Question regarding infill %

Talk about anything.
Post Reply
Az_h1
Posts: 3
Joined: Thu Jun 02, 2016 7:44 pm

Question regarding infill %

Post by Az_h1 » Thu Jun 09, 2016 4:52 pm

Hello everyone,

I am a new user of Up mini and was wondering about the infill percentage of the four infill options in the software. Is it correct to assume the infill is about 100/75/50/25 %? I could not find any information in the manual.

Thank you for your help!

tel
Posts: 242
Joined: Tue Feb 09, 2016 1:35 pm
Location: Australia
Contact:

Re: Question regarding infill %

Post by tel » Fri Jun 10, 2016 4:23 am

I would assume its closer to 10,20,30,40. Certainly nowhere near 100.

ftdesigns
Posts: 140
Joined: Mon Oct 15, 2012 5:19 pm
Location: ny/nj metro area

Re: Question regarding infill %

Post by ftdesigns » Sat Jun 11, 2016 3:14 pm

tegregg wrote:I would assume its closer to 10,20,30,40. Certainly nowhere near 100.
Sounds right. Even the highest infill % setting is not 100%

Az_h1
Posts: 3
Joined: Thu Jun 02, 2016 7:44 pm

Re: Question regarding infill %

Post by Az_h1 » Mon Jun 13, 2016 2:59 pm

Thank you very much for your help!

brainzilla
Posts: 141
Joined: Sat Nov 29, 2014 5:41 pm

Re: Question regarding infill %

Post by brainzilla » Wed Jun 22, 2016 6:57 pm

Well I guess it as close to 100% as the 0.4mm nozzle can get.
A higher feed-rate leads to the outer contour being pushed out of shape.
At least with the UP! slicer...

arhi
Posts: 256
Joined: Sun Mar 08, 2015 10:51 pm

Re: Question regarding infill %

Post by arhi » Wed Jun 22, 2016 7:25 pm

I'll post if I manage to find my old research with skeinforge and different infills ... strength of the object (ABS and PP tested not PLA nor HDPE) printed with hatch pattern (what up! slicer uses) between 85% and 100% infill is hardly measurable. same object (I was testing plates, beams, cubes and boxes) printed with hexagon pattern (default for skeinforge back then) between 70% and 100% there's almost no measurable difference. All the other things (layer height, temperature, nozzle size, direction of the print..) affect the strength way more the infill percentage .. it was published on bfb forum some ~9 years ago, Tony (he worked closely with bfb then) did some similar tests and came to same conclusions, Forrest (back then part of reprap core team, he then decided to leave the team, I stayed and am there still) also reproduced most of the tests I made .. I'll check if I have the docs and pics available somewhere ..I have the backup of the bfb forum somewhere on my site..

User avatar
Tiertime-Jason
Posts: 286
Joined: Mon May 05, 2014 4:31 am

Re: Question regarding infill %

Post by Tiertime-Jason » Thu Jun 23, 2016 11:29 pm

The study sound very interesting, could help us to understand how to get stronger parts with lesser materials. :idea:

arhi
Posts: 256
Joined: Sun Mar 08, 2015 10:51 pm

Re: Question regarding infill %

Post by arhi » Fri Jun 24, 2016 12:39 pm

few years back I lost all my docs (long story, don't have stomach to tell it again) so I don't have any of the docs I made then but there should be some remnence of that research online, just need to find it..

I remember all the conclusions I just don't remember the exact numbers (as I was measuring how much stress each can take..)..

now I did most of it with PP as it's the cheapest filament I can get but what I did test with ABS shown identical results so I think material is not too important in the relative sense, in absolute they of course behave differently..

I was mostly interested in material spending so most of the tests were about identical beams breaking/bending under load .. and as I said, with hexagon infill (like this https://picasaweb.google.com/lh/photo/m ... directlink ) anything over 70% is pure waste of material as differences in load capacity are measurable but insignificant. The standard, fast, hatch infill (the one UP use for e.g.) is lot weaker.. around 85% is the point over which the "more plastic" does not give significant strength, again it is measurable but in no way significant... now notice that hex infill pattern is much slower to print and introduces significant stress to your printer, especially with heavy heads as we used at the time (talking about 7-9 years ago), hex infill would shake your bolts in no time if you go faster so we are talking some 10x slower print! .. that's also time before acceleration and before smart path finders with corner speed calculation etc .. no clue how much faster you can print now the hexagon but I'm pretty sure that UP! can print the hexa pattern with this puny head and smarter firmware at least 20x faster then my old darwin could... so I wonder why hex infill is not a thing these days..

so, as I said, I do remember those few numbers from the infill pattern and volume % but that's about all I remember on the numbers .. the rest of the tests shown some expected results but if I don't dig out the real numbers these are common things that you all already know
  • 80% of the "free fall dia" is the strongest layer height build (free fall dia is when you extrude into air, let filament fall, let it cool, measure thickness of the "string")
  • thicker the layer, stronger the part (irrelevant to how good your lamination is, even at super high temps and additional acetone bath, thicker the layer, stronger the part)... the strongest parts I build (from HDPE for my CNC) I built using 1.5mm nozzle and 1.4mm layer height (free fall dia for HDPE trough 1.5mm nozzle is 1.75-1.8mm )
  • on a beam, when you print "surface" (top/bottom), you print them 90degrees from one another, now if you print then at 45 degrees to the edge of the beam beam can take more load before bending then if you print them at 90 degrees to the edge of the beam BUT the 90degrees to the edge will take more load "in total" as it will bend earlier but break later. tensile strenght is also higher for 90 degree print.
  • easier way to strengthen object is to add holes :D (the perimeter of the holes will do the trick) but you have to have proper infill to perimeter and perimeter to perimeter overlap
all this of course with very finely tuned printer, if your printer is not fine tuned and you are not extruding properly (over/under extruding), your perimeter/infill overlap is not good, your perimeter/perimeter overlap is not good you will get skewed results ... I seen prints with infill ratio of 70% that have 101% of plastic, plastic overflowing everywhere because of overextruding.. or 100% "hollow" prints because of underextruding, the greasing caused by overtemperature... all that needs to be solved before you go into details with orientation of the part and precission of the print ...

for e.g. looking at cpu's and fpga's available I really can't understand limitations we have attm... we had a script 6 years ago that was printing 4 0.1mm layers of perimeter at "fine" speed and then one 0.4mm of infill with "draft" speed .. 6 years later, UP!, the commercial printer, still can't do it?! I don't understand why :( ... 7-8 years ago we introduced some basics of acceleration, 4D, 5D ... I still see my UP! shaking like crazy, I don't see any corner speed calculations being actually done, maybe I'm wrong but.. again, I seen way smoother motion from linuxcnc on similar or even worse paths .. for e.g. looking at studio, what was wrong with how original sw 2.18 for e.g. looks like, why waste hours on making new interface, those hours could be spent making better slicer, better firmware... 8 years ago I have never seen ringing on my prints, almost all my UP! prints have ringing... ok I don't mind ringing it's not a big issue but still .. I get frustrated with two step front one step back ..

arhi
Posts: 256
Joined: Sun Mar 08, 2015 10:51 pm

Re: Question regarding infill %

Post by arhi » Fri Jun 24, 2016 12:47 pm

one thing I forgot to add and is fairly significant .. unfortunately no numbers but at low infill ratios (1%, 5%, 10%..) hexagon was much stronger then hatch infill (no clue if it was 5 times or 50 times really, it was "much stronger") and as the infill ratio grows then even out so after 50% infill the differences become insignificant .. but on the lower ratios the difference was "to not believe", not only for the beams but for crushing power, both from top and from sides and diagonals .. I remember I printed bunch of 7x7x7cm cubes for that test.. took ages to print .. and those I had to print in natural transparent PLA as at the time I didn't have a chamber so everything else would warp like crazy.. my house was smelling on sugar for weeks :D

orthoa
Posts: 7
Joined: Tue May 31, 2016 5:42 am

Re: Question regarding infill %

Post by orthoa » Tue Jun 28, 2016 10:46 pm

Hexagon would be significantly stronger we have been doing some tests on the UP BOX prints and with the "solid" hatch infill you are only getting roughly 20-30% of actual contact between layers. If we could print in a hexagon you would see roughly double the strength.

Post Reply