helium filled hard disk drives

[Rick in PA]

An interesting article from TechRepublic.

I found the concept intriguing. My 2 cents, the trick will be keeping the helium in the drive. Those are some mighty tiny, slippery molecules.

 

[Shannow]

LOL

 

[SwampSurvivor]

Isn't there a helium shortage?

 

[2009Edge]

Using it in a large capacity hard drive puts the helium (H2) to better use than in balloons.

 

[Astro14]

10 years ago, my neighbor, who worked for Seagate, described the head of a disk drive as the equivalent of a 747 flying at 10 feet...

I can only imagine how precise the tolerances are in a disk drive now...and have no doubt that boundary layer effects must be accounted for...

 

[JHZR2]

Don't they run hydrogen or helium in electric plant generators to help with heat dissipation? I'd guess that HDDs are somewhat sealed as-is, but it will be interesting to see how they combat loss. Of course they likely only want a 3-5 year life span anyway...

 

[scurvy]

Originally Posted By: JHZR2
Don't they run hydrogen or helium in electric plant generators to help with heat dissipation? I'd guess that HDDs are somewhat sealed as-is, but it will be interesting to see how they combat loss. Of course they likely only want a 3-5 year life span anyway...


Generator rotors in most power plants run in a hydrogen atmosphere.

It's not just to cool the rotor (although it exceeds beautifully in that capability - roughly 10x better cooling capability than air) but also to give it a low drag environment to reduce windage losses.

The big units cool the generator stators with water as the stator bars themselves are hollow copper tubes.

When I worked for a generator/turbine repair shop, the hydrogen seal halves for generator casings were hand lapped together. They'd be checked for contact first with bluing, hand scraped to get close, reblued/rechecked, then cleaned well, their mating surfaces spread with lapping compound and a machinist would sit in front of it and move one seal against another in a big horizontal arc, over and over. They called it 'driving the school bus'. Then they'd clean them, reblue & recheck for contact.

 

[HerrStig]

Solid state drives will solve the problem for good.

 

[Nick R]

Originally Posted By: HerrStig
Solid state drives will solve the problem for good.


Pretty much this.

 

[JTK]

It is amazing how small and light H2 or He molecules are. I can see it at work if I've got to purge a section of our H2 plant out of service so we can work on it. Pressurizing and blowing-down a vessel in H2 service might take 5 min. That same vessel switched over to N2 gas can take an hour to blow down.

Joel

 

[PandaBear]

Mechanical hard drive is not sealed. They have a vent hole that said DO NOT BLOCK.

Air is needed to keep the head flying, and the fly height are controlled by the arm spring rate and an electrically heated resistor that pushes the head downward.

Aerodynamic and turbulence is mainly a head design issue (to be precise the air bearing of the head). It needs to handle the varying altitude and temperature, as well as the difference in head speed along inner track and outer track part of the platter.

Sealing a drive is asking for trouble down the road when the seal starts to leak when the drive heats up and air pressure build up. This will eliminate any aerodynamic benefit of having a "lighter" air.

The bigger problem IMO in having more platter is the media and arm rigidity. This He drive seems like a concept rather than a production ready technology.

 

[JimPghPA]

Years ago there was an article in Popular Science or Popular Mechanics about using hydrogen or helium in some kind of lifting for flight device. The article said that a very small amount of a gas with a larger molecule such as carbondioxide or carbonmonoxide was added to the lifting gas. The larger molecules not being of enough concentration to significantly reduce the lift, but being large enough to get stuck in the very small pores of the containment material, significantly reduces the loss rate of the hydrogen or helium through the containment material.

Helium is a monatomic gas and therefor is very good at conducting heat.

Neon, and Argon are also monatomic and can be usable for providing better heat transfer compared to air.

 

[mechtech2]

Originally Posted By: Nick R
Originally Posted By: HerrStig
Solid state drives will solve the problem for good.


Pretty much this.


Way pretty much this!

 

[JHZR2]

Originally Posted By: PandaBear


Air is needed to keep the head flying, and the fly height are controlled by the arm spring rate and an electrically heated resistor that pushes the head downward.



With the right bearing design, a lot of mass can be supported with sub-ambient pressure light gas, and minimal windage loss.

 

[Shannow]

Originally Posted By: scurvy
Originally Posted By: JHZR2
Don't they run hydrogen or helium in electric plant generators to help with heat dissipation? I'd guess that HDDs are somewhat sealed as-is, but it will be interesting to see how they combat loss. Of course they likely only want a 3-5 year life span anyway...


Generator rotors in most power plants run in a hydrogen atmosphere.

It's not just to cool the rotor (although it exceeds beautifully in that capability - roughly 10x better cooling capability than air) but also to give it a low drag environment to reduce windage losses.

The big units cool the generator stators with water as the stator bars themselves are hollow copper tubes.

When I worked for a generator/turbine repair shop, the hydrogen seal halves for generator casings were hand lapped together. They'd be checked for contact first with bluing, hand scraped to get close, reblued/rechecked, then cleaned well, their mating surfaces spread with lapping compound and a machinist would sit in front of it and move one seal against another in a big horizontal arc, over and over. They called it 'driving the school bus'. Then they'd clean them, reblue & recheck for contact.


:manhug:

axial hydrogen seals were invented by the bloke who invented purgatory...or pushing the wheel of dharma.

hydrogen in a hard drive makes sense, as it can be electrolytically produced, as a top-up. Seal leaks, drag, and sealed for life on helium don't seem like 40,000 hours of life (shelf, service, or other) propositions.

 

[PandaBear]

Originally Posted By: JHZR2
With the right bearing design, a lot of mass can be supported with sub-ambient pressure light gas, and minimal windage loss.


Agree, but most of the power consumed by the drive after spin up is the seeking motion, and a heavier gas with heavier spring rated arm will give you better shock tolerance for the same fly height, therefore less crashes and better reliability, or you can lower the fly height for the same reliability to increase recording density.

Originally Posted By: Shannow
hydrogen in a hard drive makes sense, as it can be electrolytically produced, as a top-up. Seal leaks, drag, and sealed for life on helium don't seem like 40,000 hours of life (shelf, service, or other) propositions.


For a $600 hard drive it might make sense to put a H generator in, but not for a $40 hard drive. Also it is hard to tell how long a hard drive will sit on the shelf as a cold spare before it is used, and therefore it is hard to rely on top off at run time. Relying on a generator also means 1 additional part to break, and additional heat and power consumption (which is probably the only benefit of using H/He in a drive).

Their goal of reducing between platters height can be achieved much easier with other methods, and unless they are put in a very stable fixtures, external shocks can crash the head much easier with the thinner gas with reduced ride height.

More rigid material for the platter, better aerodynamics on the arm, better aerodynamics on the internal of the drive case, etc would work better than using He in the drive.

 

[Rick in PA]

They're in production. Looks like they figured out how to seal them (a five-year warranty):
The Register

 

[Subdued]

HGST?

They were the brains behind the maligned IBM DeskStar 75GXP (AKA DeathStar) which ended up being the most unreliable drive in desktop storage history. The problem there is they were also one of the fastest desktop drives of their time as well.

I have no doubt their quality has improved but the mention of HGST always makes me giggle al ittle bit

 

[ueberooo]

Originally Posted By: Rick in PA
They're in production. Looks like they figured out how to seal them (a five-year warranty):
The Register


Looks awesome. I'd take these 6 TB suckers over any solid state junk, anytime.

If plumbers can keep boilers and radiators from exploding, then a team of highly paid folks with a bunch of engineering degrees can solve this problem of pressure variation.

 

[PandaBear]

Originally Posted By: Subdued
HGST?

They were the brains behind the maligned IBM DeskStar 75GXP (AKA DeathStar) which ended up being the most unreliable drive in desktop storage history. The problem there is they were also one of the fastest desktop drives of their time as well.

I have no doubt their quality has improved but the mention of HGST always makes me giggle al ittle bit


Hard drive in the enterprise storage vs consumer computer are different quality and grade (and speed and price). Why they keep the HGST brand has little meaning on who they are selling to (most likely the server farms). I have not heard of any companies selling bad SCSI drives ever (but they tends to pay $200-500 each to begin with).