yeah, i’m leaning towards some kind of flow issue, maybe temp related? in the last pic you can see some roughness and delamination:
i just wanted to jump in and throw out some other ideas since it seemed like everyone was getting to the point of ruling out machine settings. but your idea of a partial clog could be the culprit as well.
Hello!
I also believe that I might be doing something wrong in the slicer settings. I initialy started palying around with E steps and k factor as I noticed these inconsistences, but both before and after the changes, there were similar issues.
I wait to see what this latest print results will be, as I recalculated everything.
In any case, I use Cura. Below you can see my settings. I snipped everything, just in case! 
I also use this filament currently --> https://gembird.com/item.aspx?id=10368 , but I tried with other too, with similar results.
Not much is standing out. Personally I’ve had issues at .12, so I print at .16mm layer height.
Similar effects were also to 0.16 and 0.2, but as I mentioned in a previous comment for some reason my E-step was reset to the original value… So lets hope that with the print that I currently have going, there will be some extra clue to help find out what is wrong.
It seems like surrounding geometry will affect the line appearance as a general rule - is it possible to fine tune it away? probably, but i think to a degree thats just part of 3d printing too… i think it looks pretty good to be honest, at least compared to the original prints you had shown.
You can tell that the lines look different where the tapered bottom meet the beam and then again where the tapered top beam meets it again.
Perhaps a tighter infill would help it some? Maybe if it was a little hotter it might hide it better too?
I don’t know that it will get a whole lot better than that, the material surrounding the extrusion on all sides will affect how it lays out .
Perhaps if you reduced the cooling or had it in an enclosure the temp would remain consistent longer and help to hide it as well?
The first print I posted had more visible problems, that is why I started with that. It is a lithophane so layer to layer can change a lot, and that really helped me to figure out if those lines had a pattern.
Printer is in an enclosure (not the original, but still it is an enclosure), so there should not be any big temperature variation during the print.
If I notice anything new, I will update! 
So the print is finished… Same gcode, but different E value. Result: effect enhanced, as there is more flow… Look below.
In the second pic, the one on the top is with default E value.
Also I was checking regularly about the nozzle temp and it seemed not to have ups and downs.
Unless I missed it above, would you post your model and the gcode?
I suddenly think it’s a gcode/model issue.
Should be simple to clear up.
No you didn’t miss them, I never uploaded those.
Here is the stl
Lower_Frame.stl (146.1 KB)
The gcode file is too big for me to upload it 
I’m mostly interested in the gcode. Can you put it on google drive with a link? Or some other service?
I know the forum here won’t let you upload a .gcode file you have to zip it. Even zipped it’s too big?
Lower_Frame.zip (1.9 MB)
zipped it is!!
The model does have a couple small self-intersections:
But the gcode looks like it dealt with it just fine.
There are some noticeable changes in the toolpath as it transitions in and out of the region that is overextruded:
Because the bridge is solid there’s no place for any overextrusion to go, like there is with the sparse infill.
This structure matches the internal structure of the print, where the bridge goes from solid, to infill, back to solid:
I think you need to take a break from this print specifically, and do some tests with something that has a similar shape, namely solid infill that’s long and thing. Maybe even a smaller piece of this model.
I have a prediction that if you print a small 20mm cube at 100% infill it will be horribly overextruded. I still something is wrong with the slicer. Perhaps it’s infill overlap at the wall-skin boundary, that’s something I had to tweak in my settings.
The thing is this is not happening only in this file. Nor a bridge is always the cause of it. I didn’t notice something specifically there.
If you check the below image, you will see to main points where I have this inconsistency.
One where the bridge ends and one where the 2mm deep and long semicircular hole starts.
That’s here?
I don’t know if you’re interested, but here’s it sliced with my settings in S3D:
Lower_Frame-S3D.zip (642.3 KB)
It seems you’ve tweaked most of the knobs already with little difference. These are just my default settings at 20% infill, .16mm layers.
is it possible that there is a little backlash in the linear modules that is contributing to this?
My settings compensate for 0.035mm of backlash at the start, that would partially test for that too.
The official backlash spec’d value is 0.02mm, I think I’ve since changed the start gcode to only compensate 0.02 but I haven’t updated my S3D profile, so that one was sliced with 0.035. Oh well. Easy to change, it’s the M425 X0.035 Y0.035 Z0.035 F1 S0 line at the start, change all the 0.035 to 0.02
Although my X axis has developed a more severe backlash, it now gets 0.11mm :(. Something needs to be tightened.
