I hate rafts and supports and avoid them as fervently as I can. In fact, you could say that I'm a bit rabid about the topic.
To be honest, of hundreds of objects I've printed with the Cube printer, the number that required supports could be counted on one hand that had lost a couple of fingers. In fact, I think it was only one... when I first got the 1st generation Cube.
But, recently, I have run into some designs that just would NOT stick to the table for the length of time they needed to remain attached. Of course, I blamed it entirely on the glue and made my feelings known to the Cubify design team.
Then it dawned on me that while there might be some potential for a stickier long lasting glue, the real issue just might be my adament refusal to consider using a raft. So, I decided to test that theory and reprint one of my non-raft failures with a raft.
Ooops! Duh... that worked like a charm.
I think I know why. I was trying to print 8 beam-like objects that were 130mm x 10mm x 6mm. The adhesion had to work with the 6mm edge. The length offered more than a little opportunity for warping. So, I cut pentagonal holes along the length hoping to help mitigate that potential. While it helped, it probably didn't take enough stress off to completely stop warping.
You would think that a perfectly flat surface would adhere better than waffle surface. But, upon reflection, I can see why that might not be accurate. A raft probably offers more surface area for adhesion because the glue, which has some depth of its own, has the opportunity to ooz up into the seams between the first layer of lines of the raft and wrap around the filament adding a bit more grip vertically as well as under the filament. The actual area is probably considerably bigger than offered by the flat side alone.
Whatever the reason, one can't argue with the results. So, I've decided to be a bit more flexible about my no raft rule. When you need it, you need it... sigh....