Layer fill methods

It seems to me that the slowest part of each layer is the filling in of the part known to be solid. I remember reading on the blog about an idea where the model was honeycombed into cubes or pyramids internally, so that it would take less fill per layer, and thus less time.

Are there any other possible ways to either reduce the amount of infill, or change the way its done?

I was wondering if maybe the outline of the layer could be printed in the normal fashion, and then filled in by pouring in a liquid to a specified depth. I don't know of any materials that would be a good choice, but it seems that using a liquid fill method could be much faster than laying single mm wide traces across the relatively large surface. Alternately, could an extruder head be made that outputs in a wide slot? I have been wondering what the effects of changing the size of the extruder nozzle are on the layer thickness, if any. Do larger holes produce thicker layers? If not, then I suppose just a coarser nozzle could be used, but otherwise a slot nozzle or other method would be better.

Any ideas on reducing fill time?

-Samuel

Forrest has experimented with moving his infill stripes further apart, effectively creating a lattice-like infill pattern which reduces material usage and print time and helps solve the HDPE shrinking issue while still maintaining some structural integrity.

I like the liquid fill idea, but like you I have no idea what sort of liquid one would use (Fernando or ohiomike, any ideas?) Thermoplasts themselves don't really behave like liquids unless you get them unmanageably hot.

There hasn't been too much hard experimentation done as of yet with the effects of different extrusion parameters, this is currently an area of active inquiry. There may be a few posts in the builder's blog scattered around, and nophead has some interesting information a page back on his blog at [hydraraptor.blogspot.com]. Check it out.

i really like the liquid idea... its something i've been tossing around for a while.

it would be cool to do an infill method where you print the outline / shell, then print a very sparse infill... enough for structural integrity, etc. then, when the model is finished, you fill/inject it with some sort of epoxy or resin that would settle in and harden.

if you work it right, the finished object might even be stronger than just raw plastic. the plastic might even act like a sort of rebar for the object. now wouldnt that be cool??

Wow, Zach's idea sounds awesome. That has the same advantages as Samuel's idea, but also is less prone to spilling (you can fill it in a more controlled environment) and even more importantly is feasible with a 1.0 Darwin, with the proper software!

[www.3dreplicators.com]

You don't need solid objects. You get beaucoup structural strength using HDPE with 33-50% infill.

nice... the sparse infill concept seems to work pretty well. if we do it in such a way that you can optionally saturate the interior of the object with epoxy resin for added strength, then thats just icing on the cake!

The much more elegant way would be to make better designs for your objects. I you need a cube, say, which is going to get most load along one diagonal, the thing to design is not a solid cube, but an emtpy one with an internal structure that reinforces it along that way. Add some light supports inside so you can print the upper face of it, and you get a cube with has nearly the same mechanical properties, but weighs about a tenth of a solid one, and prints in a tenth of the time.

If you fill your object with a lattice, you always get a weakness diagonal to it, which you have to compensate with a lot of material. If you need the thing to be heavy for other reasons, you