PLA printed object pulls up from the corners on buildplate

[toilet-ovule-add 0]

Hi, 

 

I am having some issues with PLA which i have not had before. 

When printing with the UM3 and Cura default settings (Fine) the prints on the build plate side and top layer seem to flow out a bit.

Also the corners of the object are pulling up. Printed object is sticking well on the build-plate.
Is this related to the build plate temperature? Making a raft (PVA) or skirt (PVA) do not solve the issue.

 

Any tips or hints ?

 

[Smithy 1,135]

Do you print on the glass directly or with some glue stick?

When have cleaned the glass the last time and put a fresh adhesion layer on it?

Which bed temperature do you use? Should be 60° for PLA.

Have you opened the window during the print or is there some other cool airflow near the printer?

 

In another thread you wrote the printer is in an enclosure, is the printer still there? I ask because of the upper layer, could be that there is too much heat inside the printer too. Printing PLA with a fully enclosed printer is normally not a good idea.

 

 

[toilet-ovule-add 0]

 

Smithy,

 

Thanks for your reply.

 

As mentioned i use the cura default settings and have printed directly on glass and on a raft (PVA), so bed temperature is indeed 60 degrees. Never use(d) glue and have never had the need for it. The whole model sticks like glue (no pun intended) on the bed, only the corners of the object seem to pull up or melt. The printer is placed in a cabinet with moisture absorbers so no issue with cold airflows. 

 

If printing in a fully enclosed environment is not a good idea, then why did ultimaker sell enclosures in the past ?
The enclosure is about 10 times the size of the printer so the environment gets warm, but not hot. If the outside temperature issue  the whole model would have been affected. De layers between the lowest and highest are perfectly straight.

 

Cheers,

Peter

 

 

Edited October 20, 2018 by toilet-ovule-add

[Smithy 1,135]

I asked because I have read about problems with full enclosures where the temperature inside was too warm/hot for PLA prints - but it was just an idea. 

 

I tried only a few times to print directly on the glass bed, but had not so good experiences. Therefore I now always have my glass bed glued with a thin layer of glue stick and never have seen warping or lifting corners with PLA.

 

The "flow" on the lower/first layer is normal, it is called elephant foot and comes from the first layer squish to the bed. You can live with it or adjust the setting "first layer horizontal expansion" to -0.25 or so, to adjust it a little bit. But I have no idea what could be the reason for the upper layer.

 

Is your setup new (I mean the enclosure) or have your printed successfully in the past with this setup?

 

 

[rcfocus 0]

I have the same issue recently. For a new bed (glass), there is no issue. But after the glass is used for months, I start to get the issue. So, I think the characteristic of the glass surface must be different. There might be very small particles... or something... that can not be fully cleaned. I use 75% ethanol solution to clean the glass after each print. I may have to try other methods.

 

You can try to take off the glass and clean it very carefully, or you can start to use glue, or change the glass with a new one.

[toilet-ovule-add 0]

13 hours ago, Smithy said:

I asked because I have read about problems with full enclosures where the temperature inside was too warm/hot for PLA prints - but it was just an idea. 

 

I tried only a few times to print directly on the glass bed, but had not so good experiences. Therefore I now always have my glass bed glued with a thin layer of glue stick and never have seen warping or lifting corners with PLA.

 

The "flow" on the lower/first layer is normal, it is called elephant foot and comes from the first layer squish to the bed. You can live with it or adjust the setting "first layer horizontal expansion" to -0.25 or so, to adjust it a little bit. But I have no idea what could be the reason for the upper layer.

 

Is your setup new (I mean the enclosure) or have your printed successfully in the past with this setup?

 

 

 

I always cleaned my glass with a drop of water and and some drops of ammonia. Rub this out with a paper towel till you feel a lot of resistance and the plate is dry. Ammonia removes fat much better than alcohol. 

Never have had the elephant foot issue and the flow on the top layer. See the attached photo of the frame which is printed in the orientation as shown. Printed about 10 of these frames successfully with the same printer and enclosure.

 

Will try the horizontal expansion setting next time, and see if this will help. 

[Smithy 1,135]

7 hours ago, rcfocus said:

But after the glass is used for months, I start to get the issue. So, I think the characteristic of the glass surface must be different. There might be very small particles... or something... that can not be fully cleaned.

 

Yes, that's exactly why I use a thin layer of glue stick every time. I don't have to clean the bed after each print and such a layer of glue lasts days/weeks before I have to clean the glass and apply a fresh layer. And I don't have to worry that the glass was not clean enough and a larger print comes off. I apply the glue very rough, zig zag and then distribute it with a wet tissue to get a very thin layer. Thats enough. 

 

20 minutes ago, toilet-ovule-add said:

Will try the horizontal expansion setting next time, and see if this will help. 

 

This will only help for the initial first layer. If you set the other horizontal expansion setting, then the whole object will shrink or grow in x/y, depending on the value.

 

 

[JohnInOttawa 103]

Good morning.

 

Looking at the second print image, I see three areas where the print appears to flare out - the interface with the plate, the other side of the main plate, but also the bottom of the 'foot' (my term).  I just want to first confirm the upper surface in that image is what is against the build plate and that there are, in fact, three levels where there is widening. 

 

Can you verify which Cura version was used for this print?  Has it changed since you last did this successfully?  If so, any chance something in the profile got changed without your input?  I'd think about things that affect flow rate, anything like speeds or settings affecting top layers, changes to temperature for the top layer or mid print, stuff like that.

 

A couple of things you have likely already checked - belts, rods, lubrication, even Z screw cleanliness and lube...

 

Finally, and this may seem strange - have you measured to ensure that the wide areas are the problem and not that the narrow areas are somehow undersize?  I ask this because the warping suggests something is shrinking as it cools beyond the limits of adhesion. 

 

I appreciate none of this actually solves your issue.  Just things I'd look at.

[geert_2 556]

Lifting corners is normal when printing on bare glass: in my experience the environmental effect plays a big role and is hard to predict: temperature, moisture in the air, wind,...

 

If you want to be sure, use glue. Or if you want to keep printing on bare glass without glue, try my salt method:

- first thoroughly clean the glass,

- then clean again with pure tap water only (no soap, no dish washer,...),

- wipe the glass with a tissue moistened with salt water,

- genlty keep wiping while it dries, so it leaves a thin almost invisible mist of salt stuck to the glass. It should look just a little bit dusty.

 

The glass bed has to be heated to ca. 60°C. Without heating there is no bonding at all. After cooling down, models come off by themselves, without any effort. The ease of applications (just wiping with a moist tissue) and ease of removal are the most attractive aspects for me.

 

For PLA this greatly increases bonding for me. But it does not work for other materials.

 

Also, it is not optimal for thin, very tall objects like poles: the salt does not flex, and can not absorb any shocks. In this case you might be better off trying glue.

 

Whatever bonding method you try, stay with the printer during the first tryouts, and observe what happens.

 

[jbebel 0]

I found that pre-heating the glass plate to 60 and letting it sit for a few minutes once it got there to even out greatly enhanced both the quality of my first layer and the adhesion on the corners.

[toilet-ovule-add 0]

On 10/21/2018 at 12:25 PM, JohnInOttawa said:

Looking at the second print image, I see three areas where the print appears to flare out - the interface with the plate, the other side of the main plate, but also the bottom of the 'foot' (my term).  I just want to first confirm the upper surface in that image is what is against the build plate and that there are, in fact, three levels where there is widening. 

 

The upper surface was indeed the side against the build plate. There are two points where it flairs out, at the build plate side and on the other side of the main plate. The legs/feet are straight.

 

On 10/21/2018 at 12:25 PM, JohnInOttawa said:

Can you verify which Cura version was used for this print?  Has it changed since you last did this successfully?  If so, any chance something in the profile got changed without your input? 

 

I currently use Cura 3.5.1. I have never printed this object with an other version of Cura, it is my first concept print. I used the standard 'Extra Fine' profile.

On 10/21/2018 at 12:25 PM, JohnInOttawa said:

A couple of things you have likely already checked - belts, rods, lubrication, even Z screw cleanliness and lube...

 

Yup, also issues with any of these would affect not only the first and last layer.

On 10/23/2018 at 2:26 PM, geert_2 said:

In this case you might be better off trying glue.

 

Yes more people are suggesting it, but the print sticks fine, only the corners seem to retract/melt from the build plate.
 

On 10/21/2018 at 12:25 PM, JohnInOttawa said:

Finally, and this may seem strange - have you measured to ensure that the wide areas are the problem and not that the narrow areas are somehow undersize?

 

Sorry, do not understand this. What should i measure ? It is a rectangle shaped box, where the retraction occurs on all four corners.

 

[JohnInOttawa 103]

OK.  You might want to try this again in Cura 3.4.1 just to rule out a software issue.

 

Good you checked.  But just for clarity - Issues with belts, rods, lube and Z do not affect all layers equally.  For Z in particular, it depends where the dirt or other source of friction is.  Pulleys slip sometimes as a function of jerk and that is dependent on geometry and print speed.  Not all layers print at the same speed with the same acceleration, so tension and tightness issues can surface in wierd spots.

 

The last comment pertains to shrinkage of the model as the filament cools or print temperature and fan settings change.  Granted this would be a dramatic example.  But since no root cause has yet been found, thought I would verify that the dimensions are too big in spots, and not too small.

 

Cheers

J

[toilet-ovule-add 0]

23 hours ago, JohnInOttawa said:

But just for clarity - Issues with belts, rods, lube and Z do not affect all layers equally.  For Z in particular, it depends where the dirt or other source of friction is.  Pulleys slip sometimes as a function of jerk and that is dependent on geometry and print speed.  Not all layers print at the same speed with the same acceleration, so tension and tightness issues can surface in wierd spots.

 

Ok, checked it again to be sure, z-axis is clean and slightly greased, all belts have the correct tension (on both sides).

 

23 hours ago, JohnInOttawa said:

But since no root cause has yet been found, thought I would verify that the dimensions are too big in spots, and not too small.

 

I do not get this. Which dimensions are you referring at, the dimensions of the printed model ?

 

I am now running a new test print with a lowered build plate temperature (55c) and a temperature of 60c for the initial layer.
A PVA raft is printed to see if this together prevents the elephant foot and retraction from the build plate at the corners of the model.

So far so good...

 

[toilet-ovule-add 0]

 

And it came out perfect now. No elephant foot and lifting corners, (almost) perfect. Additionally i keept the door of the closet the printer stands in slightly opened to lower the environment temperature. 

 

 

Thx for all the replies!

 

[Smithy 1,135]

Fine that you found a setup that works for you.

 

But just for the records, a thin layer of glue (it is also PVA based) and the setting "initial layer horizontal expansion" against the elephant foot, brings the same result, but in my opinion much easier, without dual extrusion and without the expensive PVA not really needed here.