Beam of air: Extending the range of a fan

[Anachronist]

TL;DR Summary

Need advice on laminar vs turbulent flow techniques for focusing air from a fan at a distance

If one Googles for "laminar flow nozzle", one finds many interesting tutorials on creating a nozzle for a laminar water jet; a stream of water that remains coherent over a long distance without breaking up. These typically consist of a large-diameter tube with regions inside (like sponges and straws) to remove turbulence, and a knife-edge circular orifice from which the water jet emerges.

I want to know if there is something similar can be done with air from a fan, to extend the air jet as far as possible with minimal spread. Because the medium of the flow is the same as the medium in which it flows, I guess this would be analogous to making a long-distance water jet underwater, only this is an air jet in air. I am skeptical that those laminar flow water jets would perform well underwater.

Would a swirling flow hold together better as the air propagates down the direction of flow? I could make an axial fan cover that induces swirl.

For a non-swirling flow, would it be best to have a gentle radial gradient in the flow, with the highest velocity at the center, reducing gradually toward the edge? This could be done with something like the bliss caps on the exhaust stacks of some Navy ships.

Or would it be best to use the laminar water jet principle and strive to make every point in the flow column move at the same velocity? This would create a large velocity shear between the flow and still air, creating turbulence. A squirrel-cage fan rather than an axial fan might be better for this idea.

I would test this by making different covers for a computer fan and probably running dry-ice vapor through it to visualize the flow.

 

[berkeman]

Does the air stream have to be continuous or can it be pulsed?

If it can be pulsed, maybe it could be based on the old toy air blaster bazookas and pistols. You could figure out why they were able to propagate the blast of air so far with little dissipation, and then extend that to some reasonable pulse rate (2-4Hz maybe?).

https://alphadrome.net/data/albums/userpics/10004/AGENT_ZERO-M_SONIC_BLASTER.jpg

 

[Anachronist]

Does the air stream have to be continuous or can it be pulsed?

Ah, I remember those pulse toys! In my experience they were used mostly to annoy other people! It looks like it's using a nozzle with an exponentially decreasing radius, over a large diaphragm.

Sorry, I thought that might have been clear when I said this is for a fan. It would be continuous. My objective is to cool a hot object from a distance with minimal air blowing on other stuff adjacent to the object.

I didn't want to distract from the problem with my specific application, but specifically, I am experimentally trying to find a way to cool the extruder stepper motor on my 3D printer (https://all3dp.com/1/original-prusa-i3-mk3s-plus-review-3d-printer-specs/) without adding mass to the print head on which the motor is mounted. I figured I could install a 40mm diameter fan on the same Z-axis mount that the extruder gantry is on. Extra mass on the Z axis doesn't matter, but it does on the extruder that moves along the X-axis. The fan would then always be level with the extruder motor, blowing on it, but at varying distances (5-28 cm) as the extruder moves back and forth.

I don't want the fan to blow on anything except the extruder motor, to maximize cooling efficiency. I'd like to see if I can make a stream of air that maintains itself at about 40mm in diameter over a distance of 30cm before spreading out.

An axial fan isn't mandatory. A blower-type fan could work too. I thought maybe if a swirling flow would be optimal (and I really don't know) then an axial fan would be better.

 

[Anachronist]

After some searching, I found that there are compressed-air nozzles designed to create a laminar stream of air to exert a focused blowing force over an extended distance (example 1, example 2). They apparently surround a high-speed central jet of air with slower-moving protective sheath of air. If the same principal applies to no-pressure (the air from a fan is basically incompressible flow), then the Navy bliss-cap idea in my first post might do the trick. I expect I'd have to remove swirl or turbulence from the air jet too.

 

[Lnewqban]

Speed of airflow is your friend for achieving your goal.
Nevertheless, it is difficult to reach high speed with a computer fan (or any axial type), as high speed means relative high discharge pressure: a centrifugal type fan could do a better job.

The ejection effect is always present and the moving air will suck surrounding air into a diameter growing airstream.
Please, see:
https://mechanicalengineeringsite.com/ejector-working-principle/

 

[Baluncore]

If you are able to deliver the air efficiently and accurately, then there may be an advantage in delivering denser colder air, for example from a vortex tube. https://en.wikipedia.org/wiki/Vortex_tube

Since HP air can be delivered down a very thin flexible tube, I would make a stepper motor enclosure from aluminium shim, then inject the cooling air into the space around the motor. The thin supply tube could also act as an inter-cooler.

A vortex tube might be mounted integral with the stepper enclosure.
The vortex tube waste outlet is hot, do you need to heat the extruder of your 3D printer?

 

[Anachronist]

Speed of airflow is your friend for achieving your goal. ... a centrifugal type fan could do a better job. The ejection effect is always present and the moving air will suck surrounding air into a diameter growing airstream.

That makes sense. I am thinking now that a centrifugal fan with a nozzle that sucks in air to control the shear around the central flow might help, in the way that the high-pressure nozzles linked above create a protective sheath to keep the air jet collimated.

If you are able to deliver the air efficiently and accurately, then there may be an advantage in delivering denser colder air, for example from a vortex tube. https://en.wikipedia.org/wiki/Vortex_tube
...A vortex tube might be mounted integral with the stepper enclosure. The vortex tube waste outlet is hot, do you need to heat the extruder of your 3D printer?

In all my years I had never heard of the vortex tube to separate compressed air into low and high temperature streams. I can think of uses for that, but not for this project. I don't really have a source of high-pressure air in this case, and I also want to avoid adding any mass to the extruder (which holds the stepper motor). Aiming the waste heat at the extruder nozzle would also aim the heat at the part being printed, which needs to cool quickly as the molten plastic leaves the nozzle.

 

[sophiecentaur]

I am experimentally trying to find a way to cool the extruder stepper motor on my 3D printer

Did you consider using a long helical plastic tube, suspended from above and attached to the 'head'. It could be long and self supporting and impose virtually no load on the mechanism. There are all sorts of tubes available for delivering liquids and gases and some are very lightweight.
I understand that many 3D printers are a bit noisy but, if additional noise is an issue then this alternative could be quieter than high speed air jets.

 

[Anachronist]

Did you consider using a long helical plastic tube, suspended from above and attached to the 'head'. It could be long and self supporting and impose virtually no load on the mechanism. There are all sorts of tubes available for delivering liquids and gases and some are very lightweight.
I understand that many 3D printers are a bit noisy but, if additional noise is an issue then this alternative could be quieter than high speed air jets.

I'm having trouble imagining how that would work. This is what the extruder ("print head") looks like. The extruder motor that I want to cool is the cubical-looking object with the words on it. That motor is what forces the plastic filament (not shown) into the hot end at the bottom and out the nozzle. The angled round-ish object at the bottom is a centrifugal fan that blows air on the melted plastic coming out of the nozzle so that it hardens quickly.

I want to avoid adding any mass or inertia to the print head. That's why I was thinking of mounting a small fan with an appropriate collimator to one of the orange parts on each end of the X-axis (the two rails that the print head slides on left to right). I can re-print one of those orange parts to include a fan mount. The orange parts move in the Z direction (you can see the long vertical worm gear on the right edge of the image). It's OK to add more mass to the Z axis. The collimated "beam of air" would be aimed at the extruder motor at all times, and move vertically with it because both fan and extruder are attached to the Z axis motion.

Now, when you mention a helical plastic tube suspended from above, I am imagining a flexible coiled plastic tube attached to a shroud around the extruder motor, with the shroud guiding the air from the tube around it. Is that what you meant?

You might wonder why I want to blow air on this motor. 95% of the time it isn't necessary. However, over a print run that lasts many hours with many filament retractions, the motor's own self-generated heat creeps along the motor shaft and into the drive gear that's controlling the filament. This part cannot get too hot or the filament softens at the drive gear, and cannot be pushed cleanly into the hot end. The heat from the hot end is isolated from the drive gear by that aluminum heatsink behind the blower fan, and it has a separate fan that you cannot see in this image, blowing across that heatsink. That's quite effective. It's the heat creep from the extruder motor's self-generated heat that I'm trying to prevent.

This whole project is more of an excuse to figure out a way to focus a jet of air over a longer distance than it would normally have just coming out of a fan.

By the way, this printer is exceptionally quiet. We're in a small apartment with a combined living room and dining room. The printer is 4 or 5 meters away from the dining room table, and we don't even notice when it's running while we have dinner.

 

[jrmichler]

I am thinking now that a centrifugal fan with a nozzle

A properly sized centrifugal blower with a nozzle will blow air across the width of most small printers. A blower something like this, which I randomly picked from the Digikey catalog:

For this blower, I would size a nozzle for 0.10 CFM at 4.5 inches water pressure. I made no attempt to pick the right size blower for your application, this is the first one that came up. At the price of $15.00 plus shipping, you can easily buy two or three to find the right size by experimentation. Just size the nozzle for each blower so the the blower operates near the middle of the fan curve.

 

[hutchphd]

Having just blown the leaves around my lawn, I guarantee a centrifugal blower with a modest nozzle will certainly do the job directionally.
/

 

[Anachronist]

OK, a blower fan then. I'd probably use this one. I know such a fan can easily blow air across the width of the printer. The question is, how does the air jet spread? Ideally, the amount of air impinging on the motor housing should be about the same 5 cm away or 30 cm away.

I can design a nozzle for the blower to try to focus the jet, using ideas from the high-pressure nozzles I linked earlier in this thread. How would I visualize the spread of the air coming out, and the turbulence around the jet? I tried my hand at CFD a couple years ago and realized that I'd need to go to college again to learn enough to model this thing. I have the equipment needed for Schlieren photography but my 6" mirror isn't really big enough. Dry ice + water vapor fed into the blower might be the easiest.

 

[hutchphd]

I believe you are overthinking this. There will always be some divergence in the flow. I would build a cardboard and duct tape prototype (KISS) and proceed accordingly. Your ability to micromanage this flow will not be that refined. Then build a final version with materials of your choosing.

 

[sophiecentaur]

I want to avoid adding any mass or inertia to the print head.

That's certainly a relevant consideration. There is, already a fairly hefty connecting 'umbilical', carrying the filament and power. The spiral tube I was envisaging would be very lightweight and flexible and with an equilibrium position with the nozzle at the mid point of head motions so not a lot of added load or inertia. It could be suspended from a long arm, vertically above the centre of the movement of the head. The advantage would be that all the cooling air would be directed directly at the right spot. Normally, I'd say that would look a bit bizarre but 3D printers tend to look that way, in any case.
I have a friend who bought a 3D printer, some while ago and he now realizes the steepness of the learning curve that's involved so you have my sympathy. All his problems seems to be thermal and yours are the same.

I can see that you find your idea very attractive and you'd rather do it 'your way'. I'm totally fine with that - been there - but if you want a working system, you could consider alternatives.

 

[hutchphd]

I can see that you find your idea very attractive and you'd rather do it 'your way'. I'm totally fine with that - been there - but if you want a working system, you could consider alternatives.

I will echo this and point out the most reasonable engineering solution seems to me to be:
https://www.amazon.com/dp/B082MY2MX3/?tag=pfamazon01-20

This will likely do everything you need, but has zero elegance. Probably quieter too. Enjoy.

 

[Pratyeka]

a much simpler solution to prevent heat from propagating along the shaft of the stepper motor to the gear is to insert a coupler between the shaft and the gear made of non-heat conducting material. That would mean modifying the position of the motor, since you have to extend the shaft. But I have made prints that lasted more than 50hrs with no problems without any form of extruder motor cooling, even in ambient temperature above 30 Celsius.

 

[sophiecentaur]

to insert a coupler between the shaft and the gear made of non-heat conducting material.

10/10! Nice.
But isn't that the sort of thing that we should expect from state of the art equipment? Perhaps the low end producers just don't do extended tests. My mate gets a lot of 'stringing' from the end of one item to the start of another; same sort of problem that just hadn't been spotted.

Perhaps the top end stuff would include that design feature.

 

[Pratyeka]

To get rid of stringing requires patience and fine tuning of retraction, coasting at end and wiping distance, all depending on the material being printed, speed and temperature. I have successfully reduced stringing to none with most PLA and PETG I have worked with.