Sharing here as well.
What a difference good part cooling can make. Especially with bridging. The diameter of the hole is 31mm. Left side has no part cooling fan. Right side is 100% cooling fan and the same printing speed as the outer surfaces. This is with PLA. No support material used.
The part cooling fan ducts I used are here: https://www.thingiverse.com/thing:2901039
With just one fan / duct do you have to take care orienting the print on the bed? What settings do you use to prevent the hot end from cooling off / ensure it maintains correct temp? (ie: fan speed %, at what height do you enable the fan?)
I’ve been working on getting my ducts to be more flat like yours, but I either run the fan so fast the hot end can’t keep up, or else features on the model extending away from the hot end/duct curl up due to lack of cooling.
@Alex_Wiebe I had the same issue, and I noticed that fan over 65% reduce the hotend temperature.
Days ago I noticed that the hotend was 178C instead of 190! As I turned off the cooling it went back to 190.
Strange that it never happened in the past. Still to be investigated.
@Alex_Wiebe No need to orient the print any particular way. The duct design is such that it blows a concentrated stream of air right across the tip of the nozzle horizontally across the top of the print. See the picture below. I can feel the air stream on my hand up to about 10 -12" away. There is very little turbulence from the nozzle tip in the air stream so geometry on the far side of the nozzle cools just about as well as the near side.
It does not affect the hot end maintaining temp at all for me. I get rock steady hot end temps with the fan all the way to 100%.
By using a centrifugal fan that produces a high Cfm and over pressure I can restrict the nozzle opening and direct the air exactly where I want it and not have blow back through the fan. (By comparison cover up more than about 1/3 of the open area of a regular 40mm fan and you get air blowing backwards through it. Really hard to concentrate that on a focused area.) Being in one direction it removes the heat from the vicinity of the part quickly. I also print in 0.1mm layers or 0.05mm layers for precision prints so there is not a lot of mass to cool off.
I will have the fan off for the first layer (0.1 - 0.2mm height) and then at 33% for the second one. 66% for the third layer and 100% for the 4 layer at which time I keep in on full speed for the remainder of the print. At the 4 layer or 0.5mm in height I also drop the bed temperature from 60 to 45 to keep large flat parts from curling up.
missing/deleted image from Google+
@Alex_Wiebe Here is another picture of a part with the same features in different orientations with no difference in crispness and precision. No curling to speak of on the thin flat nearly horizontal fingers. This would have been a nightmare if supports were used and hardly possible without good cooling as is.
missing/deleted image from Google+
Do you have a picture of a side profile showing the distance of the nozzle to the tip horizontally and the height of the nozzle opening to the bottom of the tip?
@Alex_Wiebe Here is a side profile of the duct lined up with the nozzle. The sloped part at the tip should make a line that crosses the tip of nozzle. The end of the duct is about 12mm from the tip and the bottom of the duct is about 1mm above the nozzle tip.
A fun side effect of the horizontal direction of the air stream is that it will sort of whistle across certain infill and geometry profiles. 
missing/deleted image from Google+
@Jeff_Parish I have built exactly yours, and I never noticed a change of temperature, until recently, when I replaced the thermistore. May be do I need to recalibrate the termal process (e.g. Marlin receive resistance data slide different that confuse the process of the heath) ?
@Roberto_Coli Changing up your thermistor could affect the stability but the part cooling air should not be affecting it. Does the temp still drop when you are 30 - 40 mm off the bed with no print just as a test? i suspect there is something else gong on as well. I’d check your thermistor settings and run another PID tune.
Yeah cooling is pretty important, i made a fanduct only because bridging. Always had the problem if the air steam hits the nozzeltip directly and i have a long bridging area, it blows down my bridging string. Now i have a duct that only sourrounds the nozzle with air, way better when bridging than before, it only hits the nozzle when i print flat surfaces, but then i only let him run on 30-40% so my nozzle will not cooled down.
But for real, i want a fanduct that changes the direction of the airstream with the movement of my nozzle. So it will always blow in the direction ov movement:joy:
