A rather amusing bridging behavior in slic3r 0.9.9. Single perimeter, 0 infill, two bottom layers. Apparently the bridging routine doesn’t take account of the amount of infill behind the bridged perimeters, and assumes that something will be there to anchor the ends of the bridge a few mm behind them.
I don’t understand. Can you highlight in the pictures the behavior?
In slic3r, bridges are made by sweeping the print head rapidly back and forth across the gap, the goal being to drape strands of filament horizontally across said gap. These strands have to have something to rest upon at either end of the bridge.
The funny thing here is that slic3r is assuming that it has support for the bridge filaments in the zero-infill region of the print. It still makes a bridge, but the edges of the bridge are hanging in midair. You can see pictures of the generated gcode here:
The 8th layer is the one just beneath the bridge, while the 9th and 10th layers are the bridge itself. All other layers are simple hexagons making up the walls of the bucket.