Archived from groups: alt.video.ptv.tivo (
More info?)
w_tom wrote:
> Demonstrated was how myths are created. Others claim power
> cycling is destructive - using speculation. Not one could be
> bothered to provide a single number.
As opposed to your one set of completely unsubstantiated numbers from an
unverified source? Care to try again, Yoda?
I provided no specific numbers because I didn't design these bearing and
have no knowledge of the design requirements, load limits, and lifecycle
costs associated with them. Given those numbers I could likely work up
some numbers, but they would be strictly probabilistic due to the
failure profile of this type of design. An extensive stochastic
analysis would be required to arrive at some sort of probability
function of failure, and I don't see you offering to pay my salary for that.
> All hard drive manufactures provide numbers for life
> expectancy and power cycling. No numbers were posted here: as
> junk scientists routinely do. Randy S posted:
>
>>Hard drives use bearings at the hub, so there is definitely
>>precedent for start/stop cycles to adversely affect their
>>operation ...
>
> This only proves that a hard drive might fail after power
> cycling every day for 10,000 years. I noted how junk
> scientists speculate without numbers. Then Randy S goes right
> ahead again; using personal speculations as if they were facts
> - without any numbers. His only proof? Personal insults.
You make conclusions from nothing. What I said *proved* nothing, and I
did not say it did. All I set out to demonstrate was that you *also*
proved nothing. Simply stating something as fact does not make it so.
How does saying a hard drive might be more prone to failure with
frequent power cycling somehow bring up some assinine 10,000 years
number? That's your junk number not mine!
> Randy S - previously demonstrated was how those without
> basic knowledge create facts from speculation: provide no
> numbers. Then you went right ahead and did it again. Clearly
> you are only a mechanic. You provided no numbers for jet
> engine bearings either. Somehow you just know - which is why
> we are happy the engineers are making the decisions for you.
> What do they first learn - and you do not? The numbers.
4 years of undergrad work in aerospace and mechanical engineering and 4
more in graduate engineering work makes me more than a mechanic, thank
you. I didn't *maintain* bearings in jet engines, I designed them for 4
years after my undergrad degree. Besides, that's an assinine statement
as well. I worked besides line mechanics with 30 years of experience
who knew a *hell* of a lot about how bearings worked and how they failed
without having a college degree in engineering. I'd take advice from
any one of them well before I'd take yours.
> Using Randy S reasoning, then jet engine bearings will
> immediately fail after only a few cold starts. Since he
> posted no numbers, then this conclusion is accurate based upon
> what Randy S posted. We must never power off jet engines - if
> Randy S logic is valid. Or we cite Randy S as a purveyor of
> junk science reasoning.
? I said no such thing. We designed engine bearings taking start/stop
cycles into consideration, they were one of the largest causes of wear
in the lifecycle. The useful lifetime calculation for an engine bearing
included *both* the engine run time *and* the number of start/stop
cycles it had experienced. The relative weighting of those factors
would vary between applications, but they are both certainly significant.
> Most drives spec around 100,000 power cycles. This one IBM
> drive had a particularly low number of 40,000 cycles.
> Believe power cycling to be so destructive? Then post numbers
> without wildly speculating. And unlike Randy S, learn to post
> numbers of science without silly insults.
And I you before to provide some verifiable numbers, preferably
independent and unbiased. By refusing to do so you have basically
proven that you are making these numbers up. I see no reason to believe
otherwise.
How about I take literally 20 seconds and google bearing wear start-up?
Hmm, look what the first link leads to:
http
/www.bearings.machinedesign.com/guiEdits/Content/BDE_6_1/bdemech6_21.aspx
Let's take a quote, shall we?
"Hydrodynamic bearings wear very slowly. Wear occurs during start-up and
slowdown when speed is too low to produce sufficient fluid pressure to
support the bearing surfaces on a lubricant film."
Wow! Where have I heard a statement like that before? Maybe in my last
post?
> Claims about power cycling are classic junk science
> reasoning when the author provides no numbers; never even
> bothered to cite specifications from disk drive
> manufacturers. No numbers means he is posting junk science -
> old wive's tales - myths.
>
> "OMG" is how drug addicts reply - rather than post numbers.
If this is true, why don't you at least *attempt* to refute the logic I
presented? The scary thing is that it's not even just logic, this comes
from several years of work in the field. Why don't you ask an
automobile mechanic or engineer at what point does most wear and tear on
an automobile engine occur? They'll confirm that it is at start-up,
before the oil pump can provide enough oil pressure to circulate
lubricant in the engine. Take a look at this page:
http/engineparts.com/motorhead/techstuff/bearingfailures.htm
Notice how, at startup, the shaft and the race are physically in
contact? This is when 80-90% of wear occurs. Once started, the oil
wedge generation by rotation seperates the parts and has an extremely
low rate of wear.
Engines that are designed to cycle often (like your typical car engine)
have to be designed with bearings that can take a lot more wear, or can
be easily and more often replaced. Engines that operate continuously
(like turbine generators) have far fewer start/stop cycles and can be
designed with much tighter clearances and lower wear profiles. This can
increase efficiency greatly.
Can you even *name* the parts of a bearing? Can you tell me the
difference between a ball, roller, and journal bearing? Can you tell me
where you would use a roller bearing in place of a ball bearing, and
why? How about a caged vs. a cageless design? Can you actually
demonstrate any knowledge in this field at all other than your so-called
*junk* science? I'm afraid you're just another Internet wacko who
subscribes to innumerable conspiracy theories, I suppose you'll be
telling us when the aliens will be landing next.
Startup/shutdown cycles*are* uncontrovertably a factor in bearing life.
Are they significant enough to avoid turning a pc off at night? I
can't say, and, guess what, neither can you. In fact the answer may
differ even from drive to drive. But stop feeding us this drivel about
how somehow *YOU* have the answer.
Randy S.
Facebook
Twitter