Simple question for the Engineers: Mechanical/Electrical fatique in Hard Drives (and DC motors).

[Samsonid]

Hello,

This is a technical issue which I have been pondering for a while, but apparently I do not have enough technical background in physics to do the calculations and give a definitive (scientific) answer.

Here is the Question:
How many hours of continuous operation of a DC motor would yield as much wear and tear on that motor as ONE "Cold Start"?

In other words, how many hours would it take to inflict as much mechanical / electrical wear-and-tear on a motor (at normal load) as compared to the initial surge of curent to the windings of that motor when it is first started?

A real life example: Leaving a motor running for 7 hours overnight would cause a) more, or b) Less electrical/mechanical damage, than performing a single "cold strart" next morning? How about leaving the motor running for a week, compared to a single cold start, how about a month? What is the equivalence?

P.S. I would like to also take into account the damage through mechanical friction in the moving parts (such as those within the bearings)

So, which one would it be? a)More damage in cold start than 7 hours of operation, or b)Less damage in cold start than 7 hours of operation? What is the rule of thumb? At what point is it best for the motor to be shut down instead of leaving it run?

*Let's neglect the electricity cost at this moment, this is purely a motor longevity question.

Thanks
Thomas A.

 

[Mday]

um, this really depends on how the motor was designed, and specifications of the motor.

in other words, there is no answer to your question.

but for 7 hours, i would say a cold start would definitely cause LESS wear\tear on the motor.

 

[Belegost]

I think this would be best determined empirically. Setup two identical motors, set one to run constantly, and have one on a switch that opens and closes on a certain delay. Record the number of times the switching motor starts before it wears out, and the amount of time the constant motor runs before it dies. This should give you a fair average for that particular motor. Of course it would be best to have 5-10 motors running and an equal number being switched to add redundancy.

 

[KIAman]

Depends on what DC motor you are talking about. The classical Brush-type motors recieve their wear at the point of contact of the brushes against the commutator in which case leaving the motor on would cause more wear than cold starts.

If the windings in the motor are not up to par, then a cold start would definately be worse because of the fact that as the motor needs to accelerate, the amp draw gets larger and larger until the speed of the motor is a continuous velocity then the amp draw will drop to the nominal rate. This is why if you create a too huge of a load on the motor (for example not letting it turn), then the amp draw will get so huge to the point where the windings will literally burn up and *poof* a dead motor.

So basically it all depends on the design of the motor. I assume you are asking this question because of the old pros and cons about leaving the computer on all the time or can turn it off and on. Unless you got some really old computer you can safely turn your computer on or off anytime you choose. As others might disagree, it will not shorten your computer's life unless you continuously turn it on and off really fast.

As for a filament bulb, leave it on if you can, but that just gets impractical and they are designed to burn out anyways so yer out of luck there.

 

[Samsonid]

>I assume you are asking this question because of the old pros and cons about leaving the computer on all the time or can turn it off and on

Your assumption is correct ! and Thank you for offering some of your insight into this question.

My biggest concearn has been the effect of cold-starting the hard drive on a daily basis.
(I could care less about the fans, they can be replaced in no time)

The information which keeps me confused is something that I read in Scott Mueller's "Upgrading and repairing PCs" (page 1424). He regards the computer more or less like a light bulb and suggests leaving it ON (ideally on a 24/7 basis) so as to minimize extreme temperature variations caused by the ON/OFF power cycling and maximize longevity.

However, that advice seems to go against my physical intuition; I would expect that minimizing the hours of operation is a better recipe for component longevity.

Then I remembered an article, possibly from a consumer reports magazine, explaining that almost 85% of wear and tear in automobile engines occures during those few seconds of ignition; since the lubricant is still not fully circulating. The corrolation here is that the motors inside the PC (especialy the hard drive) also have mechanical parts in friction. The hard drive in my machine has fluid dynamic bearings. Does this mean that at every shut-down the fluid film raptures and thus 85% of the damage will occur tommorow when the film has to re-establish itself?

Personaly, I think there has to be an optimum rate of on off cycles that only an experienced engineer can come up with by solving a few sets of equations.

The results can be surprising since I remember vaguely from the engineering courses in college that solving equations in order to "Optimize" a condition in a system would often yield results that would challenge common sense.

The hard drive manufacturers do offer the MTBF and start-stop cycles data, but I am still not quite sure how to interpret those values to answer my questions (if these values are even relevant for this matter).

Belegost's post suggesting the empirical solution would, certainly, give out the answer without a doubt. Unfortunately I wish there was enough time in my schedule for this kind of staff.

Well, I am hoping that this kind of information is readily availabe somewhere without having to resort to empirical tests or long hours of search. Someone must have done these tests already; what the heck, we have sent a man to the moon... which is probably a lot harder.

 

[Samsonid]

Well, the answer I am realy looking at is along these lines:

Someone to come up and say, hey Thomas we have done very extensive tests over here at Seagate Laboratories over the years and all our experimental data suggest that,

*A single cold start of a hard drive is EQUIVALENT to 5.465 hours of normal opeartion.

So, if you plan to use your computer 7 hours from now, it is best to turn it off so as to save 2 hours out of your hard drive. If you plan to use it only 3 hours from now, of course it is best to leave it on since starting it up 3 hours later would be as if you had it on for 5.465 hours instead.

->5.465 was a ficticious number just for the example.

 

[KIAman]

Hmm, I'd suspect that unless your room tempeture is about freezing, and if your harddrives operating temp is 50 degrees C, then i really dont see any "extreme tempertures differences."

About the only thing about the computer that resembles anything like the light bulb is the moniter.

The reason why the bulb anolagy is so scarey is that when bulbs turn on, the filament inside the bulb is a resistor that the amp draw is great enough to produce light, also a great deal of heat. Now here is a pefrect example of temperature difference: when its off it is room temperature (or atleast gets there) when on it could reach up to 100 C or greater. Now that is extreme temperature difference and when things heat up, they expand, so the constant expansion and contraction of the filament will cause it to die.

Harddrives have nothing in common with light bulbs.

So my point is, the make and quality of the drive will have a MUCH bigger impact on the life of the harddrive, rather than the turning it on or off, unless you are intentionally trying to break it by rapidly turning it off and on.

Most people's main reason for leaving their computer's on is simply because they offer some sort of service or recieve service all the time, or are part of the network, or any other reason that relates to the fact that computers are still usefull after the primary user has left.