I read this article:
http://www.eere.energy.gov/industry/bestpractices/energ...article_id=188?print

A few questions arise:
1) Many literatures in internet suggests to turn the shaft once a week. Isn't this toof requent?

2) Also many suggest to alternately run the machine with an installed spare unit instead of to run as duty/standby. This is alright for the motors but will damage the pump mechanical seals as has been discussed extensively in this forum. Apparently, there is a clash of requirements here.

3) It said that the vibration from the running machine can be trasnmitted to the standby unit and thus cause false brinelling. How much vibration transmitted from the running machine can the standby machine withstand?

Josh:

Yes, way too frequent. My usual recommendation is that the shafts are turned at least 3-4 times per year. If you are complying with good storage techniques, this should be frequent enough.

In this case, a lot of the information used by the editors of the US DOE's Energy Matters material is actually from motor repair shops or EASA. Not all of the information from either source is always the best for field applications.

None. Just as with motor storage, you want the machine located as far from heavy traffic as possible or other sources of vibration. If you have a measureable level of vibration at the bearings with the machine de-energized, it will eventually turn into an issue.

MotorDoc

I was more interested in turning the shafts of motor and its driven pump which are on standby mode. I guess 3 to 4 times per year still applicable.

Can the motor be damaged by frequent equipment switchover (or alternate running) such as biweekly?

You are saying NONE of vibration from the running machine will be transmitted to its standby machine, aren't you?

Your last sentence appears to suggest that there should be no measurable vibration at the bearings with the machine de-energized. Is this correct?

Josh:

I am enjoying an actual connection to the internet within a steel mill, so I will make this fairly brief.

Concerning frequent equipment switchover: Starting an electric motor is the most stressful time of it’s operation. This is where most failure modes of the winding and mechanical condition are triggered. Meaning that if there is a weakness or degraded insulation system, this is where it will most likely fail, for instance. During the starting process, inside the electric motor, the windings flex, the rotor is seeing line frequency slip and a resulting high current (seen on the motor leads as 4-8 times running current for general purpose motors) and there are torsional forces on the frame, feet and mounting. As we are discussing a pump, there are stresses in the shaft, coupling, impellor and in the seal. If the machine has been sitting for a while, the seal will see potential problems whether it is a mechanical or packing system.

For a mechanical seal, the actual lubrication for the seal surfaces is from the vaporization of the fluid that the pump is pumping or from the seal enclosure if nitrogen, potable water or oil is provided. In effect, the seal literally rides on a cushion of vapor, and that is how it keeps its seal. In a packing system, some amount of fluid is allowed to travel through the packing material through a lantern ring in order to maintain cooling and lubrication of the surfaces. If either is allowed to rest for too long, the seal surfaces can be damaged during the starting process.

So the answer to the question is less one of should you do it as it is one of what are the operating conditions. If the motor and pump combination are running continuously then it does cause some stress to shut one down and start the other. If the pumps are cycled as part of their operation then it has very little impact on the operating stresses that one or the other machine will see. In the first case, it is a judgment call, in the second case, it is a good practice to switch between machines, bearing degradation is just one of the considerations.

What I meant by ‘none’ is that ideally you want no vibration transmitted. How the vibration will impact the bearings will depend on the phase of the vibration, and the resulting amplitude, at the feet of the motor. When I have performed forensic analysis of bearing failures on stationary machines due to vibration, I would usually check the vibration at the bearing housing with the machine de-energized. The goal is to eliminate as much of that vibration as possible, including using phase readings to help identify the location of the cause of the vibration (ie: nearby machine, through piping, traffic, etc.). In a few cases, I have recommended/installed vibration isolators in order to reduce the impact of a vibration that is not easily corrected, stiffened the source of the vibration and a few other methods. Stiffening the affected motor actually has an opposite effect as the vibration occurs between the surfaces of the bearings, but the energy can be greater. The vibration also has the effect of moving the lubrication out of the way (there is a few mils of oil between the bearing surfaces).

The exact amount of vibration should be negligible. There are a few standards in development through ISO TC108 that are primarily geared towards mobile and at-sea machines. I have drafts of those somewhere, but, if I remember, they do not identify actual values. I will have to check on that.

Howard

Josh,

I don't think ocasional rotation a shaft of a stored/shutdown machine could preventing false brineling in bearings. It can be beneficial for maintaining a lubricant film in bearings to minimize corrosion or to minimize sagging of large rotors.

I would use 1.0 g peak acceleration as a guide to avoid false brineling, since any components would "rattle" with higher vibration. I would like to see any test data or study that would indicate a suitable vibration criteria.

Walt

Although it's not as effective as removing the vibration, I think that rotating the shaft has some benefit because it prevents too much damage at one location. I'd rather have 100 tiny (miscroscopic?) false brinneling sites than 3 or 4 large false brinneling sites.

I think you need to shoot much lower than 1 g peak. 1g pk/0 is ~ 1 ips pk/0 at 3600. I wouldn't want my running machines to see that, much less my secured machines.

Thanks for the responses above.

MotorDoc

"So the answer to the question is less one of should you do it as it is one of what are the operating conditions. If the motor and pump combination are running continuously then it does cause some stress to shut one down and start the other."

I understand this paragraph up to above. In don't quite understand the following sentences:

"If the pumps are cycled as part of their operation then it has very little impact on the operating stresses that one or the other machine will see." Are referring to alternate running of the machines here?

"In the first case, it is a judgment call, in the second case, it is a good practice to switch between machines, bearing degradation is just one of the considerations."

Your first case is to run machines as duty/standby and your second case is alternate running of the machines, isn't it?

If I understand your cases correctly, I am confused because you said in the first paragraph that the frequent start up during the alternate running of machines will give a lot of stresses. And in the last sentence of the 3rd paragraph, you said it's a good practice to switch between machines.

Pls clarify and enjoy the internet connection from the steel mill.

Walt

By large rotors, do you mean slender rotors? Any guide on the length to diameter (L/D) ratio which may be subjected to sagging?

Good tip on lubing the bearings by turning the shaft. If the shaft remains stationary for 2 months, can rusting develop?

Pete

I guess 1 g peak is a maximum vibration tolerable. Should we aim half of this?

What do you mean by "secured" machines? Secured means the installed standby or stored spare in the warehouse?

This message has been edited. Last edited by: Josh,

Actually, had to wait to get home.

A lot of machines are cycled - meaning started and stopped during their normal operation, such as compressors. As a seperate example, take a fire pump. They are supposed to keep pressure on a main. They do this by running, stopping, then running (in some applications). This starting and stopping may be occassional or frequent, depending on the system. In this case, it would not hurt to alternate between machines because they would normally see these starting stresses.

In other cases, the pump (or other machine)runs continuously, 24/7. In this case, you may alternate, but not very frequently (ie: monthly) or you may opt not to shut down and alternate the machine at all.

There are other things to consider, of course. One of those is that the seals of the machine may be damaged if not run for a long length of time.

Howard

Ok MotorDOc, that's clearer now.

Yes, I agree we should run pumps with soft packings alternately to keep the seal wet.

I intend to run pumps with mechanical seals on duty/standby. So run the duty pump as long as possible such as 2 months and test run the standby machine for 8 to 24 hours as a failure finding task and to do condition monitoring eg vibration survey.

You said the seals may be damaged if not run for long time. Do you mean the mechanical seals?

This message has been edited. Last edited by: Josh,

"Secured" means stopped/idle.

I would hesitate to tell you a limit for vibration on idle machines. I think I would investigate a little if I noticed more than 0.1 ips pk/0 overall or 1 g's true peak on the housing of idle machines, but that doesn't mean that above these levels is a problem or less is safe.

You should make a practice to inspect removed bearings to give a clue to what factors are degrading your specific machines in your operating conditions/environment. If you see multiple series of equal spaced marks... with distance between marks in a series equal to rolling element spacing, that's a good clue the cause was false brinneling. You can also see many examples in the literature from the bearing manufacturers.

Then you have to do some investigation as to what is causing the vib and whether it is more cost effective in the long run to live with it or eliminate it.

I have a few photos and spectra from one case where we saw false brinneling on a pump bearing and then measured the bearing housing on the idle pump and founnd 0.17 ips. We traced the vibration to be coming to the pump through the discharge piping and eliminated it by installing expansion joints. (in other cases, changes to pipe braces might be appropriate)

This message has been edited. Last edited by: electricpete,

I realized my response above was a little "to quick to the podium". The 1-g was for vertical vibration, since no relative vertical motion can occur below it. External vibrations can presumably induce horizontal and axial motion of the rotor/shaft relative to the bearing housing. Depending of bearing clearance and preload (if any), some micro-movement can be expected. If no lubricant film is present, then some wear over time can be expected. I did a recent Net search and there are several consistent explanations for the visual appearance and general cause of False Brinelling. I found that even the word was spelled differently; 2-N and 1-L, 2-L and 1-N, and 1- N and 1-L. Please don't criticize my spelling!
I did not see any reference to the the amount of vibration that would cause or not cause this effect. It appears that once you have False Brinelling, then you can measure vibration and assume that it was too high. Of course, perhaps the vibrations were higher in the past than they are now; a delema in picking a vibration criteria. I hope someone has scientific/engineering information to share on this topic.
Here are a few links:
http://www.skf.com/aptitudexchange/glossary.html#FalseBrinneling
http://www.emersonbearing.com/fail_falsebrinell.htm
http://en.wikipedia.org/wiki/Rolling-element_bearing

http://en.wikipedia.org/wiki/Brinell_hardness_test


http://www.akrongear.com/bearings1.htm#falsebrineling

Walt

So it looks like we need to monitor vibration on standby machines to ensure no substantial externally-transmitted vibrations exist, don't we?

Josh,

quote:

to ensure no substantial externally-transmitted vibrations exist

The key word is "substantial". If you don't know what the vibration limit should be, then why waste time and money on machines that are not running? I have not seen many instances of False Brinelling, so I suspect it does not occur very often. I think you could take a retro-active approach rather than be proactive. Retro-active means you wait for this to occur, and then take actions to prevent occurance again. I am still waiting for a better approach than this.

Walt

Ok, I will at least do once for the standby machines. If indeed no substantial vibrations, then it's ok not to do it every time.

We just found one set of pumps is affected by vibration from another pumpset nearby every ten seconds as per the vibration vendor.

I don't like the retro-active idea because it will affect some reputation and cost some money. Maybe can do this experiment in a lab.

Josh,

Rotating the shaft would not eliminate false brinelling, what I meant was we are just changing the position where false brinelling would occur thereby delaying or slowing down the its process.

I still believe that it is a good practice to rotate the motor bi-monthly.

If you want to test if the standby motor is affected by the vibration of the duty motor, why don't you perform one of cavemans tricks, the coin test and place it on the standby motor and see for yourself.

My warm Regards,

Ok, we can do the coin test after one month on standby along with the shaft turning task.

Btw, we found out that the standby motor may be affected by vibration not only from its running mate but also from the other machines nearby as mentioned above.