getting supports to "stick" better

[jaysenodell 17]

I've had this problem with trees but have been able to ignore it until now. Basically they don't "stick" and get ripped off. Not all of them, but generally "the same ones" in mulitple runs so it has to be a setting thing. In the attached file its the trees supporting the roof on the X/Y+ side of the eraser to the first internal support (screw hole to brace). 

 

What settings get tweaked to improve the tree "stay-put-iness" without making them impossible to remove?

sink_front_clip_04_PETG_Standard_Quality.3mf

[GregValiant 1,330]

Increase the "Branch Diameter Angle" to around 20° and make the "Initial Layer Diameter" around 20.

That will give wider bases.  The model is short so the trees have to get wide faster(more angle) than they would on a taller model.

Here the brim of the supports and about 12mm in diameter.

 

This with the default settings.  The support brim is only about 8mm in diameter.  That's kind of a shaky base.

 

Enabling Z-hops at about a 0.5 hop height can help as well .  One of the reasons the supports get knocked over is when there are pips sticking up and the nozzle whacks them as it travels.

 

 

Edited May 22 by GregValiant

[GregValiant 1,330]

This is a lot like stealing a thread - only different.  Your model is a good one for this.

 

You know I dabble in post processors.  I had a couple of prints with large flat supported areas and the finish of the first layer over the support was pretty bad.  I though that if I could emulate a dual head printer and put down the interface with a different material that the surface would be much improved.

That turned into "Support Interface Material Change".  It is sort of like Pause At Height but instead of pausing at a height or layer, it pauses at the "Support-Interface" on a (user selected) particular layer(s) and you switch to a non-compatible material, and then there is a second pause at the end of the Support-Interface and you change back to the model material.  That allows the top layer of the interface to be a "ain't gonna stick too good" surface.  For a PETG print, using PLA for the support interface works well and the opposite is also true.  The finish of the first layer over the support is outstanding and compares well to what is achieved with a dual extruder printer using PVA for the support.

 

The Cura set-up needs to be just so.  With a "Support Top Distance" of 0.0, and and the Top Interface at 100% density of "Lines", the support comes right off.

What it doesn't work well with is horizontal holes because there are so many pauses required that it's just annoying.  For flat surfaces - so far it has been terrific.  Yes, I am patting myself on the back.

 

Below is your project with my changes for both the support, and to be compatible with the post processor. which is also attached.  It will go into the Scripts folder in your Configuration Folder.  I haven't tested it with OctoPrint.

 

On this model the "Layer of Interest" is layer 82.  Searching the gcode for "Custom" will turn up the inserted commands and you can see what they will do.

There are auto-unload and auto-load options.  Both amounts will need to be tuned to your printer (setting to 0 will disable them and you can pull out the filament and replace it by hand).

Purge amounts need to be sufficient to clear the previous material from the hot end and nozzle. I guessed at 40mm for your project but it will need to be tuned for your printer.  Insufficient purging can lead to a layer of mixed materials and it will be a weak point.

If you don't want to fool with this then just disregard.  I'm looking for a sucker beta tester and your model is a good choice for this.

 

GV sink_front_clip_04_PETG_SuptIntMatlChange.zip

 

 

 

These are the settings for the script.  If you have questions post them here.

 

This is layer 82 which is the last layer before the roof goes on.  The supports have been adjusted.

SuptIntMaterialChange.zip

Edited June 15 by GregValiant
Update to SuptIntMateriallChange.py

[jaysenodell 17]

Oh I’m a sucker ^H^H^H^H^H^H^H willing beta guy. On the purge front I have (dis)advantage of a e3s1p from C-word which means ru out sensor and direct drive extruder. That SOB purges all in its own where it wants, the second I hit “continue” on the stupid proprietary interface. On the plus side, I don’t have to think about how much to purge. On the down side I have to zero all that out and let it do its thing or I wind up with booger string all over the build. 
 

which is my long winded way of say “hot damn! I get to play with a new thing! Thanks!”

 

I started a print of the clip before your replies so it will be a few hours before I can get the loaded. I’ll deal with the ugly but will happily run this for the next set. 

[Slashee_the_Cow 382]

Just don't use Greg's script on small internal areas which require both a floor and ceiling like this:

Trust me on this one.

[GregValiant 1,330]

Yeah, that's 8 pauses if you wanted to do all of them.

 

Something like this is what I was aiming to improve.  A box shape that needed to be printed in a certain orientation.  This one has a large flat roof on the inside so there would only be two pauses.

 

[jaysenodell 17]

On mine I think it is more like 12 pauses. But I think I can reduce it by pushing the trees to be "on build plate when possible".

 

Nope leaves 2 inside. those I can block out as that gap is bridgable. 

 

Mrs banned me from overniggt printing last night so I'm just getting it rolling this morning. I'll let you know how it goes. 

[jaysenodell 17]

Gust getting this started. But to be clear, layer 82 is the start of the PLA? Seems to me that the roof is 78 and 82 is the start of the wall. Does the script count backwards? I have read the code because of the pesky need for income and those people that keep demanding that I answer my phone...  shall look at it now before I hit print... 

[jaysenodell 17]

I think I see what you are doing there.. (in the g-code). This should be fun! Printing!

[GregValiant 1,330]

The "Layer of interest" is the top layer of the support interface as shown in the Cura preview.  The "Support Top Distance" should be 0.0.

Edited May 23 by GregValiant

[jaysenodell 17]

@GregValiant, I loaded your file with no change other than disabling the purge feature as this thing will do that on it's own. Other than my normal petg not wanting to stay down problem (which is odd becuase I had this dialed in). on the possitive side, all supports are stil attached and nothing is sliding around the plate like a canadian living near a pond in winter. the print is also much quieter than normal to. no servo thrashing. 

 

So what did you really do differently? 5.7 said 800+ changes but I can't find a way do compare projects side by side as it closes one project to open another. Is there a project comparison feature I can use to compare these and see how your settings are different than mine? 

Edited May 23 by jaysenodell

[GregValiant 1,330]

Here is "Add Cura Settings".  It adds a goodly number (around 400) settings to the end of the gcode file leaving a record of a lot of the main settings.  It was turned down by UltiMaker because it is hard to maintain.

 

AddCuraSettings.zip

 

In regards to that project - I didn't change a lot other than a couple of support settings.  Speed, accel, and jerk should be whatever you had.

 

Edited May 23 by GregValiant

[jaysenodell 17]

Well then... it might have been the internal support trees that were the noise. other than the fans, this thing is almost quiet. It isn't this quiet when doing the big tubes (giant ovals 8" tall). 

 

For those that stumble on this and are mac people. Download your GregValiant scripts. Open terminal (cmd-space and type terminal hit enter). Then do the following

jaysen@mba13 Downloads % cd
jaysen@mba13 ~ % cd Downloads 
jaysen@mba13 Downloads % cp AddCuraSettings.py ~/Library/Application\ Support/cura/5.7/scripts 
jaysen@mba13 Downloads % 

The key thing is that the cura config folder is Library/Application Support/cura/<version>

Thanks Greg. I have clearly added that to my list of "why the fork didn't I have that before" tools. Put it in 5.6 since I still use that too. 

[jaysenodell 17]

Well now. That was fun!

 

The only issue I had was nozzle drip resulting in no booger string post resume. Here you can see the missing chunk of pla as I've started to load the petg.

 

Yelow booger string is going down now. Seems happy. Stil don't know how this thing is so quiet. Maybe I'm losing my hearing.... I am old-ish. Maybe I started the rum at 8am. Who knows...

[jaysenodell 17]

Well now... 

 

That was fantastic! Not only did the supports stick but the interface is better than my build plate or the top surface. I'm not considering rafts for everything with your little post processor as the secret sauce to get flat bottoms. 

 

As to the actual issue, I think your original "wider bases" was the answer. I actually had to scrape them off the plate. I'm marking that as the answer. Thank you!

[GregValiant 1,330]

Ha!  Blind squirrels and nuts young man.