Odd failure on top level; diagonal stripe of loose filament
I have had this happen twice on slightly different versions of this model. In the slicer I see no artifacts at this layer or the ones under it; I've looked at all the features. I'm at a loss to explain what happens here. The model is about 30mm tall and the bottom is 3mm thick, flat on top and bottom. The problem is on the top layer(s) of that bottom. I am using layer height set to 0.3mm from the bottom layer through all of these layers up to about 27mm. Any advice is appreciated. 
Please save your project as a .3mf file
Files > Save Project as
Zip the .3mf and post it here. It will contain both your part and your settings for us to diagnose.
Cheerio,
RE: Odd failure on top level; diagonal stripe of loose filament
Hmmm. I did attach it to my post. I'll do it again to this reply.
Jim
RE: Odd failure on top level; diagonal stripe of loose filament
Oh, does this forum software silently fail if the file is not .zip? Now trying with a .zip of the file.
RE: Odd failure on top level; diagonal stripe of loose filament
Bingo 🙂 Yeah the forum drops anything not on its allowed types list. Which seems to just be some picture types and Zip files.
Are you sure that's the right file? It's sliced for PETG with a modified 0.20mm STRUCTURAL preset and a little variable layer height.
Cheerio,
RE: Odd failure on top level; diagonal stripe of loose filament
Yes, that is correct. I modified the layer height to be 0.3 from as close to the build plate as possible, then up almost to the top. At the top are some features I wanted to render in 0.2mm. Does this give some clue to the failure?
I don't have a Mk3.5 to trial print ... (anyone?) ... but it appears that, due to the variable layer height, the top layer of the floor of the box is being printed at 1.99nn mm - the nn represents decimals I cannot read - yet the floor thickness is defined as 2mm (not 3mm) - theoretically you are asking for an impossibly thin layer but the slicer defaults to a sensible value ...
I don't see why or how the value should vary across the width but you might have hit one of those situations when decimal and binary align poorly; I'm clutching at straws here but I would be inclined to start the variable layers above the floor 'though I can see nothing here to justify them, ordinarily they are used when a few layers would benefit from higher Z precision not as a stretching mechanic.
You might do better to start from a 0.3mm DRAFT preset and closing the layers where you need extra detail.
Cheerio,
RE: Odd failure on top level; diagonal stripe of loose filament
Thanks for the ideas. In the 3.5 there is no 0.3mm draft setting. 0.2 is the thickest unless you do what I did and go to variable layer height.
I'm not using variable layer height to stretch, I'm doing it to speed the overall print time. I only need the 0.2 detail at the top of the box, so printing the rest at 0.3 is just for speed.
You have a 0.4mm nozzle configured, 0.3mm layers are about the limit, for ordinary work I would stick to 0.2mm - the SPEED profile will be quicker than your modified QUALITY.
Cheerio,