Circuit Model for a Rotary Contact?

[David Hodge]

Hi Everybody,

First post on the forum, this one's stumped me I'm afraid. I'm basically trying to model a rotary connection in an AC circuit, and I'm having a bit of trouble with it. Ideally I'd like a reasonable approximation with conventional circuit elements.

Basically I've got a stationary copper ring and a moving shaft that is vulnerable to lightning strikes. The contact is just a annular faying connection. There's an AC current (from a lightning strike) of variable phase/waveform at ~250 kHz which flows through the moving shaft and into the copper contact ring. I just need a way to model this little snippet of a circuit, hopefully yielding an inductance and impedance value that'll let me size the connection within ~20% or so (so I know what to buy for testing).

Any help is vastly appreciated, especially if you can point out any literature that might describe an empirical model and/or some closed form solution.

I've attached a simplistic diagram of such a rotary interface to try to clarify my question.Thank you!
-DH

 

[berkeman]

Hi Everybody,

First post on the forum, this one's stumped me I'm afraid. I'm basically trying to model a rotary connection in an AC circuit, and I'm having a bit of trouble with it. Ideally I'd like a reasonable approximation with conventional circuit elements.

Basically I've got a stationary copper ring and a moving shaft that is vulnerable to lightning strikes. The contact is just a annular faying connection. There's an AC current (from a lightning strike) of variable phase/waveform at ~250 kHz which flows through the moving shaft and into the copper contact ring. I just need a way to model this little snippet of a circuit, hopefully yielding an inductance and impedance value that'll let me size the connection within ~20% or so (so I know what to buy for testing).

Any help is vastly appreciated, especially if you can point out any literature that might describe an empirical model and/or some closed form solution.

I've attached a simplistic diagram of such a rotary interface to try to clarify my question.Thank you!
-DH

Welcome to the PF.

Use the EN 61000-4-5 Surge waveforms to model lightning strikes to electrical supply lines. And can you use the traditional SPICE relay model to model the contacts?

 

[tech99]

Hi Everybody,

First post on the forum, this one's stumped me I'm afraid. I'm basically trying to model a rotary connection in an AC circuit, and I'm having a bit of trouble with it. Ideally I'd like a reasonable approximation with conventional circuit elements.

Basically I've got a stationary copper ring and a moving shaft that is vulnerable to lightning strikes. The contact is just a annular faying connection. There's an AC current (from a lightning strike) of variable phase/waveform at ~250 kHz which flows through the moving shaft and into the copper contact ring. I just need a way to model this little snippet of a circuit, hopefully yielding an inductance and impedance value that'll let me size the connection within ~20% or so (so I know what to buy for testing).

Any help is vastly appreciated, especially if you can point out any literature that might describe an empirical model and/or some closed form solution.

I've attached a simplistic diagram of such a rotary interface to try to clarify my question.Thank you!
-DH

The contact is shown in Section A-A and it looks to be very small area. The copper strip also looks to be very small cross section. To give a feel for the problem, a typical lightning strike involves a peak current in the order of 10kA, and typical lightning conductors have a cross sectional area of 50mm^2. A seem to remember that a copper wire of 1/8 inch diameter will survive about half the strikes. I suggest reading the EN standards as quoted by Berkeman. I tend to feel that a small clearance between large steel rings, so it relies on sparking, might be more reliable.

 

[sophiecentaur]

The contact is shown in Section A-A and it looks to be very small area. The copper strip also looks to be very small cross section. To give a feel for the problem, a typical lightning strike involves a peak current in the order of 10kA, and typical lightning conductors have a cross sectional area of 50mm^2. A seem to remember that a copper wire of 1/8 inch diameter will survive about half the strikes. I suggest reading the EN standards as quoted by Berkeman. I tend to feel that a small clearance between large steel rings, so it relies on sparking, might be more reliable.

It is a difficult problem and I'm not sure that you can get away with your assumptions. I feel that a single contact with a small area isn't like a 'small wire' because it is more like a discontinuity. Thermally, it is in good contact with a much bigger mass and there is an equivalent Capacitance across the non-contacting area around the actual contact. Also, if the slip ring is moving past the contact, the cooling would be even better. That would increase the RF current capability and spread out any arcing damage (when it's running).
Also, I wonder whether the 10kA figure of a typical strike is necessarily applicable here. Would the contact necessarily be the only path for the strike? But I guess it is a 'constant current' source. ("just try and stop me")
There are many commercial lightning protection devices and, as this isn't a particularly fragile system, there is bound to be something that could protect your contact. But perhaps cost is a factor?

 

[David Hodge]

Hi Everyone,

Thanks very much for your replies, let me see if I can address them:

@berkeman: Thanks for the waveform model, that's quite helpful; I'm afraid I don't follow your logic in modeling the contact as a relay though?

@tech99: 10kA seems about right for a median amperage value, I am, indeed worried about sparking; unfortunately in this instance, a traditional grounding brush is contraindicated because of corrosion and temperature concerns. Similarly the graphite rotary contacts that are used in wind turbines are no go because of the particulate they generate in natural abrasive wear.

@sophiecentaur: This probably wouldn't be the only path for lightning but it's definitely the preferred one; i.e. it'd be great to be able to size the contact and/or tweak other variables to make it the path of least impedance. I actually got this idea from a commercial vendor (link below). I'm mostly just at the trade study phase in the design, and my major variables are inductance and impedance across the rotary contact.

I suppose that a decent clarification of the question based upon your reponses would be:
Assuming Isothermy, a constant rotational velocity, and a constant, fairly small, compressive force applied from the stationary ring on the rotating shaft, is there a decent way to find inductance and impedance approximations? Perhaps an approximate solution of Maxwell's Equations integrated about the annular contact?

Thanks again,
-DH

http://www.omegashielding.com/category/contact-ring

 

[Tom.G]

Two possible research ideas.
Find out what the Amatuer Radio (Ham Radio) operators use on their rotating beam antennas. http://www.arrl.org would be a place to start.
Perhaps instead of using a sliding contact you could use a rolling contact; on the order of a roller bearing, lubricated with conductive grease if needed.

 

[sophiecentaur]

a constant rotational velocity,

Whatever speed the contact can be moving at, it will not affect the electrical characteristics of the system, which depend on the speed of EM waves.
Most sliding electrical power contacts are made with carbon brushes. Lots of advantages - including a large contact area with low friction. And they are readily available in all shapes and sizes.
It would be easy just to protect the contact (and the whole device) with an appropriate lightning protection circuit. After all, would it only be the contact that would suffer from a serious strike?

 

[David Hodge]

Thanks again, everyone, for your replies.

@Tom.G: Your suggestion got me thinking and after some digging I found the following IEEE article
about analytical modeling of slip rings. I'll http://ieeexplore.ieee.org/document/4072999/?reload=true
I think that I can extract the necessary theory from it or it's citations, but I'm going to have to digest it for a while.