Ugly step in outer shell

[Coni 1]

Hi there,

I printed my own race quadrocopter frame. It looks and works perfektly for me, but I have one problem. The outer shell always has this ugly step. It is not part of the cad design neither I can see the step in the cura layer view. So how can this happen? It shouldn't be there . .

Best regards

[SandervG 1,521]

Hi Coni,

It is a bit hard to tell from the pictures, is it more like a different shade of colour, is it an underextruded layer or is it more like a dent / notch?

If it is the latter, I am inclined to think it is in the model.

If it is underextruded there was most likely a brief moment of friction on the reel or in the bowden tube tube but it recovered. Colour gradient can happen because the temperature changed or something in the ingredients of the filament was different at that spot.

Have you printed this more than once? and did it occur at the exact same place?

[Coni 1]

Hi Sander,

I already printed this 4 times and it is ALWAYS at the same spot. But it is definately NOT in the model. Also it is not a color issue - one can feel the step. Aktually it is even not like a notch or dent it is a offset. If you look at the first picture you can see the bottom of the print. So it starts printing slim and then we have a offset to slightly thicker - but on BOTH sides of the objekt.

Very strange I have no clue. It is just at a very certain spot where the printer has to print the bottm for the battery housing.

[ignatius 23]

Did you modeled it in two parts?

[gr5 2,172]

Coni you need to be more clear. I read your text 3 times and looked at the photos and I think Sander is talking about something different - but not sure - there are 20 things in the photo you could be talking about so it's tough. However I think you are talking about the brim. You are supposed to peel that off after the print is done.

Maybe add an arrow pointing to this "step" in all 3 pictures? And the lack of the step because the cura photo has a completely different angle and the brim is the only thing I can think of that is in both photos.

Without the brim your part won't stick to the glass as well and may lift off the glass about 5mm into the print and will ruin the print. I usually use a sharp knife to remove it carefully (don't cut yourself).

[gr5 2,172]

Reading text and looking at pictures 5 MORE TIMES now I think you mean the line that sander meant? That looks like typical underextrusion from when the filament is tangled but wouldn't always be at the same height. It's probably something related to the model - maybe you should post the STL. when the printer switches from clockwise to coutner clockwise due to changes in STL file sometimes you get weird shifts due to backlash or play. That is usually fixed by adding oil or tightening belts or loosening belts or loosening end caps.

If you are indeed talking about that thing then your cura shot seems to be hiding it - better to look at it in cura from a different angle.

[Coni 1]

Sorry to be unclear,

indeed I'm talking about what appears as a thin line. The point is that you can't see that line in the stl nor in the sliced cura view - IT SHOULD NOT BE THERE

[gr5 2,172]

I'm guessing your z screw has a glitch there such that you are underextruding for a few layers and then it recovers and prints normally after that - the bed moves too far maybe. I would check that the Z stage isn't hitting anything - grab it with both hands on either side of the back of the stage and lift it up (with power off) and push it back down again. Try to feel if something happens somewhere at that height. Also try printing a simple 1cm x 1cm X 30cm tall block. Hollow. See if you get the same line on that part also which should only take 10 minutes to print.

[Coni 1]

I'm guessing your z screw has a glitch there such that you are underextruding for a few layers and then it recovers and prints normally after that - the bed moves too far maybe. . . .

 

Well now I can say that it is not a problem with the z-axis. I modified my object. I lowered the bottom for the battery housing and the ugly step just moved with it. See it in the picture:

Is it possible to analyse the gcode if the step is already in there?

[SandervG 1,521]

Yes, if you open the g-code in a text editor you can count the E steps up onto where this happens. You have to measure on your print where this happens too.

Could you make the file available in a dropbox so we could check it out too, if we see anything out of the ordinary?

[gr5 2,172]

There are a few really great gcode viewer programs. I recommend repetier host - it's very easy to use and you can highlight just one layer at a time and see the movement through that layer in another color. It's also free, and a quick small download.

[gr5 2,172]

I'm confused - is your "step" always the same distance from the print bed (indicating hardware error)? Or did the "step" (#2) move upwards to a new spot (indicating something more complicated interaction with the model)?

If #2 then it could be a clockwise versus counterclockwise issue that shows up errors caused by backlash. This is common on flat surfaces but never on curved surfaces that move out and then back in or vice-versa.

[gr5 2,172]

Also the red arrow in the photo above seems to be pointing out some underextrusion. Typically when it is just one or two layers it is caused by a tangle on the filament spool.

[SandervG 1,521]

Actually, @Gr5, I thought the arrow pointed at an edge/boarder coming out of the print, instead of an underextruded layer.

Is it in- or outwards?

[Coni 1]

Actually, @Gr5, I thought the arrow pointed at an edge/boarder coming out of the print, instead of an underextruded layer.

Is it in- or outwards?

 

Hey,

sorry for the slow reply - I didn't get any notification.

It is like the title says - it's a step outside - it's not like a groove. In the old version it was at a height of 23mm - in the new version where I lowered the battery housing inside the body for 3mm the step is now at 20mm. Measured from the bottom/bed.

Here is the gcode. Password: UMF

Thanks for your help and thoughts guys,

Coni

[Labern 774]

have you looked at your model in x-ray mode to see if there is an internal edge or face?, will show up red.

[gr5 2,172]

OH MY GOD! You should have posted that a month ago!!! This changes everything.

In layer view the printer stops on layers 31,32 do not print - nothing happens - it just skips from layer 30 to 33. No wonder! So the problem is in your STL model. There is a hoel in it somewhere such that cura can't make a complete loop so it doesn't know what to do with all the "random" line segments it has for those layers and so it prints nothing.

if you look at the part in xray view as labern suggests you will see a red spot somewhere in those 2 layers. You need to fix that spot. Possibly you can fix this with the "fix horrible" settings but those might fill in holes that you need such as vertical screw holes so I don't recommend that.

Also you don't need "support". I would turn that off - that's a waste of time and filament because it is supporting a surface that is "bridged" anyway and the printer is pretty good at doing bridging - especially with PLA. I mean the support doesn't hurt anything but it doesn't help either.

[Coni 1]

In layer view the printer stops on layers 31,32 do not print - nothing happens - it just skips from layer 30 to 33.  No wonder!  So the problem is in your STL model.

 

I cannot confirm this error in layer 31. Plus the error I am looking for is at a hight of 23mm so it should be found some where around layer 230. I've also checked the stl in x-ray view and there is no error as well :/ ( See the Picture)

I would have preferred that error - because now I'm clueless again . .

Best regads

[cloakfiend 995]

Its a model issue as its scales. Its the only possible reason. It will become clear if you show me the wireframe view in a 3d app. One that shows all polys that is, not some short hand view. Ive seen some seriously bad stuff, and its usually blender and inexperienced modelling thats the reason. You MUST model clean shells to get perfect prints, or else weird things happen. I wrote this a while back but could not be bothered to type it all again as i was timed out after a lengthy explaination. Avoid all intersections, and if you are going to have them, then zoom right in and make them as minimal as you can and youll be fine, booleaning also doesnt always create a union object and simply hides the intersecting faces inside. Check your model.

Also dont use the new cura, use 15.04 or 15.02.01.

Edited October 21, 2015 by Guest

[gr5 2,172]

Well look at the gcode file you posted. It's subtle - when the yellow traces are missing on a layer or when they are faded in color (darker) that's because it's still showing the layer below but hasn't printed anything on the current layer. You can also see that the red sides have gaps every 5 or 10 layers - somewhat regularly. Something very strange is going on.

Near layer 230? Well it skips over layers:

210,215,225,234,248,265

And many other layers are skipped.

Another clue is that there are zero blue lines on the layers that it is not printing on.

[gr5 2,172]

Hmm - okay well I take it all back.  I looked at the gcode in repetier-host and it looks fine.  There must be some bug in Cura when viewing gcode files.  :(

It looks fine at 23 mm- I don't think it's a slicing issue - probably a hardware issue.  Either underextrusion or Z platform moves too far.  Can you measure the printed height to see if it's off by .5mm  or whatever the gap is?

Edited October 25, 2015 by Guest

[cloakfiend 995]

Can you post a wireframe view of the model in tris not quads? To verify its not a model issue? But if you know your modelling then it must be the slicer.