Thanks for a very thorough answer!

It’s possible your original Blender design had an issue. Blender is not always kind to 3D printers.

I’ve had good success with previous Blender files, although this is the first time I’ve used a boolean operator to cut out anything. I usually use FreeCAD for these custom Gridfinity pieces, but the process of converting the .stl mesh to a solid part in FreeCAD seems a bit error prone (several steps involved), and I haven’t yet used booleans in FreeCAD. I could try that again.

The first thing I would tell you is to stop using cubic infill, it is evil. It never always causes me failed prints, especially larger prints. Nozzles often tend to drag across the previous layers and can easily cause failed prints. I can even hear the nozzle hitting the infill as I print. I often recommend gyroid as a good all around infill pattern.

Good to know! I use PrusaSlicer, and this grid infill is the default. The way you describe it sounds like what I experience, and I can in fact see some artifacts when I inspect it closely. Though, the sounds I hear would only start when the concave part starts, and that’s also where I see the failures. But that could possibly because there’s too much overhang over the cubic infill? Anyway, I checked out gyroid pattern, and it was pretty dense with the 15% default infill value. What type of infill % do you typically use? Seems I could get away with less here.

I look at it and I wonder, does the rolling pin need to be supported full length? A wooden rolling pin is ridged and only needs minimal support on the ends. So I might just design the cradle only at the very ends. And then design the middle to be a simple flat that connects the two end pieces. I might even skip the middle altogether and just print the ends. That saves the most material and time and still does the job perfectly.

You mean an open container with maximum depth and width between the ends that holds the ridges? That could be a good way. They way I ended up doing it was essentially just a rectangular cut-out which worked fine and is similar to your suggestion (although I could save more material doing it your way), but feels less custom… as if that is a goal in itself. I would not like the gap by just printing the ends though, as I wouldn’t be able to squeeze anything else underneath and it would not look right to me. Wish I didn’t think like that, so I could save material, but I know myself enough that I would be annoyed every time I opened the drawer…

When it comes to slicing your print, orientation matters. How you support overhangs can be tricky and often compromises must be made. While I will use the auto supports as often as I can, sometimes you just need to use paint on supports to get what you needwhere you need it. Pay attention to the top zed support gap. The defaults are never right. I always open them up more. With a .40mm nozzle, I use a .265mm gap. For a .60mm nozzle, a .365mm gap. You might even need to print your parts at an angle. Often tipping the part at 30 to 45 degree angle can make those nasty over hangs completely printable without supports. And this is only a good beginning. How fast you might print an overhang matters, the amount of cooling fan can affect the over hang, lots of fine details that you will learn about as you keep doing this.

This I will need to read up on more. I don’t actually use supports for these Gridfinity prints (but PrusaSlicer does warn me about potential instability…). The printer handles the overhang between the grids fairly well, but I guess I didn’t think about the long lines crossing the infill. In other prints I’ve only used the auto supports. Could I ask you what slicer you use?

Good Luck and never fear making a mistake!

Thanks! I must admit I do fear it sometimes when the printer makes some weird noises…