Mathematically calculating if your part will warp...
Hypothesis: There is a limit to the volume of filament, per cm2 of bed, per hour that can be printed before warping is uncorrectable. (with ASA at least)
Been running a small production line of parts on my Mk3S - same parts, 100's of them, got it all worked out for bed adhesion & getting perfect adhesion & no warping. Always using Prusament ASA & nothing else (consistency!!!)
Get my mk4 kit up & running & NOTHING BUT ISSUES... sadness filled my heart.
Warping horribly & basically useless - it "couldnt print the same parts in the same environment reliably at all"
I went through the usual steps, cleaned the surface, glue, copied all the print settings across, temps were good, fan cooling was good... no joy. I even used the exact print sheet from the mk3 on my mk4 - it wasnt working at all!
Then printed 3 times the parts at once & all was well!
The mk4 was just printing too fast - putting down additional layers while the others were still much hotter than previously.
How can it be measured?
- volume of filament, per cm2 of bed/layer, per hour: a lot of filament build up over a large area will create a large pull force on the bed. If its a smaller area - then its a "tall" print & wont have as much pull on the bed. If the lower 10 layers are the same temp - they could resist pull from upper layers, so time is a factor
- layer size / time to print a layer: different way to measure
Here's some rough calculations in a spreadsheet to show what I mean...
Temp = the temperature of the layer in the print
Force = how much the layer "pulls" on the existing print, or can resist pulling. Arbitrary numbers & formula: 40-(layer temp-bed temp)*5
Cumulative warp value = if it goes negative it warps & is pulling on the bed adhesion
Bed force = adhesion, start with a high value, this is how much pull you can withstand
Each layer = the higher the difference between the layer temp & the bed, the more it pulls
Thoughts?
RE: Mathematically calculating if your part will warp...
There is commercial software available that will predict the resultant shape of 3d printed metals and other materials. Last I saw it was in the range of $100000 U.S. It does massive amounts of analysis.
Also some commercial slicers are able to alternate passes, do the odd number passes, then the even number ones kind of thing. This prevents too much heat from building in an area. Good for a corner or thinner areas. Again high cost or proprietary for business use.
I have done limited ASA printing in an enclosure to hold the heat. Most were good, but I added a brim to keep some thin parts from warping up on the ends.