Friday, May 14, 2010

Liquid cooled extruder

I've been thinking about extruders and their problems concerning the nozzle/barrel design. Nophead had mentioned the idea of a water jacket for an extruder and I thought that was an idea worth considering. The goal is to have a short melt zone for the filament without having a large heat sink mounted to the extruder/Z axis to remove heat from the upper barrel. 

I welded this extruder from stainless steel tubingThe diagram (not to scale) shows how it's constructed- the inner barrel is tapered at the tip and then has a short thin wall transition section to the upper barrel section to minimize heat transfer to the upper barrel. The upper barrel has a water jacket around it for cooling. This assembly will be mounted to the Z axis by a block that clamps around the water jacket.

This particular extruder has a one piece barrel/nozzle. The next version will probably have a removable nozzle/heater so I can easily change nozzle size without removing the entire extruder or having to take apart the entire assembly. The machined groove at the top of the barrel is for a plastic guide for the filament to reduce friction as it enters the barrel.

I'll use a small aquarium pump with a PC radiator and cooling fan to complete the system.

I still have to make the heating element- more than likely it will be an aluminum or brass piece with a resistor mounted to it. No idea how well the water cooling will work but it'll be a fun experiment! I can easily modify the shape/size of the water jacket if needed.

I've been out in the garage cutting aluminum extrusion for the chassis so I should have another good update before the weekend is over...  


  1. Looks very nice craftsmanship. Is that a TO220 resistor for the heater? I don't think they have a high enough temperature rating.

  2. Thanks- it's not 100% perfect but it should be good enough to test the idea.

    I read a while back about someone using a TO-220 resistor but I agree that it would be really pushing it, especially since the resistor would have to conduct heat through the stainless nozzle. I imagine I'll end up using a vitreous enamel resistor like you've done.

    Based on the requirements you gave me for power dissipation and temperature control I calculated the necessary flow rate at .034 gal/hr. which sounds really low. That doesn't take into account heat put into the system by the pump as well as loss in flow rate through the system, thermal efficiency, etc. Even taking those things into consideration it would appear that water cooling is totally overkill... and may not work at all due to the necessary temperature gradient requirement of the extruder. We'll see- either way it'll be fun to see what happens!

    I also don't believe I've read about anyone using an enamel resistor with the MakerBot extruder controller so that may be another hurdle to overcome.

  3. kudos on that work, how well did that work out for ya.?

  4. I already made a new one- I'm trying to get the printer finished this week!