Raft Create bad print on final layers

[pm_dude 27]

Hello,

I have a model I'm trying to print and its all on supports. When I add the Raft to help the support stick better it create an offset on the filnal layers of the print and it pretty much ruin the print... I also tried adding a second model taller next to it to test if it was gonna shift the issue further but it doesnt...

Its like if the final layers were all made at the same height.

It actually reflect a bit what I see in Cura and it doesn't seems normal to me.

Its the first time I use Raft in a print and I dont understand why its doing this.

Sounds familiar?

Thanks

PM

 

[illuminarti 18]

That looks like a bug in Cura, but I'm not sure, as virtually no one uses raft on an Ultimaker that I know of. Even for something like this that has a lot of fine or support pillars that need to be stuck down, enabling brim is usually sufficient.

I'm more curious though why you're printing that piece with support? Isn't it designed to be printed without support, standing up?

 

[pm_dude 27]

When I print it standing I get really bad results at many places betwen the legs and all around the face because it as many overhangs.

I'll try with Brim instead... I would prefer by far having no supports at all but I just dont manage to get something nice...

 

[illuminarti 18]

Print several at once, so the layers have plenty of time to cool.

 

[pm_dude 27]

Wow that made a great difference.

Now with no support and printed standing.

I still had some bad overhand but when changing layer height from 0.06 to 0.1 it removed almost every issue. I guess thinner layers is not always what you want

Thanks illuminarti. This have been very helpfull and educative information!!

PM

 

[gr5 2,173]

I tried to duplicate the bug shown in the first photos and wasn't able to. Cura seems fine with or without raft.

 

[pm_dude 27]

Hi again.

@GR5 you are right Raft is not the issue.

The issue seems related to aborting prints... I've been wondering for a while why the initial layer wasnt always the proper height and I think aborting prints and restarting in the same session is causing this (and the offset at the end of the print).

I've change my Feeder to iRoberti version and needed to get the good tension again so I was aborting quite a few prints. The I started seing the initial layer getting thinner and thinner. I'll make some videos to show it.

Is this a know issue (aborting seems to not reset to proper height)

EDIT:

The abort seems to cause the initial layer to get thinner but its not the cause of the final layers to be a mess.

It happens only when the layer get smaller and smaller near then end of a print. The model I have an issue with ends with a small pyramid and when it reach the point where no other things are to be printed the printer doesnt seems to care about the minimum time per layer and at that point it spend about one second per layer when printing at 100mms.

[gr5 2,173]

The issue seems related to aborting print

 

What issue? You didn't say anything about leveling earlier and none of the pictures show levelling issues which only affect the first layer. I'm confused.

But anyway, yes there seems to be some leveling issues where the level isn't properly mainained.

leveling bug on um2 patch is here:

http://umforum.ultimaker.com/index.php?/topic/5935-z-axis-homing-inconsistent-on-um2-workaround-and-patch/?p=57095

 

[pm_dude 27]

oh sweet!

Here is the model/gcode I'm printing

https://www.youmagine.com/designs/benchmark

 

[pm_dude 27]

@gr5 I didnt mention any leveling issue before because it didnt seem related before. And in the end it still isnt related so you can ignore that... my bad

From what I see when looking at it printing it really look like it doesnt respect the minimum layer time (8 sec in my case)

 

[pm_dude 27]

It messup even at 20mms

 

[gr5 2,173]

This (bad tops of prints) is a common, well understood problem. Most notably on the robots antennae. The min layer time works fine. The problem is that the top point is so damn small that slowing down the print doesn't help because the head is touching the top constantly and constantly keeping the PLA in liquid state. Like a soldering iron on solder.

There are 2 common solutions.

The first one is the simplest - you just check a single checkbox under cooling called "cool head lift". This makes a huge difference but has a drawback. While the head is cooling for most of those 8 seconds it leaks (even with retraction it still leaks slowly as the PLA higher up the tube heats up and expands and slowly forces out a string). Then when it goes to do the next layer the string gets attached to the side of your part. It is not too bad to cleanup and it is a huge improvement but you can do better.

The second solution is to print something else just as tall. Or 2 things. So for example you could print a 10mm-on-a-side square (or round) tower on either side of the print. And then throw those away. Or you can just print 2 of these pyramid "benchmark" things. Or 3. Or 4. Two might not be enough.

Also it helps (not as much as the above 2 things but every bit matters!) to print at a lower temperature. Say 180C when you get to the tip of the pyramid. You can use tweak-at-Z to do this.

I think the best compromise is to just print 1 towers on either side. Make sure they are at least 1 layer taller than your print. Make sure to use "brim" as towers that skinny can get knocked over without brim.

 

[pm_dude 27]

Ok that makes sense. Thanks for explaining more in details. I know illuminarti gave the advise of printing multiple one at once too. I had tried with a poll on the side but didnt got much difference since it was so small