How do the developers of slicer software get there default settings? Are they just found via experimentation or are they based on some theoretical relationship between hot-end temperature, feed rate, and head speed? (or are they just tribal knowledge?)

Basically I'm interested in learning the physics of what's going on with FDM printing in depth (I'm a mechanical engineer so feel free to be technical). So far everything I've found has been very qualitative but not very quantitative (which is understandable given how complex FDM is when you think about it). I figure this might give me a lead.

Thanks!

I think most of it would be from experience playing with them (trial and error) of course things like temps can be from specs of the plastics used and a little testing to find the optimal temps. If you use a 3D printer for a few months guessing at the best for the job at hand becomes kind of second nature.

Allesandro Ranelucci is the author of Slic3r and I believe he has several machines in his lab for testing every conceivable situation.

---

_______________________________________
Waitaki 3D Printer

A lot of the process depends on the fluid properties of the heated plastic of relatively uncontrolled composition, which, taken with the measurement of temperature in the melting chamber of a hotend, and its calibration, leads to a pretty uncontrolled process. The users need experimentation to tune them for their machine's calibration, their source plastic, and the application.

Indeed DaveX, there is a lot that is uncontrolled with consumer/reprap 3D printing today, but that just means a theoretical model would be that much more useful. Even though material properties vary from spool to spool, I'm willing to bet the same spool will have consistent properties (well assuming you don't leave it in a humid bathroom for 3 months in the middle of its lifecycle). Outside of the material you're only inputs really are feed rate, hot-end temp, nozzle geometry, and head speed, all of which have well known variations.

How these all interact is still a complex physical problem, but if you can determine a nice algebraic formula to describe how the interact with material properties you could use that as a road map for quick (and possibly standardized) optimization. When your optimizing you just compare your current results (say extrusion width is too wide) to desired results and since material constants wouldn't change in such a comparison, they'll likely drop out. In other words, the model could be useful even if we don't know exactly what the density/viscosity/etc of our specific spool of plastic is. The model could also predict what order you'd want to change things (like should I change feed rate by 1mm/s or 0.1mm/s?).

Nozzle geometry is relatively fixed, and the easy-to measure and calibrate head speed is easy to control and adjust. If you are just looking for the marginal effects of changes around an uncalibrated operating point, maybe a simple-minded Taylor-series expansion around the operating point would be enough to predict the effects of 1mm/s changes. For bigger questions, like "given that this setting works with this quality of output, how fast/hot/cool can I go?" you'd need better-than-average calibration and control of materials to predict that you'd start having problems at 72mm/s, 217C, or below 183C.

Maybe look at [www.extrudable.me]