Cure noisy fan

[Paul Komski]

What's the best lubricant (if any) for a noisy heat sink fan - (an Arctic Cooling Alpine 64 (http://www.frostytech.com/articleview.cfm?articleID=1922) in this specific case)? I have the case open and till today a couple of light taps cured it but now it has an almost continuous loud whine.

The last time this happened to me I added a drop of 3-in-1 oil and it just made the whole situation worse; possibly there was some dust that it sort of made a "glue" with.

Any pros? any cons? ... that could affect its longevity?

[david eaton]

I have had quite good results with WD40.

[Sylvander]

I stopped the fan in my external USB HDD enclosure from whining by very slightly loosening the screws that hold it to the casing.

Why do this?
Here's the theory:
1. Rotating devices should generally be attached to their static mounting point using "Flexible mounts" [I was taught this in my Engineering class].
With sophisticated flexible mounts, damping/dampers would be included, but not in a simple setup [such as this].
Why?
(a) Any out-of-balance creates a tendency to vibrate. That is, the centre-of-mass does not lie on the central axis of the shaft.
(b) It's best to allow the device to move slightly so that it now rotates around the centre of mass, rather than the central axis of the shaft.
(c) So now there is no out-of-balance, but instead the mount points are moving back and forth, hence [if you make the mount points rigid] the vibrational forces acting upon the mount and being transmitted through to the structure/objects contiguous with the mount points.
(d) If you introduce flexibility, the device vibrates slightly back and forth, but that causes little to no problem as it does not transmit to the wider structure with the consequent noise.

2. Not only that, but with a rigid mount...
If the out-of-balance is in the rotating part [the rotor]...
And the bearing is rigid...
The shaft will tend to bounce around inside the shaft [any wear producing a small gap?]
That is probably what is causing the screeching, hence why lubrication produced no improvement.

[Whyzman]

I would have to say that your Alpine apparently cut classes when fluid bearing "Characteristics and Principles of Operation" were discussed. The White Papers (http://www.hitachigst.com/tech/techlib.nsf/techdocs/089C4B963AEE9A6F86256D340075052F/$file/FD_White_Paper_FINAL.pdf) point out that this simply ought not be happening. By design these are self-correcting: It is also very difficult to make a mechanical bearing which is atomically smooth and round; and mechanical bearings deform in high-speed operation due to centripetal force. In contrast, fluid bearings self-correct for minor imperfections.Fluid bearings generally have very low friction -- far better than mechanical bearings.Fluid bearings are typically quieter and smoother (more consistent friction) than rolling-element bearings.The bottom two quotes are taken from Wikipedia...

2. Not only that, but with a rigid mount...
If the out-of-balance is in the rotating part [the rotor]...
And the bearing is rigid...
The shaft will tend to bounce around inside the shaft [any wear producing a small gap?]
That is probably what is causing the screeching, hence why lubrication produced no improvement.However, according to this quote from Wikipedia: It might seem that bearing stiffness, as with maximum design load, would be a simple function of average fluid pressure and the bearing surface area. In practice, when bearing surfaces are pressed together, the fluid outflow is greatly constricted. This significantly increases the pressure of the fluid between the bearing faces. As fluid bearing faces are comparatively large areas, even small fluid pressure differences cause large restoring forces, maintaining the gap. [bold mine] Isn't the gap a good thing??

My suspicion is that the fluid has escaped whatever containment device was used. Notwithstanding all the carefully instituted compensating measures in the Alpine 64, all would be for naught if the lubricating fluid were successful in its design to escape its confine...

If the seals containing the fluid are compromised, I suspect that any measure other than replacement of the unit might be but a short term fix...

[minus-sign]

graphite. liquid lube in these things turns to gunk way to easy.

[Whyzman]

You may have recourse...have you visited the Arctic Cooling website??

Under Support on the website appear Technical Support and RMA : http://www.arctic-cooling.com/catalog/static_info.php?tid=RMA_PROCEDURE It appears they have no manufacturer's warranty per se, but parts are available to consumers so they don't have to ship the whole works back etc. It would appear that the fan could be obtained and swapped...

[Sylvander]

1. "Isn't the gap a good thing??"
(a) Well naturally you cannot afford to have an "Interference Fit (http://en.wikipedia.org/wiki/Interference_fit)", that would produce too much friction; you need instead to have just enough of a gap to have the ideal thickness of lubricant.
My recollection is that the surface of the rotating shaft aqua-planes on a thin layer of lubricant.
But if the gap were to become too great [due to excessive wear?], there would then be an empty air-gap with no firm hold to position the shaft.
Add to that an out of balance of the rotating part, and the shaft [and whatever is attached to it] is able to bounce around inside the bearing.
If you allow the bearing [plus housing] to move [make it cushioned (and damped)], the shaft will move around in an eccentric [off center] fashion, but the bearing will be allowed to move with it rather than stay rigid and therefore no longer make the shaft bounce.
Once a bounce begins it tends to continue, and since the conditions are right for creating bounce, why would it stop?
As soon as you take away the rigidity you to tend to prevent bounce and so that situation should cease [unless the gap is REALLY big].

(b) Even if things aren't so bad as in (a) above [there is no excessive gap between shaft and bearing], it's still good to allow the rotating device to rotate around the center of mass.

(c) It's a known phenomena of rotating objects that they naturally tend to DEFORM [if they can] so as to produce rotation around the center of mass, which then tends to eliminated out-of-balance.
DON'T FIGHT IT, ALLOW OR ASSIST IT.
Put another way:
The object tends to deform [no elastic object (e.g. Steel is elastic) is of infinite rigidity] so as to make the center of mass move to become located on the the axis of rotation.

[Whyzman]

I didn't mean to infer a "large" gap. However, you've expressed my thoughts exactly...and I did stay at a Holiday Inn express last night... ;)

[Paul Komski]

When I said it just made the whole situation worse because it turned to a sort of "glue" - I should have made it clear that it almost stopped the fan from turning altogether rather than make the whining worse. It is almost impossible to get at the spindle directly and I don't see a way to dis-assemble the fan unit without forcing the spindle out from the electric motor. I have taken the fan off the heat sink and I might just try immersing it in petrol and then let the whole thing dry out. I have replaced the original but this could come in handy as a spare part - particularly for a deaf user!

I have seen advocates both for and against WD40 because it removes lubricant as well as "clean things up". Petrol has some lubricant properties but I suppose that so does WD40. Graphite grease might be good if there was some way of applying it.

[minus-sign]

Spraying/dipping/brushing: Dispersion of solid lubricant as an additive in oil, water or grease is most common used. For parts that are inaccessible for lubrication after assembly a dry film lubricant can be sprayed. After the solvent evaporates, the coating cures at room temperature to form a solid lubricant. Pastes are grease like lubricants containing a high percentage of solid lubricants used for assembly and lubrication of highly loaded, slow moving parts. Black pastes generally contain MoS2. For high temperatures above 500°C pastes are composed on the basis of metal powders to protect metal parts from oxidation necessary to facilitate disassembly of threaded connections and other assemblies.

Free powders: Dry-powder thumbling is an effective application method. The bonding can be improved by priory phosphating the substrate. Use of free powders has its limitations, since adhesion of the solid particles to the substrate is usually insufficient to provide any service life in continuous applications. However, to improve running-in conditions or in metal forming processes a short duration of the improved slide conditions may suffice.

AF-coatings: Anti-friction coatings are "lubricating paints" consisting of fine particles of lubricating pigments, such as molydisulfide, PTFE or graphite, blended with a binder. After application and proper curing, these lubricants bond to the metal surface and form a dark gray solid film. Many dry film lubricants also contain special rust inhibitors which offer exceptional corrosion protection. Most long wearing films are of the bonded type but are still restricted to applications where sliding distances are not too long. AF-coatings are applied where fretting and galling is a problem (such as splines, universal joints and keyed bearings), where operating pressures exceed the load-bearing capacities of ordinary oils and greases, where smooth running in is desired (piston, camshaft), where clean operation is desired (AF-coatings will not collect dirt and debris like greases and oils), where parts may be stored for long periods of time.

I hate wasting good parts. Had a heatsink for years; whenever I fired up ventrillo people would ask me if I was running a computer or a nuclear power plant. I didn't mind, being 'deef'. Hope it proves a good spare.