Carl, if you make this shaft and measure, say, .0005 TIR, and the eddy current meter says that there is .002 TIR, then the difference is electrical run out. It is not real run out but the system can't tell the difference. Since most steel is pretty homogenous the usual source is variations in the surface finish. These can be eliminated by polishing and burnishing.
Interestingly the spec called for this to be measured at the bearing areas. This meaningless because the sensor is not in the bearing area. It's next to it, as close as possible but may be a foot away.
Bentley Nevada has most of the market for this stuff. The turbines I work on may have there sensors for temperature, radial and axial runout, axial growth, and speed. You might contact them for advice on preparing this surface.
Ok, tdmidget, that I understand, it's electrical in the sense it's read off a sensor and that's just as I thought.
I guess it's the hard headed Dutch/German in me that seeked a reasonable answer.
Actually it's Cuda that is seeking the info on electrical runout while I am trying to understand what electrical has to do with runout.
Problem solved, thanks.
Basically a dial indicator can detect if the surface of the shaft is smooth, and thus in your perspective, it has no runout.
An electrical device can also measure the runout of a shaft, the BIG difference is that the electrical device can't JUST see the outside surface, ( that you are measuring with a dial indicator) the electrical device "sees" right into the steel itself. It can see any inconsistancies INSIDE the steel.
Now in the ideal world, the sensor wouldn't see, or care about anything except the outside of the steel, like your dial indicator does.
But in the real world, the sensor Does see the structure of the steel, under the surface and will get fooled into thinking it is seeing runout that is not there.
The inconsistancies that your dial indicator cannot see, but the electrical device CAN see, are best labled as "electrical runout" because it is a "runout" in the sense that an Electrical device will be fooled by it.
So, there is nothing "New Age" at all about the naming of the phenomonon.
Ok, I can buy the sensing internal mass but why can't they do a base line run and not have to burnish the metal? All the burnishing can do is affect the surface and a small depth of the metal.
I see no need to require a surface burnish, can you explain why.
We are getting close to the end now. If a base line reading will determine the "electric runout" without burnishing we are done and the patented process is a hoax.
"electrical device "sees" right into the steel itself."
Steel is electrically conductive so as such, the surface forms a nearly
perfect faraday shield and no electrostatic method can investigate
any features below the surface.
Either this is a) complete snake oil, or b) completely mis-named.
I could imagine a *magnetic* sensor that measures features in the
shaft's local magnetic field, that would suggest features in the
body of the shaft.
But now, that would be "magnetic runout" and not "electrical runout."
If you slow roll the shaft in a set of precision v blocks using a good dial indicator you can see the mechanical runout
what about tri lobe?
discussions like this are so far over my head i lost control when i got a chance to say something.
we use to do a lot of centerless grinding and there is a phenomenon that makes a kind of triangle cross section that doesn't show in the v-bloke roll test.
If you had followed this thread, you'd realize that the device in question is an electrical device, that is using eddy currents, not an electrostatic field.
Originally Posted by jim rozen
Now you can get all pedantic, and argumenative, (once again) or try to be helpful, and explain something in a manner that will be helpful to the person with the question.
Are you suggesting that the eddy current device is hydraulically powered? No, it is electrically driven.
Are you trying to be helpful, or not?
It's NOT ELECTRICAL RUNOUT.
Nothing ELECTRICAL is running out. No ELECTRICAL
parameter in the shaft is being measured. No ELECTRICAL
performance is being improved, modified, adjusted, or
Whoever invented that specification is a MORON and should
go back to serving coffee at starbucks.
You could use a federal gaging head which is ELECTRICAL
to measure the physical runout on the shaft but that does
make the resultant number an ELECTRICAL runout. No.
You could sweep the OD with a wire hooked up to a light bulb
and claim you were measuring an ELECTRICAL property of
the shaft operation. You could do that and everyone would
laugh at you for maintaining that position.
Look, non-destructive testing is a huge industry issue and they
do things like measure magnetic behavior of parts, and if you
could invent a widget that you hold up to a steel shaft in
motion and tell reliably that there's some sort of defect or
non-uniformity in the shaft that can be improved, the folks
in that discipline would kneel down and kiss the hem of your
garment because such a thing, short of X-raying the part,
to my knowledge simply does NOT EXIST at this time.
So yes, I'm trying to help. I'm suggesting to the original poster
that he watch out carefully because it looks like a bit of snake
oil is being peddled here.
wippin boy, I sure do like your avitar, he is cute. We have a little doll that has that pouting face and she is cute as hell. We call her Mary Lou our adopted daughter.
Ok, back to the problem, thinking about it I am thinking it does not sense the irregular mass but just the vibration of the shaft due to the bearing play. Since they have to take a base line reading by rotating the shaft slowly in V blocks or in the bearings themselves.
I went back and read several sites about electrical runout and none of them explain what electrical runout is. They describe how to detect it and how to burnish the shaft in the area of the sensor but that's all. Some say that it will detect vibration and/or movement in the shaft.
So with that I am jumping back to my thought that all they are looking at is reading bearing wear to tell them when to replace the bearings before it goes WHAM BANG. They use an eddy current sensor or a capacitance sensor to sense the electrical runout.
As far as I can tell the term "electrical runout" pertains to the use of electronics to determine when the shaft exceeds a certain vibration or mechanical runout and the area where the sensor runs has to be burnished to a smoothness to eliminate any electrical impulses in the sensor.
EDIT: it seems I was composing and posting while Jim was and I totally agree with Jim after reading about 6 sites that described everything with great accolades but no where did they say what electrical runout is.
All it is at best is an expensive glorified means to detect excessive radial and axial motion in a shaft with electronic apparatus.
Jim, I'm amazed that you know so much...I'd suggest you publish a paper and set the rotating equipment world on edge. You'll be a rich man. Do you really think that scores of PhD-level mechanical engineers with a combined total of thousands of years of experience would dream up the existence of electrical runout? I'll bet you have poured more coffee than they ever will. I have dealt with enough of the top engineers at the world's largest oil companies and they are astoundingly brilliant. Not only do they have the best educations, they also have decades of hands-on experience.
The annual IEEE-PCIC conference is coming up in two weeks - man yof the people who wrote those specs will be there. I suggest you take the mike in front of a thousand people and tell them just why they are wrong.
Carl, electrical runout has nothing do with electronics. It exists whether you can measure it or not. Or rather, the moment you endeavor to measure the vibration of a rotating shaft with a non-contact device (which pretty much will need to be electronic) electrical runout will appear.
I think the main problem is you guys are thinking in terms of great quantities of electrical force, and it doesn't make sense to you that it could exist. In fact, we are talking about relatively minute amounts - but those small amounts matter because the monitoring devices are so sensitive.
If something - anything - has molecules then it has electrical forces within it. It's not hard to reason that a 97" long alloy steel shaft will have some non-homogeneity. Some areas will have a little more of one component than the other - and they will have very slighly different electircal properties, as read by a probe, than others.
Originally Posted by Cuda
I have little to add regarding electrical runout and will accept what has been stated as a definition as true.
Having said that, looking at the original post here from Cuda, let's assume that he does a perfect job of machining and the result is no mechanical runout. (0.0)
Will that also guarantee that there will be no electrical runout? Put another way, is the electrical runout a result of mechanical runout or does it arise from some electrical quality in the piece of stock that Cuda chooses to machine independent of the machining?
It would be a shame to have a perfectly machined piece rejected due to some hidden flaw within the metal itself. Then it would seem that the choice of material would be extremely important.
Just curious since I willingly admit that I've never heard of electrical runout before.
"It exists whether you can measure it or not."
Er, I think we're getting close to the problem here....
"If something - anything - has molecules then it has electrical forces within it. It's not hard to reason that a 97" long alloy steel shaft will have some non-homogeneity. Some areas will have a little more of one component than the other - and they will have very slighly different electircal properties, as read by a probe, than others."
I will repeat this. I will put it in caps for emphasis. It is a basic element of physics.
ONE CANNOT DETERMINE ANY INFORMATION ABOUT STRUCTURE INSIDE OF A
CONDUCTIVE BODY, USING AN EXTERNALLY APPLIED ELECTRIC MEASUREMENT.
This is farady's law. One cannot detect or influence charges inside a closed
The. Shaft. Is. A. Closed. Conductive. Surface.
No electric measurement can penetrate into the body of the shaft or give
you any information about what is inside of it.
Don't like that? Take it up with Mr. Faraday.
Maybe the guys at this conference are real whiz-bang gizmo geniuses. Maybe the
technique in question uses something beside an electric measurement. But if so
they sure dropped the ball at naming it.
Originally Posted by jim rozen
Given that Mr. Faraday (and you as well) are correct, then it would seem that a machinist taking in a job where a specification exists, not only for mechanical runout but for electrical runout as well, would be taking a big chance on getting his work rejected for forces beyond his control. (no pun intended).
This might be useful information. "Electrically induced runout" is one of the "glitches" described.
That site said much the same as the others I have read.
It all boils down the the fact that "Electrical Runout" is a term used to describe the use of a sensor to read any radial and axial runout or vibration in a shaft. It seems it is used mostly in highly balanced rotor/shaft assemblies.
As Jim said, it has nothing to do with electricity in or around the shaft, nor the mass internal or external. It has to do with electronic sensing of the shafts movement.
It does require a treatment of the shaft where it is work hardened and trued in the area where the sensor reads.
Electrical runout is not what one would think it is, it is just mechanical runout measured electronically.
Jim Rozen is being extreme.... he, with his physics background, knows very well that the shaft is NOT a (perfectly) conducting surface.......
That being the case, it has a 'skin depth" at any frequency, which represents the depth to which an outside signal can "see" or "be seen" in the material. It really is the effective depth to which current is induced in the surface.
So, yes, the outside signal from the probe WILL 'see" into the material some distance. If the conductivity (resistance) varies, the induced currents will not be uniform all around, and "runout" will be indicated despite the fact that there isn't any real runout.
All this crazy 'electrical runout" specification business means is essentially just teh following:
"We don't want our probes to detect a baseline runout over XXX amount for any reason, whether the surface is bumpy, or the material is non-homogeneous, or whatever. To cover our butts, we spec this so we can use these probes without thinking, and not have to characterize every shaft individually".
To be sure the runout is truly within spec, you'd have to test it with a similar probe with the shaft on centers, etc.
And I agree the term "electrical runout" is misleading, bogus, etc. However, it also seems to be the standard industry term for the effect, so this might be a good time to get used to it.
The skin depth, at any reasonable frequencies, will show that the conductivity
for steel (which is of course much less than copper) is very very thin indeed.
Probably a couple of microns at most. If you work the equations you find that
the resisitivity can be pretty large, and the skin depth still quite small.
Steel boxes make excellent faraday cages.
And That is why a thin piece of steel will completely block a magnet!
Originally Posted by jim rozen
Yes, Jim I believe you now, I won't even try it to see if as magnet can attract anything if a thin piece of steel will block out the magnetic field. You must be right...
Magnetic lines can't possibly penetrate steel to any depth can they?
Are you REALLY trying to tell everyone that a Magnetic field is completely blocked by a thin layer of steel?
So it's really MAGNETIC runout?
Not ELECTRICAL runout?
Which is it please. I mentioned above, magnetic
methods are used often to do non-destructive testing
of steel parts.
But we're not talking about magnetic runout.
The name of the game here is ELECTRICAL RUNOUT. Your phrase, not mine sir.
My comments above still stand.
The most recent paper on this I can find is that one from 1975. Does not seem
to be much interest in this technique, aside from the guy who patented it,
Nobody seems to be able to say exactly what is being measured here, or
exactly what the benefits are. Aside from the broad "it's really important"
and "lots of folks use this" there's not much.
Is it an electric measuremetn? Is it a magnetic measurement? Does it probe
only the surface, or does it give infomation about the bulk properties? Is there
any predictive value in it? How does the 'burnishing' mentioned in the abstract
differ from simply shot-peening the surface to reduce crack propogation.
How far in does it look? Does it work on non-ferrous materials? Why is there
such a long discussion about reduction of overall magnetization in the measured
Lots of questions. Not many answers.
Snake oil? Maybe. Anyone can get a patent. I've never seen this term used in
industry. But then I'm probably too critical overall.