Partially Solved Test print failed.

Trying a test print of "Car Grille Fastener" found on thingiverse, I had a failure.

Part of it is attributable to the "nut" part not staying attached to the bed, as I found it off to the side when I checked on progress, but also noticed it has gotten "thready" for several layers.

Also found it odd that there was a thing "thread" around each object with seemed to serve no purpose. I will venture a guess this is supposed to be part of this "skirt" mavens speak about, but . . . I dunno.

The "calibration circle" had the same thin thread surrounding it and now, looking over the settings in MC again, it appears this is normal when "raft" is not selected and I should probably enable "raft" to assist with adhesion.

I would appreciate any input on this and as to why the "nut" part of the print became "thready". Could that be it was a bit loose and continued to get looser?

Pictures attached.

I did have to fiddle with the Z Offset ending up at 1.9mm. It was at "zero" after installing the firmware. I understood the default was 1mm. This eliminated the grinding of the filament, but was discovered too late ot prevent some marring of the bed surface.

 

Thanks, I had seen that, but it is clearer now.

Question. How does "auto leveling" do it's thing? I have watched the gyrations before each print and have an inkling, but, need to clarify some things. How does the firmware decide where Z "zero" is? by lowering until one of the Z axis limits is actuated? Actuate each and work with that, or an average of the two. Or three, if it also tests with carriage centered left to right?

Then, does it use the Z Offset to place the nozzle above the bed?. Further to that, if I add painters tape or some other adhesion promoter, does the zeroing or "self leveling" make that surface the new zero height?

I hope that was clear.