UM3 Recurring issue, the tower!

[Ronan 58]

We own 4 UM3's and this is a occuring issue on all of them.

When printing a tall PLA/PVA print, the ooze tower will often BREAK on the bottom (around half a dozen layers) and then topple over. Most of the time the prints are near completion (our average prints are 2 days long), but this creates issues with finishing, pva support not being properly put, etc.

Yes we upgraded to the latest firmware and we saw an improvement (with the new 'hollow' ooze tower) but the issue still happens. Sometimes it gets unstuck and falls over, but more often than not, it actually breaks on the bottom and then falls over (usually on a PVA layer).

We are using default settings, normal quality. When everything works, it rivals are bigger more expensive machines, but when this happens urgh... more post processing work in finishing the prints, and often means painting them to cover the bandages...

This happens with bare hot bed and with a PVA solution (white glue) applied thinly for a stronger adhesion.

Please advise.

[exforma23 12]

I'm new to this forum so excuse this if its a repeated post.

I'm also looking for something better than glue stick for PLA.

"3D LAC" aerosol spray has popped up in a few youtube videos as the best bed adhesion solution for PLA.

I'd like to hear if others agree.

Also, how to get it in the US since it appears to be tough to get outside Europe.

Thanks

[SandervG 1,521]

Hi @Ronan, thank you for your message.

We are continuously working on improving the prime tower. Our apologies if it has been giving you trouble. While our software team is busy implementing improvements (there are some theories), perhaps this post of @GR5 could help you in the mean time. He adds some support to the prime tower that should prevent it from tipping over.

The problem lies in the adhesion / bonding between PVA and PLA.

@Exforma23, thank you for your message and welcome to the forums!

Why exactly are you looking for a replacement for your glue? Has it not been working for you? Fair chance you may want to improve your bed level instead of looking for a different adhesive.

If you have bed-adhesion issues, they are probably different from why Ronan created this post. (Wipe-tower breaking has nothing to do with bed adhesion).

I know glue works great, but it is not the only thing that works great. There are alternatives. Unfortunately I have never tried 3D Lac.

If you are gonna use a spray, make sure you remove your glass plate from your Ultimaker so you don't cover the entire insides of the machine with sticky spray

[tomnagel 126]

Please upgrade Cura to version 2.4. The wipe tower has been disabled for all profiles, since oozing has been suppressed with smart temperature tricks at start and end of each layer.

I'm glad you mention you use the default profiles. Please keep doing that, at least until you have gotten good results.

Of course, these profiles were optimized for UM materials. When using other profiles, your mileage may vary.

[gr5 2,266]

The glass plates tend to have a little bit of a curvature where the very center is higher and then it's leveled near the 3 screws (front left, front right, rear center). This means the rear 2 corners are lower. Unfortunately this is where the prime tower usually goes (rear right corner). Moving the tower closer to the rear screw helps a lot! That way the bottom layer is squished more into the glass and it sticks better.

[Ronan 58]

Please upgrade Cura to version 2.4. The wipe tower has been disabled for all profiles, since oozing has been suppressed with smart temperature tricks at start and end of each layer.

I'm glad you mention you use the default profiles. Please keep doing that, at least until you have gotten good results.

Of course, these profiles were optimized for UM materials. When using other profiles, your mileage may vary.

 

As stated we are using Cura 2.4 and the ooze tower is on by default (its now hollow).

We only use UM materials, they give the best results (by far).

 

The glass plates tend to have a little bit of a curvature where the very center is higher and then it's leveled near the 3 screws (front left, front right, rear center).  This means the rear 2 corners are lower.  Unfortunately this is where the prime tower usually goes (rear right corner).  Moving the tower closer to the rear screw helps a lot!  That way the bottom layer is squished more into the glass and it sticks better.

 

Will try it! Thanks!

Edited March 17, 2017 by Guest

[tomnagel 126]

You mention you use the latest firmware in your opening post, but I don't see a version of Cura mentioned.

What materials are you printing with? Cause the materials I print with have the primevtower disabled by default in Cura 2.4. (PLA, PVA).

If your prime tower is on by default, I am pretty sure there is something wrong with your Cura installation.

[Ronan 58]

You mention you use the latest firmware in your opening post, but I don't see a version of Cura mentioned.

What materials are you printing with? Cause the materials I print with have the primevtower disabled by default in Cura 2.4. (PLA, PVA).

If your prime tower is on by default, I am pretty sure there is something wrong with your Cura installation.

 

Latest version of Cura, installed fresh after uninstalling Cura 2.3. I'll reinstall it again, see if that changes anything.

[tomnagel 126]

Wait. I don't know the right details but I have seen mentioned on this forum that certain settings or profiles etc. survive reinstall.

@ahoeben, can you help?

[Ronan 58]

Even reinstalling Cura doesn't change the changes done to profiles from previous versions.

I had to manually click on the little 'back arrow' to revert changes to default (which i never changed, but they were simply different in the previous version).

Now, by default, with Cura 2.4 the prime tower is no longer printed.

[gr5 2,266]

I have found when upgrading cura the profiles indeed tend to move from the old version. Uninstalling doesn't seem to make any difference. The files are stored in .../appdata/local/cura. Completely blowing away that folder and then restarting cura initializes all the profiles to defaults again.

[ahoeben 2,012]

The default profiles are readonly and stored in the application folder (c:\program files\cura by default on windows). If you make a change in the sidebar, you should see a star appear next to the name of the profile, indicating that the current settings don't 100% represent the current profile. You can make a new profile out of the combination of the current (readonly) profile and current settings.

Both the "current settings" (ie: changes made to the current profile which are not yet stored in a new profile) and any custom profiles will survive an upgrade, and an uninstall/reinstall.

Pro-tip: if you click the star next to the profile-name, you go straight into the profiles pane of the preferences, showing you an overview of the settings of the profile AND the current settings/changes.

[Ronan 58]

Nuked the folder, reinstalled, all fresh and clean. Just finished a print (seems to print a good chunk faster now too!).

Once it's done taking a bath, i'll let everyone know how it came out.

[SandervG 1,521]

Sounds like the installation instructions could be a bit clearer if @Ronan did not benefit optimally from a new version. I'll see if some instructions can be included with the next release.

[gr5 2,266]

 

@SandervG - I think the installer could have an option that "removes all old profiles". Many people won't want this but when you are trying to do a complete uninstall/install then people could check this option.

[HPx_3D 0]

Guys, we are having the same problem whenever we use PVA combined with other materials. We printed a part over the weekend with UM ABS and the tower broke loose causing what you see in the picture.

[gr5 2,266]

Move your part closer to the front of the machine and move your tower centered left/right and closer to the center of the plate. The glass is higher here such that the plastic gets squished into the glass better and that way the tower should be strong enough to lift the printer by.

[tomnagel 126]

@gr5: the glass plates have some tolerances, and are certainly not always the same. Your instructions will not work for everyone.

In this case: @HPx_3D you combine ABS and PVA. That combination does not work, because the two materials don't stick. Sometimes it works, but it is not reliable.

In any case: consider upgrading to upgrade to Cura 2.4. The printing profiles are getting better and better, and Cura 2.4 contains other tricks to suppress oozing. Wipe towers are turned off by default. Saves a lot of time too.

Do you really need ABS? Care to explain what your requirements to the printed part are?

[Ronan 58]

Would love to use ABS with a soluble material... PLA is quite limiting when 3d printing mechanical parts.

Anyone tried Nylon + PVA?

As for the print we did, came out good, but not great. Unsure if it's the model or not, so will try to print one we already did before to see if there's any posi/nega changes.

Edited March 21, 2017 by Guest

[kmanstudios 1,120]

Would love to use ABS with a soluble material... PLA is quite limiting when 3d printing mechanical parts.

Anyone tried Nylon + PVA?

As for the print we did, came out good, but not great. Unsure if it's the model or not, so will try to print one we already did before to see if there's any posi/nega changes.

 

I have printed with the Nylon/PVA and it works like a charm with minimal effort on the user end.

I have printed pieces that have hollows that need support as well as 90° overhangs and get good results. That would be using the 'Everywhere' option and not 'touching build plate'

So far I have only used the Ultimaker Nylon but will soon try a third party brand.

[tomnagel 126]

PLA is quite limiting when 3d printing mechanical parts.

Very much depends on what you are looking for. Stiffness of PLA is often a little higher, but it is more brittle. But there are PLA variants which have some other polymer mixed into it to make it less brittle (but what that does for stiffness I don't know)

The real big drawback of PLA is that it is not temperature resistant, it gets soft between 50 and 60degC, which e.g. means that you cannot use in your car interior.

I often find that PLA is undervalued by mechanical engineers who somehow think ABS is better in every sense.

@Ronan: Nylon/PVA is a supported combination and works very good. Ultimaker has put a lot of effort in the printing profiles, so if you use Ultimaker materials, you should be printing hasslefree.

[SandervG 1,521]

+1 for Nylon & PVA

[gr5 2,266]

>Would love to use ABS with a soluble material... PLA is quite limiting when 3d printing mechanical parts.

PLA has a bad reputation as being weaker than ABS. It's just not true anymore. Most of the modern PLA brands such as Ultimaker brand and in particular PLA/PHA blends are just as strong as ABS. I've tested materials in bend tests (think snapping a pencil) and pull tests and I've pored over datasheets. Modern PLA and ABS are virtually identical in their strength and flexibility. The major caveat is that often ABS can have bad layer bonding and can be much weaker than PLA along layers. So really ABS should be the material with the bad reputation.

Another mechanical difference comes with drilling or machining PLA versus ABS. Because the process of machining heats PLA above it's softening temp so easily it makes it turn into a gummy gunk when you try to dremel it or drill it.

[gr5 2,266]

>the glass plates have some tolerances, and are certainly not always the same. Your instructions will not work for everyone.

Hmm.  Maybe.  My sample size is smaller than the one you have access to.  But you haven't convinced me yet.  I'm thinking the process of tempering glass makes the center thicker.  I could be wrong but if I'm right then it explains a lot of tower failures.  How many glass plates have you checked?  It's pretty easy - just a few on a flat table and measure with a straight edge (like a metal ruler).  Actually you don't need a straight edge - you just need a flat table.  My plates all have lifted corners and I can kind of rock the glass if I push down on a corner.  I find the center is higher and all the sides are lower.  Both for a 3 year old UM2 plate and a 6 month old UM3 plate.  And my UM2 ext.  And my UMO HBK.  So that's only 4 of 4.  But a few other people I talked to *seemed* to have the same issue.  Not 100% sure.  One said his right rear corner was 2mm low.  That doesn't seem possible.

If you assume the center is higher (like a sphere surface or a gentle hill) and imagine in your mind what that would look like if you leveled at the front 2 corners and the back center it would basically tilt it back so that rear edge is lower.  And the center (where one prints the most) is a bit higher (for excellent adhesion on smallish prints).

I found if I flipped the glass over it was the same thing.  Higher in the center.

Keep in mind we are only talking about .1mm to .3mm for my glass plates.  It's pretty minor except in the rear corners.  And it doesn't take much force at all to make the glass level (put a business card under each of the 4 corners and push down in center with about 1/2 kg of force.  So it's easy for the aluminum bed to be a greater cause of non-flatness than the glass.

[Ronan 58]

PLA is pretty strong, but not as strong as ABS. The bigger issue, by far, is using PLA outside. Just doesn't work in the real world environment. It's not as UV stable as ABS, and temperature is a major issue.

When printing models, PLA all the way, works amingly well. But when we have orders for mechanical parts, or 'working' objects, PLA never cuts it unless its something used indoor, doesn't need to be too strong.

Also ABS is quite easy to finish with a protective layer of acetone. PLA not so much (even with UV stabilize top coat sprays).

Edited March 22, 2017 by Guest