I am printing at about 12th scale. Definitely not 1/8th as I have a model that size and the engine alone would fill up the buildplate. BUt it has all the nuts and bolts and stuff, so we shall see.

[phi90 1]

Thanks for the advice! A lot to go through.. 

My print settings are the same as in that screenshot, so that is fine. Also I am pretty sure that the support roof/floor isn't the problem as the stringing is more (or equally) dominant on the surfaces that aren't overhanging. 

 

12 hours ago, kmanstudios said:

Also, some of your stringies are able to be taken care of with the wave of a heat gun. Finally, without any size comparison I am looking at the line widths you say you sliced at (Fine = 0.1mm) so I am guessing that the thinness of the edges are not much larger than 0.3 - 0.4mm in thickness. So these could be really tiny imperfections.

That is correct. The faults are quite small (but detrimental to the overall look). I haven't tried using a heat gun yet! 

 

11 hours ago, kmanstudios said:

But I see a lot of difference on what would be assumed the underside of the arch print. In the orientation of the pic, it would not need support. But if it was an arch with the flat sides down and the arch going over, the side you listed as having support is straight and consistent on the corners (from underarch to side walls) while the other side is wiggly and loose until there is enough density to merit a surface to build off of and the corners between the underarch and sides are straggly as well until it has enough density to build upon.

Yup, that is correct. I laid the arch on the side so you can see the underside. 

 

2 hours ago, kmanstudios said:

I am curious, if you printed with the UM2 without PVA, what were you using for support? Also, have you played with your retraction and temp settings for the PLA as well?

woops, that was a typo. It was actually UM3. I was using PVA for support. 

 

Retraction and temp settings are a good idea. Will play around with them as soon as I get the chance! The printer is shared with a couple of people so I can't just play around with it whenever I want. 

[kmanstudios 1,120]

What is the size of the thing yo printed? I will tell you that I have noticed that if it is in a bunch of small areas and voids, that the PVA and PLA can cross pollute each other. I have found it does much better on larger expanses.

 

It was @gr5 that suggested the heat gun to me and it works wonders. I use the heat gun on my SMD station.

 

You also said it looked better before when printed...what print was that and what did you do different?

[gr5 2,026]

17 hours ago, kmanstudios said:

I cook mine at about 230°C when printing.

What?  You mean in an oven?  It will melt all over the place at 230C.  Even at 230F it will likely slump quite a bit.  I think PVA gets soft around 80C?  I forget - I tested it once and wrote it down but I don't want to search through my notes for 20 minutes.  I should have put that value in my materials properties file on my computer.  Oh well.

 

[kmanstudios 1,120]

1 hour ago, gr5 said:

What?  You mean in an oven?  It will melt all over the place at 230C.  Even at 230F it will likely slump quite a bit.  I think PVA gets soft around 80C?  I forget - I tested it once and wrote it down but I don't want to search through my notes for 20 minutes.  I should have put that value in my materials properties file on my computer.  Oh well.

 

I definitely need to clarify the way I phrased that. Thanks for pointing that out as I do not wish to be confusing.

 

If you take a look at the image I have attached, and these are the default temps for PVA, that the printing temp is 215°C and the starting temp is 225°C. So, sometimes I will print at a higher temp to make it cook a bit more and then get a bit harder/sturdier when I am trying things. For instance, this is the print that is on the old machine right now. It has 38 hours to go. I am most interested in how the wiring and fan blades will come out.

 

 But, when I say 'cooking the filament' in this case, I actually meant during the printing process. And, I meant the PVA. The PLA materials I am using print quite nicely at 195°C and that temperature differential does help a lot. ie, the PLA cannot melt into the PVA.

 

When I put a print in the oven to Anneal it, I do refer to it specifically as 'annealing' at a temp. So, yeah, if I annealed at that temp, it would most definitely do as you say.....turn into an Irish print....Puddle O'Goo.

 

[phi90 1]

I got a chance to play around with it a bit more and I am starting to suspect that the fault is actually just with the PLA, while the PVA is making it difficult to diagnose.

I printed the same geometry with big support-xy-distance, small support-xy-distance and no support (see below)

Sorry for the bad quality of the images, this was taken with my phone.. But you can still make out the important stuff:

 

the two prints on the left look very similar, even though they had a different support xy-distance. I think that if the PVA was at fault, they should look different. 

The print without any support shows that the PLA is stringing. I wasn't able to see that when the PVA was in the way as it significantly decreases stringing. That wasn't immediately apparent to me, but now it seems kinda obvious. 

 

Since other prints come out perfectly fine I suspect the stringing is mainly a retraction issue rather than a temperature issue. These are the settings I have ben printing with: 

 

 

will play around with those when I get the chance

 

Edited March 21, 2018 by phi90

[gr5 2,026]

Some filaments are worse than others for stringing.  The "worse" filaments need a lower printing temperature.

 

Although you might not be retracting.  You have so many "islands" on a given layer that there are probably 100 retractions on each layer.  Too many retractions on the same spot of filament can cause a grinding failure so Cura tries to be careful and limit that.  If you remove that limitation (it's maximum retraction count and minimum extrusion distance window) you may get a failure.  Or you might not.  I found 10 retractions on the same spot is pretty safe.  20 is pushing it.  40 consistently causes failure.  But every printer is different.  Even among the same model of printer.  It depends on other things as well like how fast you are printing (how high the pressure is in the print head which affects how hard the extruder is pushing on the filament).

 

 

[kmanstudios 1,120]

I started to use @gr5's suggestions about the retraction count and it has made a difference in other models as well on the sides and such.