2

For some recent prints I have been using rafts as generated by Cura to aid in bed adhesion; which helps, but is not as reliable as I'd like it to.

However, it seems to me that the raft generation could be so much better; if I explain what I have in mind, maybe you can explain to me why it is a silly idea; or save me from reinventing the wheel, if some other slicer already implements similar functionality.

What seems silly to me about Cura rafts is that the bottom-most adhesive layer consists of continuous straight lines. This is literally the worst geometry, in terms of being able to relieve its internal cooling strains. Peeling always starts at the extremities of those long first lines, where the accumulated strain overcomes the adhesion, as you would expect.

What seems to me much more optimal is to have many non-connected small 'suction cups' as a first layer; little circles or maybe literal point-like dots. And cover a solid percentage of the surface with those unconnected cups. Like a fine honeycomb pattern. And then use subsequent layers to connect these cups with wavy lines of decreasing spacing, to build a somewhat continuous platform, while still providing a form of natural strain relief within those lines. If you take enough layers with a sufficiently compliant structure, you should be able to master arbitrary spans, with arbitrary shrinkage.

Has this been tried already? Am I missing something? Or should I start writing my own G-code generator and figure out why it might be a silly idea myself? Speaking of which; are there any python libraries to aid in generating G-code from a somewhat higher level format?1

1 Forget about that last question; gcody seems to fit my criteria for not reinventing too many wheels.

Greenonline
  • 5,831
  • 7
  • 30
  • 60
Eelco Hoogendoorn
  • 151
  • 3
  • 5
    Welcome to 3D Printing! What else have you done to ensure proper bed adhesion? What preps are you doing? How do you level your bed? What kind of filament are you using (PLA, ABS, etc)? What temps are you using on your bed and nozzle? The only time I use any type of brim is if the base of the print is too small to have good adhesion. I use a skirt quite often to prime the nozzle. I've never used a raft. I'm sure there must be reasons to, but I've never needed one ... which brings me back to ... *what's happening that you need a raft in the first place?* – Pᴀᴜʟsᴛᴇʀ2 Oct 01 '19 at 20:53
  • 1
    Thanks for the input, but I am not looking to fix a particular print here, but rather to discuss rafts more in general. Will fix my introduction to reflect that; it might give the wrong impression. – Eelco Hoogendoorn Oct 01 '19 at 21:10
  • 1
    As you should know as an experienced SE user, this is not the place for discussions. Please join us in [chat] for discussions. – 0scar Oct 01 '19 at 21:15
  • Poor choice of words perhaps; there are a few delineated questions on my post that potentially have very clearly defined answers. – Eelco Hoogendoorn Oct 01 '19 at 21:26
  • Before reinventing the wheel, you should fix your question to not contain all these rhetoric questions. Because some of these can't be answered by us, others are just tangentally to the seemingly actual question. To me this reads like "I have an idea of using a circle pattern for bed adhesion, then some layers of gyroid infill to provide a raft. Would this have benefits?" and then you asket the 2nd last paragraph – Trish Oct 02 '19 at 08:21
  • I second @Pᴀᴜʟsᴛᴇʀ2's question about material. For PLA at least, rafts are generally considered antiquated. You should not have adhesion/warping problems with PLA unless something is wrong with your setup, and rafts waste a lot of time and material and are generally bad for bottom surface quality. – R.. GitHub STOP HELPING ICE Oct 02 '19 at 12:44
  • @Trish yeah you summarized it well. Seems I am more effective in learning about rafts by generating gcode than by figuring out how to write an iso-compliant stackexchange post though, so I think I'll focus on the former. – Eelco Hoogendoorn Oct 02 '19 at 13:09
  • 1
    If you have trouble in getting straight lines to stick, you don't have raft problems, you have bed adhesion problems :) Maybe focus on that first as the Cura rafts work perfectly for me and others (but I usually try to avoid them!). Those straight first layer lines are extra wide lines, these "should" stick relatively easily. I've even got rafts to work with POM, which is not easy to stick to the build plate. Please note that many people have problems is printing concentric circles (non sticking circular perimeters are posted here multiple times), they detach quite easily. – 0scar Oct 02 '19 at 13:57
  • @R..Rafts for PLA are only needed in some few *fringe* cases: if the print bed is not uniform smooth but perforated. There was a printer in the early days that had such a setup, it could not be used without raft. Slim geometries also demand extra adhesion on *unheated* beds, and that can sometimes better be achieved with a raft than a brim. – Trish Oct 02 '19 at 14:13
  • 2
    Perhaps this question should be changed to just ask why it is common for a raft to have the bottom layer consist of parallel lines. Does this offer any advantages/disadvantages over other patterns. – Perplexed Dipole Oct 02 '19 at 14:55
  • @oscar; the question is, do the circles not stick because they are circles, or because they are extended objects? My mechanics intuition says a large circle and long line of equal size will take a comparable shrinkage to adhesion ratio to come loose. The key is to avoid contiguous structures along which shrinkage strain can accumulate; if they are dots or tiny lines or circles matters little I think – Eelco Hoogendoorn Oct 02 '19 at 15:43
  • 1
    NOW you are onto something: when is a circle and a line not equal if their area is equal? The secret lies in how extrusion is done and that the pattern of circles would consist of many separate paths which are interrupted with moves between each while a line pattern could be one continious extrusion, guaranteeing a good surface coverage. – Trish Oct 03 '19 at 12:07
  • True, youd want to cover a large percentage of the area; though tiny circles/hexes could in theory cover an arbitrary percentage, what I am currently leaning towards is to simply print straight lines with small skips of extrusion in them, as this structure nests well and prints fast. – Eelco Hoogendoorn Oct 03 '19 at 13:23
  • Here's a non-answer: use better technology. What would you do if you could skip the raft altogether? I installed a WhamBam Systems bed system. It uses PEX material on a flex-steel sheet, that sticks to a magnet applied to the AL print bed. I've only printed with PLA; it sticks really well to the PEX with nothing else needed. When the print is done, the flexsteel sheet it lifted off the magnet, you let it cool for a minute or so, then flex the sheet in the X and Y direction and the part pops right off. I really like it. – CrossRoads Oct 06 '19 at 16:16

0 Answers0