Experimental Predator Ducts

Last crawled date: 4 years, 2 months ago

The goal of this design is to push as much air as much as possible directly under the nozzle to improve bridging and overhang performance. The idea being to have the air cool the print where the plastic will be the hottest to maximize the amount it is cooled. This does have the side effect that some of the air will be blowing on the bottom of the block, but a block sock will greatly mitigate this problem and help with the overall cooling performance. I have found that pretty much all of the other designs here either direct the air downwards where the plastic should have already solidified or just blow in the general direction of the filament nozzle and this is why I have made my own design.
Now featuring version 5.3 one for 4020 fans and another for 4010 fans with roughly 60-80% more room airflow than version 4. They should have enough space for a v6 block(don't own one so I cannot confirm). Has 360 degree cooling similar to version 4. 5.0 and 5.1 did not fit, 5.2 was usable but I wasn't satisfied enough with it to share. MY 5015 to 4020 adapters similar to what is pictured can be fan here: https://www.thingiverse.com/thing:4132675
The 4010 version is just a quick and dirty adaptation of the 4020 version, not going to test or use it, but I figured I'd make one since it took like 5 minutes to modify. Added an overhang test picture that was printed using the version 5 duct, I clearly have the print temp way too high, and need to re-level the bed(just switched from PTEG to PLA), but the overhang quality is the best I have seen so far, even my Army Ant Duct doesn't do this well, print speed was a bit over 100mm/s(estimated average).
DISCLAIMER: Modify your printer at your own risk. Messing with the cooling system can cause clogs potentially to the point of damaging the hot end. Also be wary of stripping out the screw holes on the effector plate, I had one that was not formed completely and after about 10 or 20 swaps it eventually stripped, I just flipped the effector plate and cut a new slot for the wires, luckily the threads are a bit better on the other side in my case, otherwise if this happens to you you can use longer screws or even screws with nuts if it's really bad.
Your effector plate will need to be close to level if you want to use version 4 or 5, if you have any arms that longer/shorter than the rest you will want to fix this problem prior to using. Personally had one that was too long and another that was slightly too short, I just bought a replacement pair off of amazon to replace them. I made a measuring jig with a small board and finishing nails to make sure they were all close to equal. Strangely enough when I first got the printer it didn't really give me any issues, but later on I had some inconsistencies leveling the bed and this fixed that problem.
I printed them with PETG, but ABS or any other high temp plastic should be fine. I designed them with fairly thick walls on the side facing the hotend to prevent warping as I have found that thinner walled ducts, even the supplied injection molded ABS duct that sits further away and the other ducts that are available here warp/melt and so far have not had any problems with my designs melting despite the fact that they are closer to the hotend.
These models do not seem to slice correctly with anything much larger than a .4mm line width. I would also recommend preventing supports from being built inside the model, as it is very hard if not impossible to remove them. Using prusa slicer I have found that setting the support threshold to 2 degrees will prevent supports from being generated in the model and in the exhaust holes and also allow for the plate to be supported. With Cura, setting it to supports on build plate only should be good enough. Cura also seems to handle slicing these duct designs with few fuss, with Slic3r you will either need to enable or disable the detect thin walls feature depending on the design.
These are some of the first things I have ever made in CAD, so it's far from refined as I am still learning. It was created using Sketchup, the STL files have been fixed with 3D Builder and should be ready to slice.
I made a demonstration model that can be hooked up to a garden hose to show the shape of the flow, it's not perfect but it does show that most of the air should flow downward below the block rather than at it. The demonstrators that are pictured here were made from my 1.4 and 3.0 designs
Version 3 should have a bit more optimized airflow compared to version 1.4 or 2, and personally the version I have been using but will soon switch over to my 5015 design here: https://www.thingiverse.com/thing:4097656
Old update: version 4.1 and 4.2, 360 degree cooling and improved internal geometry help the air flow downward a bit more. 4.2 just has a few minor tweaks to direct the airflow a bit better. You will have to take it off the bed level sensor or use my modified bed level sensor housing to re-level the bed with version 4.