Nozzle diameter vs actual extrusion diameter

In another thread we were discussing Esteps and how the calibration process for me yields a number that results in over extrusion.

So last night I manually extruded (in air not on the bed) some ABS filament and measured its diameter. My nozzle is marked .35 but the filament extruded measured an average of .47mm after it’s cool.

Is this normal? Or should it be closer to .35mm? Perhaps the plastic expands after it leaves the nozzle. But would it expand that much?

Perhaps people with .35mm and .5mm can run that test and report their findings?

If my nozzle is mismarked it might explain why my Esteps calibration results in over extruded parts?

It is indeed normal, and the amount of expansion will depend on the specific polymer you are using. This is known as die swell.

Not only will this vary depending on type of filament, but also manufacturer and color as well. Each manufacturer will have a slightly different formula, and the individual pigments can affect this as well.

It would be interesting to experiment with a longer nozzle. But of course that would require more force and would probably be prone to clogging. Still if you could get it to work with a longer nozzle, the extrusion would be more consistent across material brand and speed.

Additional nozzle length would act as a heat sink eventually, which would stabilize the diameter of the extrusion, but would render it solid and unable to stick to the layers it was being extruded onto, which would kind of make for a somewhat useless 3d printer. That’s one of the reasons the variable diameter nozzles haven’t really been feasible as of yet,

I was thinking the nozzle would have to have a longer minimum diameter section which remained constant through the heated zone. So it shouldn’t effect extruded temperature. It just would require a longer heated zone. As I understood the Wikipedia article, the longer the plastic is kept in its elongated shape the less the die swell would be a factor. I don’t think it has to cool for the cross section to be more consistent.

The longer the narrow section is, the more likely you get jams is another consideration. But if you wanted to put a 4 inch long nozzle on a 3d printer hotend and heat the whole length, you could probably eliminate some of the die swell. Its probably easier to just account for it software side, since it’s basically going to be a constant at the scales we deal with.