I do not believe this can be mechanical levelling of the bed, although it could be that the bed itself is not flat enough. If mechanical levelling would help, surely the problem would be worse towards one edge or one corner of the bed, not towards the middle? Besides the first layers look fine in all cases.
Here is a new set of tests mostly around the periphery. Note that I used a different filament (also a different nozzle) to eliminate that factor:
This is actually nine separate runs which I have placed on the plate to show their approximate position when printed. The central one ‘MM’ as before is unacceptable while RM and FM have noticeable pillowing, but not so bad. The other six are perfect. It is hard to get the lighting right to show the extent fully, but MM, RM and FM all showed clear signs of grinding during printing.
I also did a test turning bed heating off immediately the first layer was printed, resulting in cooling to about 40C when the top layers were printed. There was no appreciable difference. I did try printing with the bed at room temperature throughout, but it broke away from the bed after a few layers. Once I got a good print in the centre of the bed by both reducing the bed temperature in this way AND slowing down both cooling fans to about quarter speed, but I was unable to repeat this with just one factor at a time.
During these tests I found a couple of bugs in the software:
- As shown in the photo, the priming stroke for single-colour prints does not avoid the model footprint on the bed, resulting in a filament of unwanted plastic attached to the print.
- Snapmaker Orca has settings for first layer and for subsequent layer bed temperature, but the first layer setting is used throughout and the subsequent layer setting is ignored. When doing the temperature test, I had to watch the print and manually reduce the bed temperature once the first layer had printed.
As suggested I will raise a support case and refer them to this thread.