L10 Bearing Life and Mech Seals 4

[Milkboy]

Hey Group

Bearing manfs have L10 for bearings

Why would mech seal manfs not do the same?


TIA



-
Milkboy

 

[MikeHalloran]

Because the failure mechanisms are different.



Mike Halloran
Pembroke Pines, FL, USA

 

[Milkboy]

Could you expand on that Mike?

Ive got a guy asking for the life expectancy of a seals and he expect the manf. to have data similar to L10

Its just not done, but Im stumbling to find the words
to explain why its not done for seals and it is for bearings.

I can see his point. Take a Type 1 Seal and run it for years and get data in the form of L10

Dry Run Bearing = Fail
Dry Run Seal = Fail

Worn Out Surfaces = Fail > log the hours
Worn Out Faces (Carb) = Fail >log the hours


I have never seen a life expectacy for a mech seal given in term of testing etc, only expectancies from experience.
My guy wants some kind of cert from recorded data

Any advice you can offer

Thanks



-
Milkboy

 

[Artisi]

Can the guy asking for the life expectancy of mechanical seals in terms of number of hours run under the operating conditions give you a guarantee of absolutely clean product at the seal faces, exact alignment and exact true running of the pump shaft ie, free of any vibrations, end thrusts, axial deflection, constant temperature of the pumped product etc etc. Unlikely - so how can he expect you to give absolute run life.
As Mike has pointed out - you are not comparing apples with apples

 

[iainuts]

Can the guy asking for the life expectancy of mechanical seals in terms of number of hours run under the operating conditions give you a guarantee of absolutely clean product at the seal faces, exact alignment and exact true running of the pump shaft ie, free of any vibrations, end thrusts, axial deflection, constant temperature of the pumped product etc etc.

Same exact issues can be said about bearings. I think it's a great question, and wouldn't mind understanding it better myself.

Automotive manufacturers for example, must have a very good idea of how long an oil seal will last. Off-the-shelf oil seals from National or Chicago Rawhide however, don't come with any way of predicting seal life like bearing manufacturers do.

Contrast that with seals used in rotating equipment. Seals in pumps or compressors wear at a fairly predictable rate. These manufacturers will often give an expected life for their equipment, including seals. So I don't think it's a lack of ability to predict seal life.

Seal life is relatively predictable and consistant given typical applications. I'd say that some seals come with a fairly well known life, such as those in a specific piece of equipment. It's only when we get away from a known installation that seal manufacturers don't have a canned formula for seal life. They can probably predict with reasonable accuracy, whether or not a given seal will work and for how long in some application, but there's no canned formula for how predicting longevity.

Perhaps we don't have formulas because the type of seals and applications are so varied. That isn't to say it couldn't be done. Wear rates are generally determined by contact stress and velocity, though they are also highly dependant on the fluid being sealed, surface finish and hardness, etc... I think it's the larger number of variables that would need to be taken into account that prevents any easy formula from being created for all seals as an industry standard similar to the L10 equations for bearings.

 

[MikeHalloran]

In terms of life predictability, bearings and seals share this: All bets are off in the presence of misalignment/ contamination/ abusive installation practices, etc.

Beyond that, it's one of nature's miracles that properly installed ball bearings _have_ a statistically predictable life, and that someone was able to fit a usable equation to it.

I haven't messed with mechanical seals much, but I have messed with shear- seal valves, which I think are tribologically similar. What I found, for otherwise constant conditions, was that the normal pressure on the seal faces has some optimum value. Too little pressure, and the seal leaks excessively. Too much pressure, and the seal faces wear fast. The 'sweet spot' is pretty narrow; but if you can find it, the life becomes essentially infinite. Pressure above the sweet spot puts you in a range where the life is proportional to the volume of material available to be worn away. I don't know what the entire solution surface looks like, but the part that I do know about is not nice and neat.

Your guy has expectations that we can't meet directly, unless I missed a meeting.

Oh. Yeah. The few mechanical seals that I've replaced mostly didn't fail by wear of the sealing faces. In most cases, the bellows/ static seal behind the dynamic seal died of old age, and in some cases the resulting fluid leak went undetected long enought to cause a bearing failure that in turn fractured a seal face. I honestly don't know if there's a way to model progressive hardening and embrittlement of elastomers, but I'm pretty sure it's strongly dependent on the thermochemical environment.

How much research and analysis is your guy willing to pay for?







Mike Halloran
Pembroke Pines, FL, USA

 

[Artisi]

Remember that the L10 bearing life means that 10% of the bearing will fail before the calculated life is achieved - and it always seems that it's your bearings that seem to fall in the 10% range.

 

[RossABQ]

The L10 is based on essentially a metal fatigue failure, which is predictable. Seals simply wear, which is highly indluenced by the environment, RPM, starts/stops, and the other factors listed above. It would be the same as asking "how long do brake pads last?".

 

[Milkboy]

Thanks for your responces guys
Much appreciated


M

-
Milkboy