Direct vs Bowden Extrusion when printing in temperature fluctuating environment

Question

I have a Bowden extrusion printer in my garage. (Printing PLA is smelly and is nothing I want to be exposed to for an extended period of time!)

I enclosed the printer within a plastic sheet housing to reduce the energy consumption of keeping the hot bed tempered. I noticed that a strong temperature fluctuation influence how the PLA part turns out.

If I start printing in the morning a print that takes 2 hours to print -- in a cold garage, cold filament, cold tube, I end up with a solid first 4 layers and lacy last 4 layers. The last 4 layer are so thin that they end up to be almost see through.

last layer

first layer

I am assuming that the PLA and tubing heat expansion difference is the reasoning for that undesired under extrusion phenomenon.

The cold Bowden tube is perhaps expanding more then the PLA filament and the tubing is warming up more quickly being directly exposed to the heat during the heating up of the unit; obviously, the PLA filament is inside the tube and is kept colder for much longer.

How can I solve this -- keeping the printer in the garage? And, would a direct extrusion mod remedy this under extrusion?

Most likely preheating the printer would help, but that would call for an insulation upgrade from plastic sheet to Styrofoam. Insulating the Bowden tube may also be worth trying. (Happy printing.)

Answer 1

I don't think your temperature based explanation makes sense. This looks to me just like a first layer smashed against the bed, possibly too much due to nozzle being too low, followed by underextrusion when there's actually the right amount of space to extrude into. If you have fewer than 3 top layers or less than 0.6 mm of top layer thickness, you will almost certainly have an underextruded top like this just from the first top layer sagging into the space between the infill lines. Since the walls look okay, this is my best guess; if the walls looked bad too I would suspect poorly tuned filament diameter/E-steps-per-mm. Also check that your infill is not too sparse; more than 4 mm or so between infill lines gives a lot of room for top layers to sag into and may increase the number of top layers you need.

I see you added in the comments that your layer height is 0.1 mm. At such thin layers, you probably need at least 5 or 6 top layers; the first 2-3 are expected to sag and not fill the space.