Thursday, May 15, 2008

 

Inter-layer cooling?


Now Adrain's got the polymer extruder valve tuned, we've got loads more control over the extrudate. Which begs the question, why nozzle wipe? From experience this procedure ups the risk of what we affectionally call "nozzle crap" - unwanted bits of filament getting attached to the nozzle. So now there's the option to skip that and move straight to the next layer. The only side effect is cooling is eliminated, which might be a problem for small parts, we shall see...

Comments:
Couldn't this be done already by setting the cooling period to zero?

Vik :v)
 
You definitely need cooling when the object gets below a certain size. Basically heat is being delivered to the object at a rate determined by temperature, flow rate and specific heat capacity.

It leaves it at rate proportional to its surface area. Once it gets small enough that it is arriving faster than it is leaving then the object gets hotter and hotter and becomes a molten blob stirred around by the nozzle.

I run a fan continually. It allows me to make smaller objects without pauses but there always comes a point where I have to slow down as well. Vik's wine glass is a good test of this as the stem gets smaller and smaller. Even with a fan on I had to slow from pi mm3/s to half that rate for the stem when making it from PCL.

I think a fan also improves the build quality because it hardens up the layers below quicker giving sharper corners etc.

You need good insulation on the nozzle and good airflow from the fan to the object without being shadowed too much by the extruder.
 
Why wipe the nozzle, when you can plug it with piano wire?
 
A static cooling value need changing with different print jobs and is also dependent on the machine layout (phenotype), it cannot be stored as metadata of the .stl (for the proposed multi-material format that contains .stl's)

Why is the cooling period not calculated (with an adjustable factor)? If the build time of a layer is short, the cooling should be longer. A function as such:

CoolingTime = abs(MinimumCooling-lastLayerBuildTime);

A thermal model would be better, but this may do the trick already. Leaving the fan on as Chris does may also be an option ;)
 
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?

Subscribe to
Posts [Atom]