MotivationI melted my DiiiCooler again. I was printing ABS, which most would say requires little to no cooling. My DiiiCooler was PETG, but after a few hours it started to sag and caught on the print. I should have taken it off or replaced it with something more resilient before I started, but such is life.I started researching alternatives and found this great study on the effectiveness of various fans and shrouds. (I will use ‘shroud’ and ‘duct’ interchangeably here, although I’m sure they have different meanings.) It looked like maybe the type of shroud didn’t matter too much as long as you had a powerful fan. I was looking forward to using a shroud that would finally give me a clear view of the build plate.I began experimenting and things snowballed. I ended up doing this “comprehensive” study on the most popular fans and ducts. The results were predictable (you jumped to the end already anyway, so you know), but there were some surprises. I’m not the first to do this type of experiment, but maybe I can add to the knowledge base.Theory[Feel free to skip this part. You probably already know why fans are important. I just go on and on about how cooling works.]Filament is hot when extruded. That’s sort of the whole idea. But, when you lay down a layer of extrudate (it’s a word) you want it to be solid before you lay down another layer on top of it. If not, the layers start getting squished and you lose the integrity of your design. Not only that, but when you try to bridge or do an overhang, there’s very little or even nothing under your current layer. The more quickly you can cool your filament below the glass transition temperature, the less your layer will sag. This is why we use fans.I started thinking about how a fan cools a surface. There are a few important variables, but mainly there are two big aspects: air velocity and time. Air velocity contributes to the rate of cooling, while time is how long that moving air is in contact with the hot filament. Let’s start with air velocity.A cooling fan is capable of producing a certain pressure which, given a flow configuration, translates into an air velocity. Axial fans typically produce a lower pressure than a blower (also called a centrifugal fan). The type of duct also determines velocity due to the flow area and internal resistances. You can create a narrow duct opening and you’ll get a higher velocity (but lower total air flow). Increase the size of the opening, and the velocity decreases but you get a larger amount of air. It is always a trade-off.What about time? That one is more difficult. You only have so much time after you extrude to cool the filament before it starts to deform. You need to use that time as effectively as possible. Your fan is moving with your hot end, so you need to consider its zone of influence (ZOI), which is the characteristic length of the zone of high air velocity. For example, let’s say you have a duct that points at your hot end from one direction and gives you a beam of cool air with a width of 20 mm. Your hot end is moving at, let’s say, 50 mm/s. The hot filament will only be in the ZOI for 0.2 seconds. (The filament is extruded in the middle of the ZOI and ‘moves’ to the outer edge.) That’s not a lot of time. So, naturally, you think it’s time to get a wider duct and increase the width of your air beam. Well, you’ll be lowering the air velocity in the process. Okay then, use a duct with a smaller nozzle to get a laser-focused beam of high-speed air. Well, you’ll lower your total air flow and decrease the size of your ZOI (which, in turn, lowers the amount of time).What does this all mean? I have no idea. Well, for one, I guess it means you should probably be using a blower if you care about cooling. But as far as choosing a duct… That’s why you’ve got to do the tests.ProcedureThe plan was simple. Choose a selection of popular and different fan/duct combinations and print an overhang test. The overhang has angles of 30, 45, 60, 70, and 80 degree. I’d then compare the results and assign a purely arbitrary but relative score. The tests would use the same filament, the same gcode, and the same conditions. Only the fan and duct would change. 100% fan after the first layer.EquipmentI’ve got a Monoprice Maker Select V2.1, but the results should be applicable to most printers. I chose Hatchbox PLA as my filament. The axial fan is the stock 40mm fan. The blower is a 50mm 0.15A fan from amazon.The Ducts/Shrouds TestedNo Fan At AllI don’t expect this to work, but I needed a control. Who knows? Maybe this whole fan thing is a myth.Stock Fan and Metal DuctThis comes with the printer, so you’d think it wouldn’t be too bad. It has the advantage of being free and simple. I did the adjustment to get it pointed more squarely at the nozzle, but otherwise no changes.Stock Metal Duct, but with a blower (via adapter made by me)If the stock fan is okay, then this should be even better, right? It involves removing the stock axial fan and replacing it with the fan adapter and blower. It sticks out a bit, but blasts plenty of air. It’s advantage is that it’s an easy mod and the duct won’t melt.DiiiCooler by Pawpawpaw85 with blowerThis is the de-facto standard that everyone says is the duct to use. I’ve had good luck with it, but I had to see how it compared. The advantage is omni-directional cooling. The disadvantage is that it can be a pain to print and can be a headache attaching and removing (which is why they keep melting). Also, it blocks your view so you can’t see that sweet, sweet hot-end-on-build-plate action.DiiiCooler with axial fan (via 40mm fan adapter)The DiiiCooler is great, but is it the fan or the duct (or both)? This test changes the blower for a 40mm axial fan using the adapter that the designer provides. It’s definitely less air flow, but maybe that doesn’t matter. (Spoiler: it matters.)Simple 50mm part cooling fan for wanhao i3/ alike by DevilZModsIf I were a betting man, I would have said this design would win the day. After reading the fan/shroud study, I thought the best plan would be to point your blower at the nozzle and pat yourself on the back for being so clever. The advantage of this design is the high air flow, since there is very little between the fan and nozzle to slow it down. The only disadvantage is that this particular design wants you to glue a little nozzle on the fan. I taped mine and it stayed on fine for one print.WANHAO i3 Fan duct by ThorpedThis is another very popular fan duct that blasts air from three sides. So, it should be 75% as good as the DiiiCooler? That math checks out, but I’ve got to run the test anyway. It has the advantage of giving you a nearly unobstructed view of the hot end. The big disadvantage is the method of mounting, which is on the back of the extruder carriage and, depending on your printer model, might require tapping a mount hole. Also, it isn’t designed for my printer, so I had to use an adapter.ResultsCheck my photos for a visual guide to the notes and scores below. Score is from 0 to 10; 0 being a total disaster and 10 being the perfect print.Fan/Duct TypeNotesScoreNone (control)Not so great. I guess a fan is a good idea after all. Still, not to total failure. 30 and 40 degree overhangs had only minor defects. 60 degrees got tore up real good.2Stock Fan and Metal DuctBetter than nothing, but honestly not that much. The 60 degree angle fared much better, but otherwise only a very moderate improvement.3Blower w/ Stock DuctThis was a huge surprise. 30 and 45 degree overhangs are excellent. 60 and 70 are both pretty good. In fact, 80 isn’t even that bad.8DiiiCooler w/ BlowerTime to shine. This should be the best – and it is… but not by that much. It is only very slightly better than the stock duct with a blower. Everything up to and including 60 degrees is about perfect. There’s a little stutter at 70 and 80 is about as good as you can hope for.9DiiiCooler w/ Axial FanThis was a little shocking also. No better than stock. Also it’s an abomination.3DevilZMods simple 50mmI had such high hopes, but alas… It was about in the middle of the pack. 30 is great, 45 had a little issue, really started to flop on 60 and up. Bummer.7Thorped’s Wanhao i3 Fan ductAgain, I really wanted this one to work out. It’s okay and scored in the top half of the group (such as it is). Nearly the same results as DevilZMods simple 50mm. Hiccup on 45, 60 and 70 are okay, 80 is not that terrible.7.5Summary of all resultsConclusionFirst, throw out your axial fan and get a blower if you haven’t already.Second, the opinion of the masses seems to be true here. The DiiiCooler is the reigning champion, and for good reason.But, there were some surprises. The runner-up was the stock duct with a blower jammed onto it. And it was really close! Why is this? Maybe, since there is so little flow resistance in the stock duct, the air flow is high and spread over a larger zone of influence. Maybe you don’t need omni-directional cooling after all.Also, why didn’t DevilZMods simple 50mm or Thorped’s Wanhao i3 fan duct perform better? I really wanted one of these to come out on top, since they provide an unobstructed view of the hot end. Maybe, in the case of DevilZMods design, the air is too laser-focused and has too small of a zone of influence. Maybe Thorped’s design has too much flow resistance, since the air has to flow around a corner. Both designs seem to direct flow down toward the build plate, rather than in toward the hot end. I just don’t know.Anyway, my plan from now on will be to use the DiiiCooler for PLA and swap in the stock duct with a blower for materials that don’t need much fan (e.g. ABS and PETG). That way, there’s nothing to melt.This was a tough test. You probably shouldn’t be printing overhangs this extreme without supports anyway. I mean, I was printing a 0.4mm width and 0.2mm thickness, so any overhang greater than about 63 degrees means there is NO support from the previous layer. Even 60 degrees is basically trying to print without any support. The fact that some of these ducts worked so well for 70 and 80 degrees is nothing short of amazing.What Next?I’m done for now. You do what you want. But, there are plenty of unanswered questions.”What about MY favorite duct/fan/mod?!” You test it. (Although I kinda want to try the Cobra).What about bridging?How does the cooling rate affect inter-layer adhesion and strength?What about other filament materials?Edit: Words. http://ift.tt/2j9qjDP