3Dlabprint part not showing internal supports, too thin ?

[charentejohn 0]

I have one of their models, P47 plane, and all wing sections are ok and print the internal walls, apart from the first, it has nothing showing when slicing or printing.

I am new to 3d so can't think what is wrong.

Is there a way to 'enhance' the internal supports ?  I can see such things as inner walls line width but changing it makes no difference.

To be honest not sure what the difference between a shell and a wall is ?

 

Anybody any idea what settings I need to make these appear ?  I did an infill of 3% and that shows ok.  I guess they have made the walls really thin, so how do I get Cura 3.1 to see them.

Any help appreciated/

[DidierKlein 729]

Hi and welcome,

 

Can you post some pictures maybe? Or the file?

[drayson 75]

Hi!

 

I recently printed the Edge540.

Things I leaned so far:

There is no Support needed so please be sure that "support" is disabled. Also, you do not need any Infill...

Those parts are thin/single-wall-prints

There are print profiles provided from 3dlabprint (but for older Cura), which need to be imported in Cura. They bring all the relevant settings so you just have to modify filament diameter and nozzle diameter.

Please also crosscheck the line witdh you selected and compare it to the nozzle diameter - could also be a root cause.

Further, I learned that if I place more than one section the same time, the internal structure of them is likely not recognized correctly.

 

And finally, as Didier mentioned, please share some screen shots :-)

 

[charentejohn 0]

Thanks for the quick responses, I have tried various settings and then went back to the one supplied by 3dlab.

This produced the slice in the photo.   At least the first layer shows up ok and the first two supports on the left of the photo.  All the rest only print once.

I am printing one at a time.  I did disable support and then put it in at 3% which added supports I could see, so in general it works.

Could be the curves of their internal supports causing a 'stepping' of the slices and so too thin ?  Are there any settings that would make the inner supports thicker, the part is seriously underweight so no problem.  

[charentejohn 0]

Where I said support at 3% I meant infill, sorry but I was thinking of their supports which are perforated and quite thin.

[DidierKlein 729]

it could be to thin indeed, there is a setting called print thin layers that you can check maybe

 

Or change the line width just slightly to see the differences?

 

Or i would advise to contact the designer maybe he knows the trick?

[charentejohn 0]

Thanks, still waiting from reply from 3dlabprint, they can be a bit slow.

Checked print thin walls but still missing.

Actually I was just playing about with infill to see what happens, turns out that 'quarter cubic' infill at 3% gets the weight spot on.  So failing any better solution from 3dlab I will use that.

Looks to be a square grid inside the wing section which should give good support.  Better than the original which you can see from the photo are perforated and wavy, may be some engineering reason but the cubic one looks solid to me.

 

One last question if I can, if I use the quarter cubic is it possible to print the infill thinner than the 0.4 nozzle size so that I can increase percentage for the same weight.  Basically more infill supports for the same plastic ?  Probably a step too far as 0.4mm will be very strong.  I can print the rest as per the plans.

[kmanstudios 1,120]

I have noticed that a lot of those files are just quick exports out to STL and not making them really water tight. Just because the model can go out in a format, does not make it a properly printable file. Most of those programs are outputting 2D sheets and not manifold 3D objects.

[charentejohn 0]

They do send gcode files but they are for a Prusa printer, they may be better but do not work on mine, go too fast and too hot.

May try again and use tuning to dial some of it down ?

I can see the internal supports are curved and perforated, and obviously very thin, two walls with a gap in the middle.  A real stringing test, means they print a very thin wall jumping from side to side of the perforation then on to the next one.  Means loads of stop start, I think the infill from Cura will be better in many ways.

If I can I would lose their supports in the other sections and replace with infill.   I guess the fusilage will be better as it is basically cylinders with vertical supports in them.

 

[kmanstudios 1,120]

1 hour ago, charentejohn said:

They do send gcode files but they are for a Prusa printer, they may be better but do not work on mine, go too fast and too hot.

May try again and use tuning to dial some of it down ?

I can see the internal supports are curved and perforated, and obviously very thin, two walls with a gap in the middle.  A real stringing test, means they print a very thin wall jumping from side to side of the perforation then on to the next one.  Means loads of stop start, I think the infill from Cura will be better in many ways.

If I can I would lose their supports in the other sections and replace with infill.   I guess the fusilage will be better as it is basically cylinders with vertical supports in them.

 

Then maybe the walls are too thin for proper sampling.

[tinkergnome 927]

Have a look at this thread, and please report back, if it makes a difference

 

 

[charentejohn 0]

Thanks that looks promising, off to bed now but will start again in the morning.  I used the 3dlab settings again and supports reappeared but their settings give line separation, too fast I think, so I will need to 'tune' down as print starts.  A lot of options to tune in Cura, I think I have selected something that has done this, that said the infill looks like better support for wings (not fusilage) so I am keeping the profile that ignores them  

[drayson 75]

Suggestion: could you provide the file of one section which makes issues to me?

I´ll give it a try with my settings and of course share them to you... What do you think about it?

[ahoeben 1,948]

Try slicing with these two settings from the Mesh Fixes category:

 

Union overlapping values: unchecked
Merged meshes overlap: 0

 

You can use the "Search" box on the top of the custom settings.

[charentejohn 0]

Hi, photo attached of the 3dlab settings.  I tried the meshes one,both were unchecked, checking did not make a difference.

With this I can see the supports inside, you can see them on the photo, but nothing has changed in the list of options in cura.  I went through them all and the only difference was this one had some dual extruder settings which I guess would be ignored as I only have one. 

 

This has been useful as I think I might use the cura infill in place of the original, looks much better.  Like I say I am no engineer but the perforations in the internal supports are a pain to print, rather than just full lines.  Also some are very thin in places, I can see they may be designed for flexibility though.

I will need the main 3dlab settings to print the fusilage though so that sould be ok.  Like I said their speed causes 'mesh' type layers at about 3mm high so needs to go slower.

[ahoeben 1,948]

37 minutes ago, charentejohn said:

I tried the meshes one,both were unchecked, checking did not make a difference.

 

Just making the settings visible or invisible does not change anything. You have to make them visible and then change the values in the sidebar.

I am pretty sure this works, because I have also tried printing a plane from 3Dlabprint. They are pretty shitty models, with many single-wall parts that confuse slicers.

[charentejohn 0]

I read where somebody said 3d printing is a hobby within a hobby, good observation.  I will be a while between relies as  (as we all know) prints can take a while.  I have a large pile of 1cm high sections in various states.   I am trying various bits and changes, one problem I have with cura is saving a profile so end up writing them down.  I know quitting seems to save it but when I say keep changes I keep getting it wrong, so I have one nearly working then have to start again.  A save profile would be easier to understand ? 

 

Single wall printing is tricky, 3dlab said so but no real solutions from them for other printers. I am having adhesion problems, some lines are just falling apart despite being 0.2mm so I may have to up the temp I think.   The wing part I am printing has a 'cone' protrusion on it, as in photos, and no way will it print without laminating, I may just give up and reinforce it.

The 3dlab interior supports are just a set of cobwebs.  To fix it I think I need to really up the temp and set layers to 0.15 or less.

I will try their Prusa gcode now and use Tune to dial back the temp from 230c but leave the rest, wish me luck.

 

I will e-mail them and ask if ok to put the wing.stl file here, can't see why not as it is just part of the plane.  Be interesting to see if anyone can find a good profile.

[charentejohn 0]

Just watched the 3dlab direct gcode print working and I think that would be ok.  I may try it fully, unfortunately it is printing to the far left of my print bed and I hadn't put tape on that.

What is interesting is their codes, looking at the photo with the various sections, it prints the interior bits first, so all the little boxes in the first layer are printed with a gap between them.  Then it goes all around the outside making the wing shape.

So is it possible to make cura produce this ?  Is this sort of 'inner walls first' then outer walls.  The .stl file sees internal and external walls as the same, internal are not fillers but part of the main structure.

Added photos to show structure, internal 'boxes' are very thick, way over what I had.  So prints, skirt, boxes, outline, fills in box joins (as in P47 gcode06)

Edited February 27, 2018 by charentejohn
To add photos

[charentejohn 0]

No reply from 3dlab, as usual so here is the problem .stl good luck as I think you'll need it.

I have posted a request on the 3dlab forum to ask for settings as the ones they provided just don't work.

wing_L1.STL

[ahoeben 1,948]

I ampretty sure you are in violation of their end user agreement by posting this.

[kmanstudios 1,120]

OK, I am not sure why you posted the file. I took a look at it, did what was suggested in another thread about this and it worked fine, that is to turn off mesh fixes such as union overlapping volumes. Otherwise, I am not sure what you are looking for.

[charentejohn 0]

Drayson said he would look at it and let me know any results.  I looked at the other thread and saw the settings mentioned.

I have just realised that if a setting is not displayed on the right it is not just not being used but is set to some default or other.  As you can see I am that new to this, not so mych a learning curve as vertical takeoff

I located the Merged meshes overlap and set it to 0 and the supports appeared so thanks that worked.  

 

One thing I can't seem to do is make it print like the 3dlab one does, you can see it printing the above formats here, it prints the boxes then goes over them, so prints the boxes then the outer shell is printed on top of them, then it does them again.  A bit like two prints in one, does one layer boxes and one layer shell then repeats.  No idea what this is or how to do it.  I have tried 0.4mm shell and wall count of two, just does and inner and outer and massively heavy.

Could be they have not given enough info in the.stl to do this. 

I am just trying to do a clean(ish) print.  Now I can see all walls again I will retry but have a feeling that the internal supports will noot print properly but I can play with high temps for that ?  3dlab temps are massive 230c and 60c for the bed, hence in the photos very liquid, I was trying to avoid that as seems a bit high.

 

 

[kmanstudios 1,120]

different printers, different slices, slightly difference in print procedure. Whatcha want is a good print...not 'their print.'

[charentejohn 0]

I am sure you are right, currently reimported their 'starter' settings and will play with temperatures.  I can see loads of people on youtube who have printed these perfectly, and I would like to be one of them     I will try a few more variations, at least the merged meshes overlap to 0 worked.

I will try my amended print later, I have now figured out how to save my settings so have tidied up the dozen or so failed profiles, progress at last.  Thanks for all the help so fa, hopefully won't need much more. 

[charentejohn 0]

Very close now, all excellent apart from the angled interior supports (perforated curved bits) which are fine for near vertical but some start at about 45 degrees and print as a series of unconnected strings.  The corner with the half cone as it's stat point (see photos above) will never be right I think, some adhesion but weak.  The normal outer walls are just about perfect.
Set at 0.2mm layer height and print temp 215 bed 60.  Tried the 2.5mm recommended and the curves and slopes are worse, as expected. 

 

Question is is there a setting to fix the stringy slopes ?   I am sure I can just reduce layers to 0.15 or 0.1 but should be ok at 0.2 really.

Part of the problem is the whole thing is considered one piece, so no infill as such to mess with.  Could do with a do sloping bits twice command   or is there a way to print the first 0.5-1cm at a different layer height ?