Bent glass?

[Blizz]

Since a while I've been having this issue:

As you can see, the buildplate is calibrated (albeit somewhat close to the nozzle), but the left and right +4cm of my buildplate seems to be bent down, as the 2 shell lines there no longer touch. (Please ignore the area in the green circle, that was me playing around with the calibration screw)

Anyone ever saw this/anything I can do about it?
FYI: I can't reverse the glass as the bottom side has a nick in it already and I really don't feel like ordering a new one from UM.
My last order with them didn't go so well...

[LePaul]

I wondered if I had something like that going on too when I first installed my build plate

However after a lot of careful calibration, it seems fine

What happens if you try something lowtech...like but a level on it. Does the bubble move slightly one way when placed in the middle?

[Amedee]

Glass plates are often not perfectly flat, but that seems extreme to me as on the middle it is way high, and then very low on the side.

It it is the glass plate you should be able to see it with a ruler, but I would rather think of a bent axis in the gantry.

[Neotko]

I had my bed banana, not the glass, on my umo+ assembled by imakr. I bed theybended it on assembly. Anyhow, after I found the spot I did press it on the oposite direction (without the glass) and after 2-3 adjustments my bed now it's flatflat. It can happen and sometimes it's the glass or the bed itself pushing the glass. You need a really flat thing to locate it. For example one of the x/y shafts (if they ain't banana).

Also check that the problem isn't the x/y. Check if the sideblocks wooble up/down when moving the head.

[Blizz]

@LePaul: I've calibrated it completely (I use the gcode that draws circles. The circle lines are all perfectly the same width and there's the same amount of place between them. The problem starts as soon as the head moves past the calibration screws as mentioned in the OP.

@Amedee: Haven't had a single issue like this for the year and a half I've had te printer. It just started recently. I'm a bit afraid it might be an axis too.

[LePaul]

I didn't have a lot of luck with the circle gcode, since it does it at 0.1mm And I was using silver filament...even harder to see!

[Amedee]

@Amedee: Haven't had a single issue like this for the year and a half I've had te printer. It just started recently. I'm a bit afraid it might be an axis too.

Glass won't bent, so if it used to be OK in the past it is more likely the axis

[Blizz]

@LePaul: I generated my own versions for my different nozzle sizes and they always worked

Edit: If you can work python, you too can now make your own

[Amedee]

I like the magical constant at line 15 here (which is probably irrelevant with the official Ultimaker HBK)

[Izzy]

One thing to check is the countersunk screws that level the bed, check that they are not proud of the aluminium plate.
I checked mine and they were raised, which ment across the back where the screw is in the middle and the 2 clips are at the edges, slightly bowed across the back.
I removed the glass plate and the levelling screws and then using a countersink carefully deepened the countersunk hole until the head was below the top level of the plate, I just used a 6" (150mm) steel rule gently across the top of the screw head with a 'rizla' paper( roll you own fags etc. they are only 0.03mm thick) therefore my screw head is 0.03mm below the aluminium plate.
Re-assembled and Re-levelled my build plate and better results.

[LePaul]

I noticed doing a big print that completely filled the build plate, I had to keep bringing the bed down for items near the back...it was fine in the middle and front.

However it wasn't terribly bad...everything came out alright after all that furious cranking while it made the skirt lines. Makes me glad that print was "print all at once"